Back

Smoking drives an epigenetic memory of aberrant hematopoiesis

Breeze, C. E.; Goodney, G.; Wang, H.; Hubbard, A. K.; Lim, J.; Machiela, M. J.; Hoang, T. T.; Richards-Barber, M.; Tran, C.; Tolentino, M.; Hansen, M.; Porecha, R.; Renke, N.; Zhou, W.; Franceschini, N.; Berndt, S. I.; Hofmann, J.; Lee, M.; London, S. J.; Wong, J. Y.

2026-05-21 epidemiology
10.64898/2026.05.14.26353250 medRxiv
Show abstract

Tobacco smoking induces DNA methylation (DNAm) changes in blood and other tissues, which may influence chronic health outcomes. However, the breadth of smoking-related DNAm changes remains unmapped, offering a space for employing novel technologies. To expand our understanding of smoking impacts on DNAm, we conducted an epigenome-wide association study (EWAS) comparing ever smokers to never smokers, using blood from a multiethnic U.S. study population (n=887). We employed the newly developed Illumina Methylation Screening Array (MSA) covering 269,094 unique sites, including 123,776 CpGs not assayed in previous EWAS. Trans-ethnic meta-analysis identified 152 differentially methylated positions (DMPs) associated with ever-smoking status (n=764); European-specific analysis yielded 129 DMPs (n=674), including 106 overlapping with trans-ethnic analysis. A separate, large-scale replication EWAS (n=2,190) confirmed 91 trans-ethnic and 77 European-specific DMPs. Among our findings, we identified 61 DMPs at CpGs novel to the MSA platform, including near both new and known smoking-associated genes. Most notably, we uncovered a dense cluster of 12 DMPs within a 1117 bp region of ECEL1P1, forming the most long-lasting, persistent smoking-associated DMR ever detected, even among former smokers who quit decades prior. We also detected new signals at AHRR, a well-known locus for smoking-related DNAm changes. eFORGE analysis revealed that detected smoking-associated DNAm changes are predominantly located in hematopoietic stem and progenitor cell (HSPC) DNase I hotspots, aligning with gene set enrichment analyses that highlighted pathways related to hematopoietic stem cell differentiation. Our findings suggest that HSPCs serve as a reservoir for an epigenetic memory of smoking. Additionally, we observed short-term cell-specific smoking-associated DNAm changes in myeloid cells. Our results demonstrate the utility of the MSA in expanding our knowledge of both transient and persistent environmental exposure-associated DNAm changes.

Matching journals

The top 4 journals account for 50% of the predicted probability mass.

1
Nature Communications
5641 papers in training set
Top 7%
19.0%
2
Epigenetics
50 papers in training set
Top 0.1%
15.5%
3
Clinical Epigenetics
60 papers in training set
Top 0.1%
9.2%
4
Science Advances
1243 papers in training set
Top 2%
6.9%
50% of probability mass above
5
eLife
5828 papers in training set
Top 20%
5.6%
6
Cell Reports
1498 papers in training set
Top 8%
5.0%
7
Scientific Reports
3612 papers in training set
Top 41%
2.5%
8
Epigenetics & Chromatin
42 papers in training set
Top 0.2%
2.5%
9
JNCI: Journal of the National Cancer Institute
19 papers in training set
Top 0.1%
2.5%
10
Proceedings of the National Academy of Sciences
2444 papers in training set
Top 23%
2.2%
11
Communications Biology
993 papers in training set
Top 16%
1.5%
12
PLOS Genetics
862 papers in training set
Top 8%
1.5%
13
Neuropsychopharmacology
153 papers in training set
Top 2%
1.4%
14
eBioMedicine
183 papers in training set
Top 4%
1.2%
15
Epigenomics
11 papers in training set
Top 0.1%
1.2%
16
Genome Biology
637 papers in training set
Top 7%
1.1%
17
Science Translational Medicine
127 papers in training set
Top 3%
1.1%
18
BMC Medicine
176 papers in training set
Top 4%
1.1%
19
Circulation Research
47 papers in training set
Top 1%
1.1%
20
CHEST
14 papers in training set
Top 0.3%
0.9%
21
Nature Genetics
286 papers in training set
Top 5%
0.9%
22
Cell Metabolism
57 papers in training set
Top 1%
0.6%
23
Genome Medicine
183 papers in training set
Top 6%
0.6%
24
Molecular Psychiatry
282 papers in training set
Top 6%
0.6%
25
Nature Medicine
125 papers in training set
Top 4%
0.6%
26
Nature
645 papers in training set
Top 11%
0.6%