Mathieu Platteel

Trans-eQTLs Reveal That Independent Genetic Variants Associated with a Complex Phenotype Converge on Intermediate Genes, with a Major Role for the HLA

For many complex traits, genetic variants have been found associated. However, it is still mostly unclear through which downstream mechanism these variants cause these phenotypes. Knowledge of these intermediate steps is crucial to understand pathogenesis, while also providing leads for potential pharmacological intervention. Here we relied upon natural human genetic variation to identify effects of these variants on trans-gene expression (expression quantitative trait locus mapping, eQTL) in whole peripheral blood from 1,469 unrelated individuals. We looked at 1,167 published trait- or disease-associated SNPs and observed trans-eQTL effects on 113 different genes, of which we replicated 46 in monocytes of 1,490 different individuals and 18 in a smaller dataset that comprised subcutaneous adipose, visceral adipose, liver tissue, and muscle tissue. HLA single-nucleotide polymorphisms (SNPs) were 10-fold enriched for trans-eQTLs: 48% of the trans-acting SNPs map within the HLA, including ulcerative colitis susceptibility variants that affect plausible candidate genes AOAH and TRBV18 in trans. We identified 18 pairs of unlinked SNPs associated with the same phenotype and affecting expression of the same trans-gene (21 times more than expected, P<10−16). This was particularly pronounced for mean platelet volume (MPV): Two independent SNPs significantly affect the well-known blood coagulation genes GP9 and F13A1 but also C19orf33, SAMD14, VCL, and GNG11. Several of these SNPs have a substantially higher effect on the downstream trans-genes than on the eventual phenotypes, supporting the concept that the effects of these SNPs on expression seems to be much less multifactorial. Therefore, these trans-eQTLs could well represent some of the intermediate genes that connect genetic variants with their eventual complex phenotypic outcomes.

Common and different genetic background for rheumatoid arthritis and coeliac disease

Recent genome-wide association studies (GWAS) have revealed genetic risk factors in autoimmune and inflammatory disorders. Several of the associated genes and underlying pathways are shared by various autoimmune diseases. Rheumatoid arthritis (RA) and coeliac disease (CD) are two autoimmune disorders which have commonalities in their pathogenesis. We aimed to replicate known RA loci in a Dutch RA population, and to investigate whether the effect of known RA and CD risk factors generalize across the two diseases. We selected all loci associated to either RA or CD in a GWAS and confirmed in an independent cohort, with a combined P-value cut-off P < 5 x 10(-6). We genotyped 11 RA and 11 CD loci in 1368 RA patients, 795 CD patients and 1683 Dutch controls. We combined our results in a meta-analysis with UK GWAS on RA (1860 cases; 2938 controls) and CD (767 cases; 1422 controls). In the Dutch RA cohort, the PTPN22 and IL2/IL21 variants showed convincing association (P = 3.4 x 10(-12) and P = 2.8 x 10(-4), respectively). Association of RA with the known CD risk variant in the SH2B3 was also observed, predominantly in the subgroup of rheumatoid factor-positive RA patients (P = 0.0055). In a meta-analysis of Dutch and UK data sets, shared association with six loci (TNFAIP3, IL2/IL21, SH2B3, LPP, MMEL1/TNFRSF14 and PFKFB3/PRKCQ) was observed in both RA and CD cohorts. We confirmed two known loci and identified four novel ones for shared CD-RA genetic risk. Most of the shared loci further emphasize a role for adaptive and innate immunity in these diseases.

PROS1 analysis in 87 pedigrees with hereditary protein S deficiency demonstrates striking genotype-phenotype associations

Hereditary protein S (PS) deficiency predisposes to venous thrombosis. Previously, we demonstrated a difference in risk of venous thrombosis between PS deficiency type I and type III. We used direct sequencing, multiplex ligation‐dependent probe amplification (MLPA), and linkage analysis to study whether this difference could be explained by molecular heterogeneity. The study contained two sets of families with PS deficiency type I (cohort 1; 35 probands, 155 relatives) or type III (cohort 2; 52 probands, 241 relatives). In cohort 1, a mixed type I/type III PS‐deficient phenotype was observed in 66% of the pedigrees. A total of 34 probands carried a mutant PROS1 allele, compared to one proband in cohort 2 (P<10−10). The probands mutation was identified in all type I, but only in 57% of type III PS deficient relatives. MLPA‐analysis in the mutation negative families did not reveal PROS1 deletions or insertions. Linkage analysis in 16 families showed cosegregation of PROS1 markers in the family with type I deficiency, but not in the 15 families with type III deficiency. The genotype–phenotype associations point to differences in genetic architecture. Whereas PS deficiency type I is a monogenic disease due to PROS1 allelic heterozygosity, PS deficiency type III is most likely a more complex or heterogeneous disorder. Hum Mutat 29(7), 939–947, 2008.

Cell Specific eQTL Analysis without Sorting Cells

The functional consequences of trait associated SNPs are often investigated using expression quantitative trait locus (eQTL) mapping. While trait-associated variants may operate in a cell-type specific manner, eQTL datasets for such cell-types may not always be available. We performed a genome-environment interaction (GxE) meta-analysis on data from 5,683 samples to infer the cell type specificity of whole blood cis-eQTLs. We demonstrate that this method is able to predict neutrophil and lymphocyte specific cis-eQTLs and replicate these predictions in independent cell-type specific datasets. Finally, we show that SNPs associated with Crohn’s disease preferentially affect gene expression within neutrophils, including the archetypal NOD2 locus.

Susceptibility to chronic mucus hypersecretion, a genome wide association study.

Background Chronic mucus hypersecretion (CMH) is associated with an increased frequency of respiratory infections, excess lung function decline, and increased hospitalisation and mortality rates in the general population. It is associated with smoking, but it is unknown why only a minority of smokers develops CMH. A plausible explanation for this phenomenon is a predisposing genetic constitution. Therefore, we performed a genome wide association (GWA) study of CMH in Caucasian populations. Methods GWA analysis was performed in the NELSON-study using the Illumina 610 array, followed by replication and meta-analysis in 11 additional cohorts. In total 2,704 subjects with, and 7,624 subjects without CMH were included, all current or former heavy smokers (≥20 pack-years). Additional studies were performed to test the functional relevance of the most significant single nucleotide polymorphism (SNP). Results A strong association with CMH, consistent across all cohorts, was observed with rs6577641 (p = 4.25×10−6, OR = 1.17), located in intron 9 of the special AT-rich sequence-binding protein 1 locus (SATB1) on chromosome 3. The risk allele (G) was associated with higher mRNA expression of SATB1 (4.3×10−9) in lung tissue. Presence of CMH was associated with increased SATB1 mRNA expression in bronchial biopsies from COPD patients. SATB1 expression was induced during differentiation of primary human bronchial epithelial cells in culture. Conclusions Our findings, that SNP rs6577641 is associated with CMH in multiple cohorts and is a cis-eQTL for SATB1, together with our additional observation that SATB1 expression increases during epithelial differentiation provide suggestive evidence that SATB1 is a gene that affects CMH.

Understanding human immune function using the resources from the Human Functional Genomics Project

Understanding human immune function using the resources from the Human Functional Genomics Project

Gene expression analysis in predicting the effectiveness of insect venom immunotherapy

BACKGROUND Venom immunotherapy (VIT) enables longtime prevention of insect venom allergy in the majority of patients. However, in some, the risk of a resystemic reaction increases after completion of treatment. No reliable factors predicting individual lack of efficacy of VIT are currently available. OBJECTIVE To determine the use of gene expression profiles to predict the long-term effect of VIT. METHODS Whole genome gene expression analysis was performed on RNA samples from 46 patients treated with VIT divided into 3 groups: (1) patients who achieved and maintained long-term protection after VIT, (2) patients in whom insect venom allergy relapsed, and (3) patients still in the maintenance phase of VIT. RESULTS Among the 48.071 transcripts analyzed, 1401 showed a >2 fold difference in gene expression (P < .05); 658 genes (47%) were upregulated and 743 (53%) downregulated. Forty-three transcripts still show significant differences in expression after correction for multiple testing; 12 of 43 genes (28%) were upregulated and 31 of 43 genes (72%) downregulated. A naive Bayes prediction model demonstrated a gene expression pattern characteristic of effective VIT that was present in all patients with successful VIT but absent in all subjects with failure of VIT. The same gene expression profile was present in 88% of patients in the maintenance phase of VIT. CONCLUSION Gene expression profiling might be a useful tool to assess the long-term effectiveness of VIT. The analysis of differently expressed genes confirms the involvement of immunologic pathways described previously but also indicates novel factors that might be relevant for allergen tolerance.

Fine mapping of the celiac disease-associated LPP locus reveals a potential functional variant

Using the Immunochip for genotyping, we identified 39 non-human leukocyte antigen (non-HLA) loci associated to celiac disease (CeD), an immune-mediated disease with a worldwide frequency of ∼1%. The most significant non-HLA signal mapped to the intronic region of 70 kb in the LPP gene. Our aim was to fine map and identify possible functional variants in the LPP locus. We performed a meta-analysis in a cohort of 25 169 individuals from six different populations previously genotyped using Immunochip. Imputation using data from the Genome of the Netherlands and 1000 Genomes projects, followed by meta-analysis, confirmed the strong association signal on the LPP locus (rs2030519, P = 1.79 × 10−49), without any novel associations. The conditional analysis on this top SNP-indicated association to a single common haplotype. By performing haplotype analyses in each population separately, as well as in a combined group of the four populations that reach the significant threshold after correction (P < 0.008), we narrowed down the CeD-associated region from 70 to 2.8 kb (P = 1.35 × 10−44). By intersecting regulatory data from the ENCODE project, we found a functional SNP, rs4686484 (P = 3.12 × 10−49), that maps to several B-cell enhancer elements and a highly conserved region. This SNP was also predicted to change the binding motif of the transcription factors IRF4, IRF11, Nkx2.7 and Nkx2.9, suggesting its role in transcriptional regulation. We later found significantly low levels of LPP mRNA in CeD biopsies compared with controls, thus our results suggest that rs4686484 is the functional variant in this locus, while LPP expression is decreased in CeD.

Serum Lipid Levels, Body Mass Index, and Their Role in Coronary Artery Calcification: A Polygenic Analysis

Background—Coronary artery calcification (CAC) is widely regarded as a cumulative lifetime measure of atherosclerosis, but it remains unclear what is the relationship between calcification and traditional risk factors for coronary artery disease (CAD) and myocardial infarction (MI). This study characterizes the genetic architecture of CAC by evaluating the overall impact of common alleles associated with CAD/MI and its traditional risk factors. Methods and Results—On the basis of summary-association results from the CARDIoGRAMplusC4D study of CAD/MI, we calculated polygenic risk scores in 2599 participants of the Dutch and Belgian Lung Cancer Screening (NELSON) trial, in whom quantitative CAC levels (Agatston scores) were determined from chest computerized tomographic imaging data. The most significant polygenic model explained ≈14% of the observed CAC variance (P=1.6×10–11), which points to a residual effect because of many as yet unknown loci that overlap between CAD/MI and CAC. In addition, we constructed risk scores based on published single-nucleotide polymorphism associations for traditional cardiovascular risk factors and tested these scores for association with CAC. We found nominally significant associations for genetic risk scores of low-density lipoprotein-cholesterol, total cholesterol, and body mass index, which were successfully replicated in 2182 individuals of the Heinz Nixdorf Recall Study. Conclusions—Pervasive polygenic sharing between CAC and CAD/MI suggests that a substantial fraction of the heritable risk for CAD/MI is mediated through arterial calcification. We also provide evidence that genetic variants associated with serum lipid levels and body mass index influence CAC levels.

A framework for the detection of de novo mutations in family-based sequencing data

Germline mutation detection from human DNA sequence data is challenging due to the rarity of such events relative to the intrinsic error rates of sequencing technologies and the uneven coverage across the genome. We developed PhaseByTransmission (PBT) to identify de novo single nucleotide variants and short insertions and deletions (indels) from sequence data collected in parent-offspring trios. We compute the joint probability of the data given the genotype likelihoods in the individual family members, the known familial relationships and a prior probability for the mutation rate. Candidate de novo mutations (DNMs) are reported along with their posterior probability, providing a systematic way to prioritize them for validation. Our tool is integrated in the Genome Analysis Toolkit and can be used together with the ReadBackedPhasing module to infer the parental origin of DNMs based on phase-informative reads. Using simulated data, we show that PBT outperforms existing tools, especially in low coverage data and on the X chromosome. We further show that PBT displays high validation rates on empirical parent-offspring sequencing data for whole-exome data from 104 trios and X-chromosome data from 249 parent-offspring families. Finally, we demonstrate an association between father’s age at conception and the number of DNMs in female offspring’s X chromosome, consistent with previous literature reports.