Adeno-to-squamous transition drives resistance to KRAS inhibition in LKB1 mutant lung cancer

Xinyuan Tong, Ayushi S. Patel, Eejung Kim, Hongjun Li, Yueqing Chen, Shuai Li, Shengwu Liu, Julien Dilly, Kevin S. Kapner, Yun Xue, Laura D. Hover, Suman Mukhopadhyay, Fiona Sherman, Khrystyna Mynzdar, Yijun Gao, Fei Li, Fuming Li, Zhaoyuan Fang, Yujuan Jin, Juntao Gao, Minglei Shi, Luonan Chen, Yang Chen, Thian Kheoh, Wenjing Yang, Itai Yanai, Andre L. Moriera, Vamsidhar Velcheti, Benjamin G. Neel, Liang Hu, James G. Christensen, Peter Olson, Dong Gao, Michael Q. Zhang, Andrew J. Aguirre, Kwok-Kin Wong, Hongbin JibioRxiv 2023 · DOI

TL;DR

This paper reveals that lung cancer cells can switch from an "adeno" to a "squamous" cell type, a process called Adeno-to-Squamous Transition (AST), as a mechanism to resist KRAS inhibitors. This lineage plasticity, particularly in tumors with LKB1 mutations, is driven by the transcription factor ΔNp63 and can be predicted by the expression of KRT6A, offering a potential biomarker for treatment response.

What Problem Does This Paper Address?

KRAS G12C inhibitors like adagrasib and sotorasib have shown promise for lung cancers with KRAS G12C mutations. However, a significant challenge is the development of drug resistance, leading to treatment failure. This is particularly problematic in a subset of lung adenocarcinomas that also harbor mutations in the LKB1 (or STK11) gene, which are known to be associated with poorer outcomes. The authors investigate the mechanisms behind this resistance, specifically exploring whether a change in cell identity, rather than just genetic mutations, could be responsible.

Methods

The study employed a multi-pronged approach:

  • Clinical Data Analysis: Transcriptomic and clinical data from patients treated with adagrasib in the KRYSTAL-1 trial were analyzed to correlate gene expression signatures with treatment duration and survival.
  • Genetically Engineered Mouse Models (GEMMs): Mice with specific KRAS and LKB1/TP53 mutations were used to create lung cancer models. These models were treated with adagrasib to study resistance development.
  • Organoid Culture: Tumors from GEMMs were cultured as organoids to allow for long-term treatment studies and mechanistic investigations in a controlled environment.
  • Transcriptomic and Epigenomic Profiling: RNA-sequencing (RNA-seq) and ATAC-sequencing were used to analyze gene expression and chromatin accessibility changes during AST and resistance development.
  • Single-Cell RNA Sequencing (scRNA-seq): This technique allowed for a detailed examination of cell-fate changes and the identification of intermediate cell states during AST.
  • Immunohistochemistry (IHC): Used to visualize protein expression of key markers (e.g., TTF1, ANp63, KRT6A) in tumor tissues and organoids.

Datasets

  • KRYSTAL-1 Clinical Trial Data: Transcriptomic and clinical data from 116 KRAS G12C-mutant lung cancer patients treated with adagrasib.
  • Mouse Lung Cancer Models: Data generated from GEMMs (KrasLSL-G12D/+/Lkb1flox/flox (KDL), KRASLSL-G12C/+/Trp53flox/flox (KCP), and KRASLSL-G12C/+/Lkb1flox/flox (KcL)) and their derived organoids.
  • Publicly Available Datasets: Used for gene signature annotations (e.g., Darwiche Squamous Signature, Hallmark gene sets).

Key Insights and Ideas

  • Lineage Plasticity as a Resistance Mechanism: The most striking insight is that cancer cells can actively change their identity (from adenocarcinoma to squamous cell carcinoma) to evade targeted therapies, rather than solely relying on genetic mutations.
  • LKB1 Deficiency Fuels Plasticity: Tumors with LKB1 mutations appear to have a pre-existing "plasticity" that makes them more prone to undergoing AST when challenged with KRAS inhibitors.
  • ΔNp63 as a Master Regulator: The transcription factor ΔNp63 emerges as a key driver of AST, promoting the switch to a squamous cell state and conferring resistance.
  • Intermediate High-Plasticity Cell State: scRNA-seq revealed a transient cell state characterized by high plasticity, which acts as a bridge between the initial adenocarcinoma state and the resistant squamous state.
  • KRT6A as a Predictive Biomarker: The expression of KRT6A, a gene associated with squamous differentiation, is identified as a potential biomarker that predicts poor response to adagrasib, especially in LKB1-mutant tumors.

Key Results

  • Patients with a high baseline squamous gene signature in their tumors showed significantly shorter treatment durations with adagrasib.
  • KRAS G12C; LKB1 mutant mouse models treated with adagrasib developed tumors with pronounced squamous features and resistance.
  • scRNA-seq identified an intermediate cell state with high plasticity signatures during AST.
  • Overexpression of ΔNp63 in lung cancer cells induced squamous morphology and conferred resistance to KRAS inhibitors.
  • High KRT6A expression in pre-treatment biopsies correlated with shorter overall survival and progression-free survival in patients receiving adagrasib.

Limitations and Future Work

  • Sample Size: The study acknowledges that the sample size for clinical data, particularly with transcriptomic data, is relatively small, limiting definitive conclusions for certain analyses.
  • Mechanism of KEAP1/NFE2L2: While the study found correlations between KEAP1 mutations and KRT6A expression, the precise mechanism by which KEAP1/NFE2L2 mutations mediate resistance is not fully elucidated.
  • Combination Therapies: The authors suggest that developing novel combination therapy strategies targeting AST is crucial for improving outcomes in patients with KRAS; LKB1-mutant lung cancer. Further research is needed to explore these combinations.
  • Broader Applicability: While the study focuses on KRAS G12C inhibitors, future work could investigate if AST-driven resistance is a common mechanism for other KRAS inhibitors (e.g., sotorasib) and other KRAS mutations.

Summary updated

Abstract

Summary KRAS G12C inhibitors including adagrasib and sortorasib have shown clinical promise in targeting KRAS G12C -mutated lung cancers, however, most patients eventually develop drug resistance. In lung adenocarcinoma patients with co-occurring KRAS G12C and STK11 / LKB1 mutations, we found a high squamous gene signature at baseline significantly correlated with poor adagrasib response. Through integrative studies of Lkb1 -deficient KRAS G12C and Kras G12D lung cancer mouse models and/or organoids treated with KRAS inhibitors, we found tumor cells invoked a lineage plasticity program: adeno-to-squamous transition (AST) that mediated resistance to KRAS inhibition. Transcriptomic and epigenomic analyses revealed ΔNp63 drives AST and modulates response to KRAS inhibition. We identified an intermediate high-plasticity cell state with distinct gene expression program marked by Krt6a upregulation. Notably, higher KRT6A expression at baseline correlated with shorter overall survival in KRAS -mutant patients receiving adagrasib. These data support the role of AST in KRAS inhibitor resistance and provide predictive biomarker for KRAS-targeted therapies in lung cancer.
BibTeX
@misc{tong2023adeno,
  author = {Xinyuan Tong and Ayushi S. Patel and Eejung Kim and Hongjun Li and Yueqing Chen and Shuai Li and Shengwu Liu and Julien Dilly and Kevin S. Kapner and Yun Xue and Laura D. Hover and Suman Mukhopadhyay and Fiona Sherman and Khrystyna Mynzdar and Yijun Gao and Fei Li and Fuming Li and Zhaoyuan Fang and Yujuan Jin and Juntao Gao and Minglei Shi and Luonan Chen and Yang Chen and Thian Kheoh and Wenjing Yang and Itai Yanai and Andre L. Moriera and Vamsidhar Velcheti and Benjamin G. Neel and Liang Hu and James G. Christensen and Peter Olson and Dong Gao and Michael Q. Zhang and Andrew J. Aguirre and Kwok-Kin Wong and Hongbin Ji},
  title = {{Adeno-to-squamous transition drives resistance to KRAS inhibition in <i>LKB1</i> mutant lung cancer}},
  year = {2023},
  doi = {10.1101/2023.09.07.556567},
  howpublished = {bioRxiv},
}
Figure