Health
Johns Hopkins researchers find cigarette smoke alters lung stem cells, determining whether KRAS or TP53 mutations lead to adenocarcinoma or squamous cell carcinoma.

Lung stem cells undergo distinct reprogramming when exposed to chronic cigarette smoke, a process that ultimately dictates the specific subtype of non-small cell lung cancer (NSCLC) that develops. Researchers at the Johns Hopkins Sidney Kimmel Comprehensive Cancer Center discovered that this environmental exposure creates unique precancerous states in different stem cell populations, making them selectively vulnerable to particular genetic alterations.
The team utilized laboratory-grown lung organoids—tiny 3D models containing multiple cell types derived from normal mouse and human lung tissue—to track cellular changes over six months of exposure to cigarette smoke condensate. This condensate consists of the chemicals and particles found in smoke. The study revealed that prolonged exposure modified both epigenetic programming, which regulates gene activity without changing DNA sequences, and the genes being expressed within various lung stem cell groups.
These modifications suppressed inflammatory and immune signaling pathways, including those involving programmed cell death. A key gene affected was ZBP1, a regulator of inflammatory cell death. Over time, the exposed cells acquired characteristics associated with a tumor-permissive state, marked by progressive changes in DNA methylation and chromatin accessibility.
To test susceptibility, investigators introduced two common smoking-related genetic alterations: mutant KRAS or loss of the tumor suppressor gene TP53. They implanted these modified cells into mice. Tumors formed only in organoids that had undergone six months of smoke exposure and contained one of the genetic alterations. Neither smoke exposure alone nor genetic mutations in unexposed control organoids produced tumors.
The results showed that the altered stem cell populations responded differently to each mutation. KRAS mutations primarily drove lung adenocarcinomas from cells derived from bronchioalveolar stem cells, located where airways meet alveoli. Conversely, TP53 loss produced squamous cell carcinomas derived from basal stem cells, which maintain and repair airway linings.
"What we have tried to do is model lung cancer from its very earliest stages," said Michelle Vaz, Ph.D., instructor in oncology and senior author of the study. "Using these organoid models, which are rich in stem cells, we were able to follow the cells over six months of exposure and see the epigenetic and gene expression changes occurring in these different stem cell types."
Vaz emphasized that genetic events alone are insufficient to start cancer; cells must first undergo changes from chronic smoke exposure to become susceptible to transformation. Single-cell RNA sequencing confirmed that KRAS-driven adenocarcinomas arose from an altered bronchioalveolar stem cell state, while TP53-deficient squamous cell carcinomas traced back to an altered basal stem cell state.
"It was very interesting that when we introduced KRAS or loss of TP53, each seemed to select for a particular type of stem cell state," Vaz noted. "We did not engineer the genetic event into a particular cell type. This is what the mutations selected for."
Silencing ZBP1 and related interferon signaling was particularly pronounced in KRAS-mutant tumors, suggesting that reducing these normal cell-death defenses may help KRAS-driven cancer cells survive. Analysis of human lung adenocarcinoma data supported this, showing significantly lower expression of ZBP1 and related interferon pathway genes in tumors with KRAS mutations compared to those with TP53 mutations.
The findings suggest that the type of lung cancer depends on both the specific genetic mutation and the molecular state of the cell in which it occurs. Chronic smoke exposure reshapes these cellular states through inflammatory, epigenetic, and gene-expression changes, creating environments where particular cancer-driving mutations thrive.
Researchers propose that these altered stem cell states could eventually serve as molecular markers for recognizing the earliest evolution of different NSCLC subtypes. Such markers might aid in identifying individuals at increased risk and preventing cancer development, though additional experimental validation is required.
The study also raises questions about therapeutically targeting pathways altered by cigarette smoke. "What's exciting as a next step is whether we can learn more about these cell death pathways and whether some can be targeted, specifically in the KRAS versus TP53 setting," Vaz said. She suggested combining such approaches with epigenetic therapy or immunotherapy to improve responses in tumors that currently do not respond to treatment.
The research involved contributions from Na Wang, Raksha Padaki, Sara-Jayne Thursby, Ray-Whay Chiu Yen, Leslie Cope, Malcolm Brock, Edward Gabrielson, and Hariharan Easwaran. Co-senior author Stephen Baylin, M.D., Virginia and D.K. Ludwig Professor of Cancer Research and co-director of the Cancer Genetics and Epigenetics Program, collaborated closely with Vaz on the project.
Funding for the work came from the Dr. Miriam and Sheldon G. Adelson Medical Research Foundation, The Hodson Trust, an American Lung Association Lung Cancer Discovery Award, the Evelyn Grollman Glick Scholar Award, a Johns Hopkins University Discovery Award, TEDCO, the National Institute of Environmental Health Sciences, National Institutes of Health grant R01ES011858, National Cancer Institute grants R01CA229240 and R01CA230995, the Sidney Kimmel Comprehensive Cancer Center at Johns Hopkins Support Grant P30CA006973, and the Van Andel Research Institute through the Van Andel Research Institute–Stand Up To Cancer Epigenetics Dream Team.
The findings were reported Oct. 5 in the Proceedings of the National Academy of Sciences. The journal reference is Wang, N., et al. (2026). Chronic cigarette smoke exposure induces distinct stem cell states driving genetic driver–specific non–small cell lung cancer subtypes. DOI: 10.1073/pnas.2604393123.



