Identification of an emphysema-associated genetic variant near TGFB2 with regulatory effects in lung fibroblasts

  1. Margaret M Parker
  2. Yuan Hao
  3. Feng Guo
  4. Betty Pham
  5. Robert Chase
  6. John Platig
  7. Michael H Cho
  8. Craig P Hersh
  9. Victor J Thannickal
  10. James Crapo
  11. George Washko
  12. Scott H Randell
  13. Edwin K Silverman
  14. Raúl San José Estépar
  15. Xiaobo Zhou  Is a corresponding author
  16. Peter J Castaldi  Is a corresponding author
  1. Brigham and Women’s Hospital, United States
  2. School of Medicine, University of Alabama at Birmingham, United States
  3. National Jewish Health, United States
  4. The University of North Carolina at Chapel Hill, United States
8 figures, 2 tables and 2 additional files

Figures

Percentage of each LHE-based emphysema pattern by Global Obstructive Lung Disease (GOLD) stage in ECLIPSE.

NE = Non-emphysematous lung. CL1 = Mild centrilobular. CL2 = Moderate centrilobular. CL3 = Severe centrilobular. PL = Panlobular. PS = Paraseptal. %LAA-950 = emphysema based on −950 Hounsfield unit threshold.

https://doi.org/10.7554/eLife.42720.003
QQ plots for GWAS meta-analyses of non-transformed LHE phenotypes.

(A) Nonemphysematous lung. (B) Mild centrilobular pattern. (C) Moderate centrilobular. (D) Severe centrilobular. (E) Panlobular.

https://doi.org/10.7554/eLife.42720.004
Overview of integrative analyses to prioritize genome-wide significant emphysema-associated loci for functional studies.
https://doi.org/10.7554/eLife.42720.006
Cell type and tissue enrichment for moderate centribloular emphysema GWAS signals.

Using Garfield (Iotchkova et al., 2019) for enrichment analysis, we tested the enrichment of moderate centrilobular GWAS loci (harboring associations a p<5×10−5) in DNaseI peaks from 424 cell lines and cell types in ENCODE and Roadmap. Significant enrichments were observed in fetal lung, fetal stomach, and multiple fibroblast cell types. These significant enrichments are denoted by colored dots located just inside the boundary of the circle of cell types.

https://doi.org/10.7554/eLife.42720.008
The locus zoom plot of GWAS p-values suggests two independent associations (Panel A), and the GWAS signal colocalizes with an eQTL signal in fibroblasts from GTEx (Panel B).

rs1690789 is located at one of these GWAS association peaks and lies within a context-specific DNaseI peak (Panel C). GM12878, H1hESC, and K562 cell lines are shown for reference, and the remaining cell types are those with DNaseI peaks that overlap rs1690789. Raw DNaseI data from only one experimental replicate are shown. GM12878 = lymphblastoid cell line. H1hESC = human embryonic stem cell. K562 = leukemia cell line. AG0449-AG10803 refer to fibroblasts from different subjects and sampling sites. HGF = gingival fibroblasts. HMVECLLy = lung derived microvascular endothelial cells. HPdLF - Periodontal ligament fibroblasts. HRPEpiC – retinal pigment epithelial cells. HSMM – skeletal muscle myoblasts. NHDF – dermal fibroblasts. NHLF – lung fibroblasts.

https://doi.org/10.7554/eLife.42720.009
Secondary association for moderate centrilobular emphysema at the TGFB2 locus.

Results from genetic association meta-analysis conditioned on the lead SNP (rs796395) at this region.

https://doi.org/10.7554/eLife.42720.010
Publicly available chromatin conformation capture (4C) results in IMR90 cells show multiple peaks of interaction for the 10 kb region containing the context-specific regulatory element around rs1690789 (Panel A – blue spikes in top figure indicate regions of high interaction frequency and light blue curved lines in lower figure indicate chromosomal interactions with the 10 kb region containing rs1690789).

Newly generated 3C assays in IMR90 fibroblasts verify the interaction between the region containing rs1690789 and TGFB2 promoter (Panel B). Primer sequences are listed in Supplementary file 1 Table 10.

https://doi.org/10.7554/eLife.42720.011
Regional knockout of rs1690789 in primary human lung fibroblasts using CRISPR/Cas9 editing.

Three pairs of sgRNAs were applied to delete a DNA region of ~100 bp spanning rs1690789 (A). The expression of TGFB2 is downregulated in rs1690789 knockout lung fibroblast cells with qPCR quantification, n = 4 (B). The editing efficiency is examined to confirm the effect of CRISPR/Cas9 regional knockout, n = 4 (C). *p<0.05 compared to control by unpaired t test.

https://doi.org/10.7554/eLife.42720.012

Tables

Table 1
Genome-wide significant meta-analysis results for local histogram emphysema phenotypes.
https://doi.org/10.7554/eLife.42720.005
LHE patternLead SNPChromosomePosition (hg19)EAFEffect alleleP value metaEffect metaSE effect meta
Moderate Centrilobularrs5607733315788990030.67a2.2E-130.0162.2E-03
rs17368582111027380750.14t8.1E-120.0243.5E-03
rs5611385019413531070.59t1.5E-09−0.0162.6E-03
rs79639512186819710.52a6.1E-090.0132.2E-03
rs13864140241454457790.38a3.3E-080.0142.5E-03
Nonemphysematous (Normal Lung)rs13864140241454457790.38a4.2E-08−0.0213.9E-03
rs717006815789129430.78a4.8E-120.0284.0E-03
rs17368659111027427610.86t6.3E-110.0365.4E-03
rs28929474*14948449470.98t6.2E-09−0.0711.2E-02
Panlobularrs1185237215788013940.35a7.1E-12−0.0035.0E-04
rs76756075*111123498440.02t2.0E-08−0.0071.2E-03
rs78070126*1165746080.97t2.8E-080.0061.1E-03
rs14577077021524878080.99a3.8E-08−0.0152.8E-03
Severe Centrilobularrs978872115788028690.38t3.4E-15−0.0056.0E-04
rs37912317308918140.40t3.7E-08−0.0047.0E-04
  1. LHE - local histogram emphysema.

    EAF - effect allele frequency in 1000 Genomes CEU population.

  2. *indicates novel association not previously associated in GWAS of COPD or emphysema (rs28929474 was associated to FEV1/FVC in smokers in Li et al., 2018, during preparation of this manuscript).

Table 2
Genomewide significant LHE loci that colocalize with eQTL.
https://doi.org/10.7554/eLife.42720.007
GWASSNPCHRPOSTissueGENE
Moderate centrilobularrs7963951218681971FibroblastsTGFB2
rs560773331578899003Fibroblasts, TestisPSMA4, CHRNA5
Moderate centrilobular, normalrs561138501941353107Lung, TestisCYP2A6, AKT2
Panlobularrs118523721578801394TestisCHRNA5
All loci have colocalization probability > 80%, reflecting the estimated probability that the GWAS and eQTL association signals arise from a shared causal variant.
LHE pattern – local histogram emphysema phenotype used for GWAS.
SNP - lead GWAS variant in locus.
Tissue - tissue of origin for gene expression data in eQTL analysis.
Gene – eQTL targeted gene in specified GTEx tissue.

Additional files

Download links

A two-part list of links to download the article, or parts of the article, in various formats.

Downloads (link to download the article as PDF)

Open citations (links to open the citations from this article in various online reference manager services)

Cite this article (links to download the citations from this article in formats compatible with various reference manager tools)

  1. Margaret M Parker
  2. Yuan Hao
  3. Feng Guo
  4. Betty Pham
  5. Robert Chase
  6. John Platig
  7. Michael H Cho
  8. Craig P Hersh
  9. Victor J Thannickal
  10. James Crapo
  11. George Washko
  12. Scott H Randell
  13. Edwin K Silverman
  14. Raúl San José Estépar
  15. Xiaobo Zhou
  16. Peter J Castaldi
(2019)
Identification of an emphysema-associated genetic variant near TGFB2 with regulatory effects in lung fibroblasts
eLife 8:e42720.
https://doi.org/10.7554/eLife.42720