BMI1 enhancer polymorphism underlies chromosome 10p12.31 association with childhood acute lymphoblastic leukemia

Abstract

Genome-wide association studies of childhood acute lymphoblastic leukemia (ALL) have identified regions of association at PIP4K2A and upstream of BMI1 at chromosome 10p12.31-12.2. The contribution of both loci to ALL risk and underlying functional variants remain to be elucidated. We carried out single nucleotide polymorphism (SNP) imputation across chromosome 10p12.31-12.2 in Latino and non-Latino white ALL cases and controls from two independent California childhood leukemia studies, and additional Genetic Epidemiology Research on Aging study controls. Ethnicity-stratified association analyses were performed using logistic regression, with meta-analysis including 3133 cases (1949 Latino, 1184 non-Latino white) and 12,135 controls (8584 Latino, 3551 non-Latino white). SNP associations were identified at both BMI1 and PIP4K2A. After adjusting for the lead PIP4K2A SNP, genome-wide significant associations remained at BMI1, and vice-versa (P_meta<10⁻¹⁰), supporting independent effects. Lead SNPs differed by ethnicity at both peaks. We sought functional variants in tight linkage disequilibrium with both the lead Latino SNP among Admixed Americans and lead non-Latino white SNP among Europeans. This pinpointed rs11591377 (P_meta=2.1×10⁻¹⁰) upstream of BMI1, residing within a hematopoietic stem cell enhancer of BMI1, and which showed significant preferential binding of the risk allele to MYBL2 (P=1.73×10⁻⁵) and p300 (P=1.55×10⁻³) transcription factors using binomial tests on ChIP-Seq data from a SNP heterozygote. At PIP4K2A, we identified rs4748812 (P_meta=1.3×10⁻¹⁵), which alters a RUNX1 binding motif and demonstrated chromosomal looping to the PIP4K2A promoter. Fine-mapping chromosome 10p12 in a multi-ethnic ALL GWAS confirmed independent associations and identified putative functional variants upstream of BMI1 and at PIP4K2A.

Introduction

Acute lymphoblastic leukemia (ALL), which comprises approximately one third of childhood cancer diagnoses, is driven by disruption of hematopoietic regulatory pathways leading to the proliferation of immature lymphocytes. Although the mutational burden of ALL is relatively low, the disease is characterized by somatic alterations in B-cell differentiation genes, as well as those involved in the Ras pathway and in cell-cycle control ^{1, 2}.

Heritable variation in genes associated with these same functional pathways is a risk factor for childhood ALL. Genome-wide association studies (GWAS) of ALL have identified SNP associations in genes involved in hematopoiesis and B-cell development, including ARID5B, IKZF1, CEBPE, and GATA3, as well as the cell-cycle control gene CDKN2A ^3–6. Association has also been detected at the chromosome 10p12.31-12.2 region, with neighboring peaks located at PIP4K2A and upstream of the proto-oncogene BMI1 ⁷, the latter region at which genome-wide significance has yet to be confirmed. The chromosome 10p12 SNPs were found to be specifically associated with the high hyperdiploid subtype of ALL ⁸, however a functional role for these loci in leukemogenesis has yet to be determined.

In a recent California-based GWAS of childhood ALL (using the same study population as current analyses), we identified novel SNP associations at chromosome 17q12-q21.1, which harbors the hematopoietic transcription factor IKZF3, and at chromosome 8q24, within a gene desert that appears to interact with the MYC oncogene via long-range chromatin interactions ⁹. We also replicated previously reported loci, including genome-wide significant associations at the chromosome 10p12 region containing PIP4K2A and BMI1. Here, we aimed to discover functional variants underlying the chromosome 10p12 associations and to identify their likely gene targets. To this end, imputation-based fine-mapping was carried out across the two 10p12 association peaks in Latino and non-Latino white ALL cases and controls from two independent California-based childhood leukemia studies, the California Cancer Records Linkage Project (CCRLP) and the California Childhood Leukemia Study (CCLS), with additional controls from the Genetic Epidemiology Research on Aging (GERA) study. Capitalizing on ethnic differences in haplotype structure, investigation of SNPs in strong linkage disequilibrium (LD) with both the lead SNP in Latinos and the lead SNP in non-Latino whites in corresponding populations was performed to pinpoint likely causal variants at both peaks.

Materials and Methods

Study Subjects

The study protocol was approved by the Institutional Review Boards at the California Health and Human Services Agency, University of California (San Francisco and Berkeley), Yale University, and of all participating hospitals. Sample acquisition for the California-based Childhood Cancer Record Linkage Project (CCRLP) GWAS of ALL is described in detail in Wiemels et al. ⁹. In brief, newborn dried bloodspots (DBS) for cases and controls were obtained from the California Biobank Program, California Department of Public Health (CDPH), Genetic Disease Screening Program. Childhood ALL cases were identified through linkage between the CDPH statewide birth records (years 1982-2009) and the California Cancer Registry (CCR, diagnosis years 1988-2011), with controls randomly selected and matched on year and month of birth, sex, and race/ethnicity (non-Latino white, non-Latino black, Latino (any race), Asian/Pacific Islander, other). Additional controls were included from the Genetic Epidemiology Research on Aging (GERA) study ¹⁰. In this study, analyses were limited to Latino and non-Latino white subjects.

ALL subtype-stratified analyses were carried out in the California Childhood Leukemia Study (CCLS), an independent case-control study of childhood leukemia that is described in detail elsewhere ¹¹. In brief, patients with childhood leukemia under 15 years of age were identified and rapidly ascertained into the CCLS from California-based pediatric hospitals from 1995 to 2015. One or two controls were matched to each case on child’s date of birth, sex, Latino ethnicity, and maternal race as indicated on the birth certificate record. DNA samples for cases and controls were obtained from newborn DBS from the California Biobank Program Genetic Disease Screening Program, or from saliva or buccal swab samples. In the current study, we limited analyses to Latino and non-Latino white subjects with available genome-wide SNP array data (n=927 cases and 750 controls), and these were largely representative of the CCLS population as a whole.

Genome-wide SNP array data in the CCRLP

DNA was extracted from newborn DBS and genotyped on genome-wide SNP arrays as previously described ⁹. Briefly, CCRLP ALL cases and controls were genotyped using Affymetrix Axiom World LAT arrays. Following quality control filtering, 757,935 polymorphic autosomal SNPs were included in analyses. Genotype data for additional GERA controls, genotyped on the same Affymetrix arrays, were downloaded from dbGAP (Study Accession: phs000788.v1.p2). In total, 3133 childhood ALL cases (1949 Latino and 1184 non-Latino white) and 12,135 healthy controls (8584 Latino and 3551 non-Latino white) were included in association analyses. Case demographic data are included in Table S1. Case-control analyses were stratified by ethnicity, with SNP associations calculated using logistic regression and adjusted for the first ten ancestry-informative principal components, calculated using Eigenstrat ¹², to control for population stratification. After adjusting for the first ten principal components, genomic inflation factors were calculated for the Latino and non-Latino white association analyses, which revealed minimal inflation of test statistics (λ_Latinos = 1.034; λ_{Non-Latino whites} = 1.004). A fixed-effects meta-analysis was used to combine results from the ethnicity stratified analyses.

Fine-mapping across chromosome 10p12

Imputation was carried out across the chromosome 10p12 locus encompassing the two association peaks upstream of BMI1 (“BMI1 peak”) and at PIP4K2A (“PIP4K2A peak”). BMI1 is centered between the two association peaks, therefore we took coordinates 500Kb upstream and downstream of this gene. Imputation of this 1Mb region was performed using the Impute2 v2.3.1 software and its standard Markov chain Monte Carlo algorithm, with default settings for targeted imputation ¹³ and using 1,000 Genomes Phase 3 haplotypes for the imputation reference panel ¹⁴. Poorly imputed SNPs were removed, i.e. those with imputation quality (info) scores <0.60 or posterior probabilities <0.90. Association statistics for imputed and directly-genotyped SNPs were calculated using logistic regression in SNPTESTv2, using an allelic additive model and probabilistic genotype dosages ¹⁵. The effect of individual SNPs on ALL risk was calculated while adjusting for the first 10 principal components from Eigenstrat. Analyses were carried out separately in Latino cases and controls and in non-Latino white subjects. Meta-analysis was carried out using the program META ¹⁶.

Following identification of two physically separate association peaks at chromosome 10p12.31 and 10p12.2, we investigated whether they were independently associations with ALL or whether this was due to linkage disequilibrium (LD) between SNPs in the two peaks. Thus, association analysis across 10p12 was repeated using the same logistic regression model adjusted for 10 principal components and also adjusting for additive effects of the lead PIP4K2A peak SNP rs10741006, and then again adjusting for the lead BMI1 peak SNP rs12769953.

LD structure across 10p12 was assessed separately in Latinos and non-Latino whites using Haploview v4.2. Haplotypes were constructed using all SNPs with minor allele frequencies >0.05 among control subjects, and blocks were plotted using the default block definition of Gabriel, et al. ¹⁷.

Epistasis analysis

Pairwise interaction analyses were carried out between three BMI1 peak SNPs and three PIP4K2A peak SNPs for a total of 9 interaction analyses. The three BMI1 peak SNPs included the most significant SNP in Latinos (rs12769953), the most significant SNP in non-Latino whites (rs4397732), and the candidate functional SNP rs11591377 that is in high LD with the former two SNPs (r²=0.93 and 0.97 respectively). The three PIP4K2A peak SNPs included the most significant SNP in Latinos (rs10741006), the most significant SNP in non-Latino whites (rs7912551), and the candidate functional SNP rs4748812 that is in LD with the former two SNPs (r²=0.82 and 0.80 respectively). Each interaction model contained the main effects for each SNP and the interaction term between each pair of SNPs, as well as the first 10 principal components. Analyses were performed separately in Latino subjects and non-Latino white subjects. P-values are reported for the interaction term of each logistic regression model performed using the glm command in the R statistical environment.

Multi-ethnic linkage disequilibrium correlation analysis

To identify putatively causal variants in the BMI1 peak and the PIP4K2A peak, at each locus we sought SNPs that demonstrated strong LD with both the lead SNP in Latinos and the lead SNP in non-Latino whites in the corresponding reference populations, i.e. in Admixed Americans (AMR) and Europeans, respectively. Thus, for the BMI1 peak we looked for SNPs in strong LD with rs12769953 in AMR and with rs4397732 in European populations, and for the PIP4K2A peak we sought SNPs in strong LD with rs10741006 in AMR and rs7912551 in Europeans. The LDlink “LDproxy” tool, a publicly available application for SNP LD analysis in 1000Genomes Project Phase 3 genotype data ¹⁸, was used to calculate LD values (r² and D’) for all SNPs within +/−500Kb of queried variants. For AMR populations, we included the “MXL” (Mexican ancestry from Los Angeles), “PUR” (Puerto Ricans from Puerto Rico), “CLM” (Colombians from Medellin, Colombia), and “PEL” (Peruvians from Lima, Peru) datasets. For European populations we included “CEU” (Utah residents from North and West Europe), “TSI” (Toscani in Italia), “FIN” (Finnish in Finland), “GBR” (British in England and Scotland), and “IBS” (Iberian population in Spain).

Scatter plots were generated for the BMI1 peak and PIP4K2A peak SNPs, plotting the r² of SNPs with the lead Latino SNP in AMR populations against r² of the same SNPs with the lead non-Latino white SNP in Europeans. Triallelic SNPs were excluded. We included only those SNPs with P < 5.0 × 10⁻⁶ in the adjusted meta-analyses results for both the BMI1 peak and PIP4K2A peak.

Enhancer element analysis

The Roadmap Epigenome Browser ¹⁹ was used to identify enhancer elements overlapping the 10p12.31 and 10p12.2 association peaks. We assessed tracks for DNase I hypersensitive sites (DHS), histone 3 lysine 4 monomethylation (H3K4me1), and H3K27 acetylation (H3K27ac) in HSCs (“Mobilized CD34 primary cells”), lymphoblastoid cell line (LCL) GM12878, and in the HSC-like myelogenous leukemia cell line K562. Predicted gene targets of enhancer elements were investigated using EnhancerAtlas (www.enhanceratlas.org), a recently developed database containing a consensus of human genomic enhancers derived from data on 76 cell lines and 29 different tissue types ²⁰. In brief, enhancers are predicted based on at least three independent experimental tracks (e.g. DNase hypersensitivity, transcription factor binding, and histone modification), and enhancer gene targets are predicted using the Integrated Method for Predicting Enhancer Targets (IM-PET) algorithm.

H3k27ac HiChIP data analysis

H3K27ac HiChIP is a recently developed protein-targeting chromatin conformation method for generating high-resolution contact maps between active enhancers and their target genes ²¹. H3K27ac HiChIP data was downloaded for GM12878 cells and for the K562 cell line (Gene Expression Omnibus, GEO dataset GSE101498), and uploaded into Juicebox software for visualization of 3D chromatin interactions across the chr10p12.31-p12.2 region. Knight-Ruiz (balanced) normalization was used to remove Hi-C matrix biases as recommended ²². Tracks for the same data were also uploaded into the Roadmap Epigenome Browser to assess locus interactions.

MethylC-Seq Analysis

Methylation levels at CpGs across the BMI1 peak enhancer locus (overlapping rs11591377) and the PIP4K2A peak enhancer locus (overlapping rs4748812) were assessed using whole genome shotgun bisulfite sequencing (MethylC-Seq) data tracks in the Epigenome Browser ²³. MethylC-Seq data was assessed for HSCs (Mobilized CD34 primary cells) and a selection of different tissue types including fetal thymus, spleen, brain hippocampus, small intestine, liver, lung, gastric, adipose tissue, and ovary. At the rs11591377 (BMI1) enhancer, we assessed the proportion of methylated cytosine at 21 CpGs spanning a 2,215bp region, excluding analysis of CpGs with <10 sequenced reads in any tissue types. At the rs4748812 (PIP4K2A) enhancer, we assessed proportion of methylated cytosines at 7 CpGs spanning a 422bp region (again excluding CpGs with <10 reads).

Transcription factor binding analysis

UCSC Genome Browser was used to identify TFs found to bind to the BMI1 peak and PIP4K2A peak regulatory elements in ENCODE ChIP-seq analyses. To assess whether candidate SNPs altered any TF binding sites, we used the “TFBIND” software (http://tfbind.hgc.jp) ²⁴, inputting an 11bp sequence that included the SNP itself (either risk or protective allele) plus 5bp up- and 5bp downstream sequence.

ENCODE ChIP-Seq analysis of K562 cells

For TFs predicted to bind to the BMI1 peak locus (at rs11591377), we downloaded available BAM files for corresponding ENCODE ChIP-seq experiments in K562 cells (from http://hgdownload.cse.ucsc.edu/goldenPath/hg19/encodeDCC/wgEncodeSydhTfbs/), including for ARID3A, BHLHE40, CEBPB, CTCF, c-Jun, c-Myc, p300, GATA2, JunD, MAFF, MAFK, MAX, TAL1, and ZNF143. BAM files for MYBL2 ChIP-seq experiments in K562 cells were downloaded separately (from https://www.encodeproject.org/experiments/ENCSR162IEM/), as these data were only released in August 2017. For each TF, separate BAM files were downloaded for each of two ChIP-seq biological replicates. BAM files were imported into the SNP & Variation Suite (SVS) software (Golden Helix), and visualized using the GenomeBrowse tool. For each TF, the number of sequenced reads was counted for the rs11591377 risk allele (G) and protective allele (A) in each biological replicate, and a binomial significance test used to calculate P values for the distribution of risk vs. protective allele reads compared with the expected distribution of 1:1. Fisher’s method was used for meta-analysis of P-values, using the R statistical package ‘metap’. To confirm that K562 cells are heterozygous for SNP rs11591377, we extracted DNA from an aliquot of cells (provided by Dr. Neil Shah, UCSF) using the Qiagen DNA Blood and Mini Kit, and genotyped using a Taqman assay for rs11591377 (Assay ID: C_____85910_10) on a Droplet Digital PCR machine (BioRad).

Replication and cytogenetic subtype analysis in the CCLS

Cytogenetic subtype information was not available for CCRLP cases. Therefore, to assess subtype-specific associations of the lead chromosome 10p12 SNPs, we utilized data from the CCLS, an independent case-control study of childhood leukemia ¹¹. SNP genotype data was available for 927 ALL cases (589 Latino, 338 non-Latino white, Table S1) including 238 with high hyperdiploidy (HD) (156 Latino, 82 non-Latino white), as confirmed by FISH or G-banding ²⁵, and 750 controls (506 Latino, 244 non-Latino white). DNA was extracted from neonatal DBS, saliva, or buccal cells, and genotyped on Illumina HumanOmniExpress or HumanOmniExpressExome genome-wide SNP arrays ²⁶. SNP imputation was performed using Impute2, as described earlier for analysis in CCRLP. Multidimensional scaling (MDS) components were calculated using PLINK1.9 ²⁷ to control for population stratification. For this analysis, we included six SNPs in total: the lead Latino and non-Latino white SNPs and putative functional variants at both the BMI1 and PIP4K2A peaks (i.e. rs12769953, rs4397732, rs11591377, rs10741006, rs7912551, and rs4748812). Association of each SNP with ALL risk (overall and with HD-ALL subtype) was tested separately in Latinos and non-Latino whites using logistic regressions assuming an allelic additive model, adjusting for the first seven MDS components. A fixed-effects meta-analysis was used to combine results from the ethnicity stratified analyses, separately in overall ALL cases versus controls, in HD cases versus controls, and in non-HD cases versus controls. Cochran’s Q-test was used to test for heterogeneity between Latino and non-Latino white studies.

Results

SNP genotype imputation across chromosome 10p12 in Latino and non-Latino white ALL cases and controls followed by meta-analysis of ethnicity stratified case-control analyses confirmed two ALL association peaks, with multiple SNPs at each peak reaching genome-wide significance (P < 5.0 × 10⁻⁸). The strongest association was at a region encompassing the 3’ end of PIP4K2A at chr10p12.2 (hereafter the “PIP4K2A peak”). The lead PIP4K2A peak SNP in the multi-ethnic meta-analysis was rs10741006, with P_meta = 5.78 × 10⁻¹⁹ and an odds ratio (OR) of 1.36 (95% confidence intervals, CI: 1.30-1.43). In ethnicity-stratified analyses this was also the most significant variant in Latinos (P = 3.19 × 10⁻¹²) and was strongly associated in non-Latino whites (P = 7.35 × 10⁻⁹); however, the lead SNP in non-Latino whites was rs7912551 (P = 7.41 × 10⁻¹⁰) (Table S2).

The second association peak was located ~ 450Kb upstream of PIP4K2A, between genes DNAJC1 and BMI1 at chr10p12.31 (hereafter the “BMI1 peak”). The lead SNP was rs12769953, with P_meta = 1.33 × 10⁻¹³ and OR = 1.32 (95% CI: 1.25-1.39). This was also the most significant variant in Latinos (P = 2.99 × 10⁻⁹) and was associated in non-Latino whites (P = 3.56 × 10⁻⁶) (Table S2). The lead SNP in non-Latino whites was rs4397732 (P = 2.05 × 10⁻⁶), which was in very strong LD with the lead Latino SNP (rs12769953) in European reference populations (r² = 0.98, D’ = 0.99) and in relatively strong LD in AMR reference populations from the 1000 Genomes project (r² = 0.82, D’ = 0.93).

Assessment of LD structure across the chromosome 10p12 loci in Latinos and non-Latino whites revealed slight differences in haplotype structure, in particular showing shorter haplotype structure in Latinos at the PIP4K2A peak (Figure S1).

Two chromosome 10p12 association peaks associated independently with ALL

After adjusting for the lead PIP4K2A SNP (rs10741006), associations at the BMI1 peak remained genome-wide significant for several SNPs, including the lead Latino and lead non-Latino white SNPs (rs12769953 and rs4397732, respectively) (Table 1, Figure 1). Similarly, after adjusting for the lead SNP in the BMI1 peak (rs12769953), rs10741006 in the PIP4K2A peak retained genome-wide statistical significance (P = 2.21 × 10⁻¹⁶), thus confirming independent associations between these two loci and ALL risk (Table 1, Figure 1).

Table 1.

Lead SNPs at the independent chromosome 10p12 association peaks upstream of BMI1 and at PIP4K2A, in analyses of acute lymphoblastic leukemia cases and controls from the Childhood Cancer Record Linkage Project (CCRLP), and additional controls from the Genetic Epidemiology Research on Aging (GERA) study. Results adjusted for the lead SNP at the alternate peak (i.e. BMI1 SNPs adjusted for lead PIP4K2A SNP, and vice-versa).

Peak	SNP	Chromosome 10 position (hg19)	CCRLP Latinos			CCLRP Whites			CCRLP Meta-analysis
			RAF controls a	P-value b	OR (95% CI) b	RAF controls a	P-value b	OR (95% CI) b	P-value	OR (95% CI)
BMI1	rs12769953	22407656	0.768	3.41 × 10−8	1.29 (1.20-1.39)	0.777	2.11 × 10−4	1.25 (1.13-1.38)	6.13 × 10−11	1.28 (1.21-1.35)
BMI1	rs4397732	22435841	0.770	6.64 × 10−8	1.29 (1.20-1.38)	0.776	1.14 × 10−4	1.27 (1.15-1.39)	6.28 × 10−11	1.28 (1.21-1.36)
BMI1	rs11591377 c	22423302	0.778	6.89 × 10−8	1.29 (1.20-1.39)	0.78	3.65 × 10−4	1.24 (1.12-1.37)	2.07 × 10−10	1.27 (1.20-1.35)
PIP4K2A	rs10741006	22856019	0.7	3.78 × 10−11	1.37 (1.28-1.47)	0.59	2.94 × 10−7	1.296 (1.20-1.40)	2.21 × 10−16	1.33 (1.27-1.40)
PIP4K2A	rs7912551	22818337	0.716	5.0 × 10−9	1.32 (1.23-1.41)	0.632	7.15 × 10−8	1.327 (1.22-1.43)	4.38 × 10−15	1.32 (1.25-1.39)
PIP4K2A	rs4748812 c	22839083	0.684	5.91 × 10−10	1.32 (1.24-1.41)	0.593	1.85 × 10−7	1.30 (1.20-1.40)	1.3 × 10−15	1.31 (1.25-1.38)

^aRisk allele frequency (RAF) in Latino and non-Latino white control subjects

^bP-values and odds ratios (ORs) adjusted for the first ten principal components. In addition, P-values and ORs for BMI1 peak SNPs adjusted for lead PIP4K2A SNP rs10741006, and P-values and ORs for PIP4K2A peak SNPs adjusted for lead BMI1 SNP rs12769953

^cCandidate causal variant

Figure 1. Chromosome 10p12 loci independently associated with childhood ALL.

A: unadjusted meta-analysis of Latino and non-Latino white GWAS revealing two association peaks, upstream of BMI1 and at PIP4K2A; B: meta-analysis adjusted for the lead BMI1 peak SNP rs12769953; C: meta-analysis adjusted for the lead PIP4K2A peak SNP rs10741006. SNPs at both loci remain genome-wide significant (P<5×10⁻⁸) after adjustment. Manhattan plots generated using LocusZoom.

Epistasis analyses between the lead SNPs in the two association peaks revealed no evidence of synergistic or antagonistic interaction between SNPs (P>0.5 for all interactions tested in both Latinos and non-Latino whites).

Given the independence of association and lack of epistasis, we hypothesized that associated variants in the two peaks impart ALL risk via discrete mechanisms and possibly by affecting different gene targets. To explore this hypothesis, we capitalized on our multi-ethnic study design and that the lead SNP at both association peaks differed by ethnicity to investigate putative causal variants underlying the association at each peak. We predicted that underlying causal variant(s) would be in strong LD with lead tag SNPs in both Latino and non-Latino white populations.

rs11591377 is a putative BMI1 enhancer variant

At the BMI1 peak, there were 6 successfully imputed SNPs in tight LD (r² >0.90) with both the lead Latino SNP (rs12769953) in AMR reference populations and the lead non-Latino white SNP (rs4397732) in Europeans (Figure 2a). All 6 SNPs were strongly associated with ALL (P_meta < 5 × 10⁻¹⁰) after adjusting for the lead PIP4K2A SNP. Three SNPs – rs11591377, rs12773841, and rs11599410 – were in perfect LD (r² =1) in AMR and in perfect or near-perfect LD (r² >0.99) in Europeans (Table S3). None of the BMI1 peak SNPs were expression quantitative trait loci (eQTLs) for expression of BMI1 or any other genes in the Genotype-Tissue Expression (GTEx) project ²⁸ or in the Genetic European Variation in Health and Disease (GEUVADIS) project RNA sequencing data (in which no significant eQTLs were detected for BMI1).

Figure 2. Multi-ethnic LD correlation analysis pinpoints BMI1 enhancer element SNP rs11591377.

A: Scatter plot showing correlation between the LD (r²) of SNPs with the lead Latino BMI1 peak SNP (rs12769953) in Admixed American (AMR) populations with the LD (r²) of the same SNPs with the lead non-Latino white SNP (rs4397732) in European (EUR) populations. Plot includes data for 164 biallelic SNPs with P<5.0 × 10⁻⁶ in adjusted meta-analysis. Black triangles denote the lead Latino and lead non-Latino white SNPs. SNP rs11591377 is one of only 6 SNPs (grey diamonds) with a high LD (r²>0.9) with both the lead Latino and lead non-Latino white SNP in respective populations. LD values calculated using 1000 Genomes Project Phase 3 data in the LDlink tool “LDproxy”.

B: Epigenome Browser screenshot showing positions, indicated by vertical blue lines, of the lead Latino and lead non-Latino white SNPs along with the 6 additional SNPs in strong LD with both. SNP rs11591377 (in strong LD with both) overlaps a predicted enhancer element for BMI1, as supported by strong peaks for DNase I hypersensitivity (DHS), H3K4me1, and H3K27ac in hematopoietic stem cells (Mobilized CD34 cells). H3K27ac HiChIP data show strong support (score = 203.96) for chromosomal looping between rs11591377 locus and BMI1 gene in the hematopoietic stem cell-like K562 cell line, which was not found in GM12878 LCLs.

We next explored the potential regulatory function of candidate SNPs. Both rs11591377 and rs12773841 overlapped a ~2kb region containing multiple transcription factor (TF) binding sites in ENCODE ²⁹ (Figure S2). A DNase I hypersensitivity (DHS) cluster was identified at this locus in 22 cell types in ENCODE, of which CD34⁺ Mobilized cells (i.e. hematopoietic progenitor cells) had by far the highest signal. In EnhancerAtlas ²⁰, this DHS cluster lies within a designated enhancer for BMI1 in 7 different tissues/cell lines, including: CD34⁺ cells, GM12878 LCL, fetal thymus, and the leukemia cell lines K562 and HL-60. SNP rs11591377 is positioned at the maximum height of the enhancer consensus score, whereas rs12773841 lies ~200bp upstream of the 5’ end of the enhancer locus (Figure S3). This locus is the only predicted enhancer for BMI1 located in the BMI1 peak region.

We then used the Roadmap Epigenome browser ¹⁹ to assess signals for DHS, which represents open chromatin regions, and for the histone modifications H3K4me1 and H3K27ac, which are known markers of enhancer elements. In support of the rs11591377 locus as an enhancer, we found strong peaks for DHS, H3K4me1, and H3K27ac in hematopoietic stem cells (HSCs) (Figure 2b) and in K562 cells, with smaller peaks in GM12878 LCLs. Analysis of H3K27ac HiChIP chromatin interaction data supported this locus as an enhancer for BMI1 in K562 cells, with a very strong signal for looping between the rs11591377 SNP locus (chr10:22,420,000-22,425,000) and BMI1 (chr10:22,605,000-22,610,000), with score = 203.96 (Figure 2b). This interaction was detected in GM12878 LCLs with a much lower score (score = 5.84), and no looping was detected in T-cells.

Methylation of cytosines at CpGs is frequently associated with transcriptional silencing at promoters and enhancer regions. Thus, we assessed methylation levels at CpG loci across the BMI1 enhancer region, using whole genome bisulfite sequencing data in different tissue types in the Epigenome Browser. This revealed variation across tissues, with HSCs almost entirely unmethylated across 9 CpGs in a region spanning 758bp (mean methylation = 0.56%) (Figure S4), whereas average methylation levels at other tissues ranged from 6.7% (ovary) to 65.7% (brain hippocampus) (Figure S5). Assessment of other BMI1 enhancers revealed these to be largely unmethylated across different tissues (data not shown), suggesting that the BMI1 enhancer spanning SNP rs11591377 is unique in its specificity to HSCs.

rs11591377 associated with differential transcription factor binding

Several transcription factors (TFs), including known hematopoietic regulators, were predicted to bind to this enhancer locus and overlap rs11591377 (Table S4). Analysis of ENCODE ChIP-seq data revealed that the K562 cell line was heterozygous for SNP rs11591377, and revealed significant preferential binding of p300 (P=1.55×10⁻³) and JunD (P=2.61×10⁻³) to the ALL risk allele (G), with trends in the same direction for c-Myc (P=0.077), JunB (P=0.194), MAFK (P=0.069), SPI1 (P=0.134) and TAL1 (P=0.14) (Figures S6 and S7, Table S5). Additional TFs did not have available data in K562 cells or showed minimal binding at the rs11591377 enhancer. We assessed whether rs11591377 altered any TF binding motifs using TFBIND software ²⁴, and found the highest motif score (0.88) for the MYB transcription factor when including the risk allele G, and no predicted MYB binding for the non-risk allele A (Figure 3). MYB-like 2 (MYBL2), a MYB paralog with a similar binding motif sequence, showed a high level of ChIP-seq coverage at the enhancer locus and with highly significant preferential binding of the risk allele in K562 cells (P=1.73×10⁻⁵) (Figure 3). Genotyping of the rs11591377 SNP in genomic DNA from K562 cells using droplet digital PCR (ddPCR) confirmed a 1:1 ratio of allele G:A, supporting that the bias in the ChIP-seq data was not due to allelic copy number differences.

Figure 3.

rs11591377 disrupts MYBL2 binding at BMI1 enhancer

A: MYBL2 binding motif showing high conservation of the rs11591377 risk allele G (red box), with no preference for the non-risk allele A (in red below). Binding motif downloaded from the JASPAR database (http://jaspar.genereg.net/).

B: MYBL2 ChIP-Seq data in K562 cells heterozygous for rs11591377 demonstrate significant preferential binding of the risk allele G (P = 1.73 × 10⁻⁵). The mean proportion of allele G out of total sequenced reads was 0.766. Bars show mean read depth of two biological replicates and error bars represent the standard error of the mean.

C: p300 ChIP-Seq data in K562 cells heterozygous for rs11591377 demonstrate significant preferential binding of the risk allele G (P = 1.6 × 10⁻³). The mean proportion of allele G out of total sequenced reads was 0.883. Bars show mean read depth of two biological replicates and error bars represent the standard error of the mean.

rs4748812 generates RUNX1 binding site in PIP4K2A enhancer

At the PIP4K2A peak, there were no variants with r²>0.90 but there were 4 SNPs with r²>0.75 with both the lead Latino SNP rs10741006 and the lead non-Latino white SNP rs7912551 in corresponding populations (Figure S8). All 4 SNPs – rs4747443, rs4748812, rs12146350, and rs746203 – were associated with ALL at P_meta<5×10⁻¹⁴ after adjusting for the lead BMI1 peak SNP (rs12769953), and were significant eQTLs for PIP4K2A expression in GTEx with ALL risk alleles associated with increased expression (Table S6).

We next sought to determine whether any of these SNPs overlapped putative regulatory regions in PIP4K2A, based on DHS, H3K4me1, and H3K27ac peaks in relevant tissues in the Epigenome Browser. Within a ~38Kb region encompassing the lead Latino SNP, the lead non-Latino white SNP, and the 4 SNPs listed above, only rs4748812 overlapped a region with strong evidence of regulatory function, with a large DHS peak found at chr10: 22,839,000-22,840,000 (Figure S9). SNP rs4748812 was also the only variant found to overlap a DHS cluster locus in the UCSC Genome Browser (Table S6).

H3K27ac HiChIP data analysis in the Epigenome Browser revealed evidence of chromatin looping between the locus containing rs4748812 (chr10:22,835,000-22,840,000) and another PIP4K2A regulatory region ~100kb downstream (chr10:22,935,000-22,940,000) in GM12878 cells, with score = 11.2 (Figure S9). HiChIP analysis in Juicebox also supported interaction between these loci in GM12878 cells (observed/expected score = 1.49), and even stronger interaction between the rs4748812 locus and the PIP4K2A promoter region at chr10:23,000,000-23,005,000 (O/E score = 2.22) (Figure S10).

Inspection of ENCODE ChIP-seq data revealed multiple TFs binding at the rs4748812 locus in LCLs and in K562 cells, including TAL1, CTCF, p300, and ARID3A. Analysis of MethylC-seq data across tissues revealed that demethylation at the rs4748812 locus is largely specific to HSCs (Figures S11 and S12). Investigation of potential TF binding motifs revealed that the rs4748812 risk allele T creates a RUNX1 (AML1) binding site (TFbind score = 0.911), which is not predicted with the protective allele C (Figure S13).

Cytogenetic subtype analysis of BMI1 and PIP4K2A SNPs

Association analyses of lead chromosome 10p12 SNPs in the CCLS revealed that BMI1 SNP rs11591377 was more strongly associated with high hyperdiploid (HD) ALL (OR = 1.56; 95% CI: 1.07-2.27) than with overall ALL risk (OR = 1.22; 95% CI: 1.01-1.49), with no association in non-HD ALL (OR = 1.06; 95% CI: 0.87-1.29) (Table S7, Figure S14). The HD-ALL association with rs11591377 appeared to be stronger in Latinos (OR = 1.71; 95% CI: 1.20-2.50), although the inter-ethnic study difference was not significant (P-value for heterogeneity, P_het = 0.58, Figure S14). The PIP4K2A putative causal SNP rs4748812 also appeared to be more strongly associated with HD-ALL than with overall ALL risk in Latinos. This pattern was not seen in non-Latino whites, although the difference between ethnic groups was again not significant (P_het = 0.32) (Table S7, Figure S14).

Discussion

Our group and others have previously carried out fine-mapping across childhood ALL association loci, identifying both causal coding variants (CDKN2A) and causal regulatory variants (CEBPE and ARID5B) impacting ALL risk ^30–32. Here, fine-mapping across chromosome 10p12 confirmed that two nearby association peaks are independently associated with childhood ALL predisposition, and likely mediate ALL risk through the effects of distinct genes, namely BMI1 and PIP4K2A. In the first reported fine-mapping of the chromosome 10p12.31 association signal, we pinpointed a putatively causal variant rs11591377 located in a predicted enhancer element for BMI1, and that demonstrated preferential transcription factor binding of the risk allele in ENCODE ChIP-Seq data. At chromosome 10p12.2, we identified SNP rs4748812 that lies within a regulatory element in PIP4K2A and is predicted to alter binding of the RUNX1 transcription factor.

The proto-oncogene BMI1 is a member of the polycomb repressive complex 1, and is a negative regulator of the cell-cycle checkpoint proteins p16 and p14ARF ³³, both of which are encoded by CDKN2A, the most frequently deleted gene in ALL tumors in both Latinos and non-Latino whites ³⁴. Hyperactivation of BMI1 promotes oncogenesis through increased cell proliferation and cell lifespan ³³, and BMI1 overexpression is commonly detected in childhood ALL ³⁵, in addition to several other hematological malignancies including AML ³⁶ and B-cell non-Hodgkins lymphoma ³⁷. Overexpression of BMI1 is associated with poorer prognosis in childhood ALL ³⁵ and in AML ³⁶. Although patient outcome data were not available in our study, it would be intriguing to investigate whether rs11591377 is associated with relapse rates in ALL.

The function of BMI1 in leukemogenesis is supported by its essential role in the maintenance of leukemic stem cells ⁴¹, and the reduction of apoptosis in CD34+ HSCs that overexpress BMI1 ⁴². Our analyses support that BMI1 is the gene target of an enhancer element ~180Kb upstream containing SNP rs11591377, with data from recent H3K27ac HiChIP experiments ²¹, as well as MethylC-seq patterns of CpG methylation across tissues, suggesting this enhancer is specific to HSCs and early B-cell progenitors. Enhancers are often tissue or cell-type specific (reviewed in Heinz et al. ⁴³), which may explain the lack of association with BMI1 expression levels in LCLs. Indeed, BMI1 was previously found to be expressed almost exclusively in CD34⁺ progenitor cells and not in differentiated cells ⁴⁴.

The rs11591377 protective allele disrupts a MYB binding motif, and the risk allele demonstrated significant preferential binding to MYB-like 2 (MYBL2) transcription factor, in addition to several other TFs involved in hematopoiesis including p300 and c-Myc. MYBL2 is a regulator of cell cycle progression and cell survival, and is required for proliferation of hematopoietic cells ⁴⁵ and maintenance of HSCs ⁴⁶. Binding of the histone acetyltransferase p300 is a strong predictor of enhancer function ⁴⁹. Moreover, c-Myc is a known positive regulator of BMI1 expression ⁵⁰. Therefore, we posit that the rs11591377 risk allele upregulates BMI1 through binding of MYBL2 and recruitment of p300 and c-Myc, which enables leukemia cells to persist via increased cell proliferation and reduction of apoptosis.

The BMI1 locus may also have general effects on hematopoiesis. A recent GWAS of blood cell traits identified a genome-wide significant SNP rs3011641 in near-perfect LD with rs11591377 (r²=0.94 in Europeans and r²=0.99 in Latinos), in which the major allele was associated with a higher percentage of myeloid cells that are granulocytes ⁵¹, and which in our study was associated with ALL risk at P_meta = 2.4 × 10⁻¹⁰. Mirroring this finding, the novel 8q24 locus identified in our recent GWAS of ALL displays chromosomal contact with a region overlapping several GWAS SNPs associated with blood cell traits, including granulocyte percentage of myeloid cells ^{9, 51}.

We also determined the likely functional variant underlying the ALL association peak at PIP4K2A. The sequence including the rs4748812 risk allele matches a RUNX1 binding motif, TGAGGT, which is the second most frequent RUNX1 binding site in B-cells ⁵², thus supporting a broader role for RUNX1 beyond ETV6-RUNX1 fusion ALL. RUNX1, a crucial transcription factor in hematopoiesis, is located on chromosome 21, which is always triploid or tetraploid in HD-ALL. The effects of rs4748812 on PIP4K2A expression may, therefore, be more pronounced in hyperdiploid leukemia. PIP4K2A is a member of the phosphatidylinositol-4-phosphate 5-kinase family, and is involved in the synthesis of PIP2 (Phosphatidylinositol 4,5-bisphosphate). Activation of PIP2 is known to suppress apoptosis ⁵³, and overexpression of a PIP4K2A paralog, PIP5K1A, has been shown to reduce stress-induced apoptosis ⁵⁴. The overexpression of PIP4K2A associated with ALL risk alleles may thus lead to suppression of apoptosis via increased PIP2 synthesis.

Both BMI1 and PIP4K2A play a role in regulation of apoptosis and cell survival. High hyperdiploid leukemia cells, characterized by gross chromosomal aneuploidies, are inherently unstable and hence prone to apoptosis ⁵⁵. Chromosome aneuploidies cause oxidative stress ⁵⁶, and BMI1 expression may protect HSCs from oxidative stress-induced apoptosis ⁵⁷. Survival of a high hyperdiploid leukemia cell likely relies more heavily on suppression of apoptosis than leukemia subtypes with more stable karyotypes, such as ETV6-RUNX1 fusion. Indeed, this may explain the stronger association of BMI1 and PIP4K2A variants with HD-ALL, demonstrated previously ⁸ and as suggested in our results, in particular for Latinos.

Previous GWAS identified genome-wide significant SNPs across chromosome 10p12.31-p12.2 ^6–8, however the functional variants were not determined. In this study, we have pinpointed putative causal SNPs at both association peaks upstream of BMI1 and at PIP4K2A, providing strong evidence of the function of both loci, in particular for BMI1 peak SNP rs11591377. Our results were based on a refined analysis of previously generated datasets, and additional experiments will be required to elucidate the precise leukemogenic effects of both loci. In this study, we capitalized on the finding of different lead SNPs in Latinos versus non-Latino whites, and on ethnic differences in haplotype structure, to reduce the list of associated SNPs to a handful of strong candidates. This LD correlation analysis should prove a useful strategy for causal variant discovery in other multi-ethnic GWAS datasets.

Novelty and Impact.

We report the first fine-mapping analysis of a chromosome 10p12.31 region associated with childhood acute lymphoblastic leukemia (ALL). In a large California-based genome-wide association study of childhood ALL, we confirm independent genome-wide significant associations at nearby peaks upstream of BMI1 and at PIP4K2A. Using a multi-ethnic linkage disequilibrium correlation approach, we pinpoint putative causal variants at both loci, including SNP rs11591377 that lies within an enhancer for BMI1 and showed differential transcription factor binding.

Abbreviations

ALL

acute lymphoblastic leukemia

AMR

Admixed American

GWAS

genome-wide association study

SNP

single nucleotide polymorphism

LD

linkage disequilibrium

CCRLP

Childhood Cancer Record Linkage Project

GERA

Genetic Epidemiology Research on Aging

CCLS

California Childhood Leukemia Study

DBS

dried bloodspots

HSC

hematopoietic stem cell

ChIP-Seq

chromatin immunoprecipitation sequencing

GTEx

Genotype-Tissue Expression

LCL

lymphoblastoid cell line

HD

high hyperdiploid