Unlike the leopard, cancers do 'change their spots' and thereby resist treatment

 

Recent posts from Cancer Ecology Commentary have focused on understanding how the epigenome, the set of DNA and chromatin modifications regulating cell gene expression, informs our understand of the connection between intrinsic influences from oncogenes and the extrinsic effects arising from the environment in explaining carcinogenesis. Using a model adopted from developmental biology to explain the emergence of cell lineages, the role of lineage plasticity, the capability of altering phenotypic character or even reversing cell differentiation toward an undifferentiated state, was highlighted. That reversibility allows for lineage transition, a necessary means of responding to organ injury but is also subject to stochastic perturbations in the epigenetic regulation controlling cell fate permitting lineage infidelity as a first step toward neoplastic transformation.

A question then occurs - what experimental evidence is there to support such pathologic lineage plasticity in the development and progression of cancer? A new report from Eric Gardner, Harold Varmus and a team from Cornell University (Gardiner et. al, Science 383; 603, 2024) report a relationship between neoplastic sensitivity to certain oncogenes driving the development of two separate histologic subtypes of lung cancer, adenocarcinoma (LUAD) or small cell lung cancer (SCLC), arising from two distinct cell lineages of the lung and the occurrence of histologic transformation between them. In the clinic, histologic transformation from an adenocarcinoma phenotype to transformed SCLC is observed not uncommonly in cancers of the lung or prostate following targeted therapy. The resulting high-grade neuroendocrine cancer invariably proves resistant to rescue treatment and is associated with shortened survival making this histologic event an important form of treatment resistance.

Their principal finding – the cell lineage from which a lung cancer phenotype arises, LUAD from type 2 alveolar cells (AT2) of the distal airway or SCLC from pulmonary neuroendocrine cells (PNEC) of the proximal airway, the cellular context of a neoplasm, determines the differential sensitivity to selected oncogenic drivers. The investigators also suggested a mechanism through which histologic transformation occurs, identifying a bottleneck in this transition characterized by an intermediate state of cell dedifferentiation identified as having a basal stem-like phenotype and implicating the occurrence of epigenetic reprogramming. Figure 1 taken from Gardiner et. al illustrates those findings.

Figure 1 caption Histological transformation through a stem-like intermediate. Gardner et al. developed a genetically engineered mouse model of histological transformation (HT) of EGFR-driven lung adenocarcinoma (LUAD) to small cell lung cancer (SCLC) and closely examined this phenomenon with single-cell RNA-sequencing. They found that the intermediate state between LUAD and SCLC was stem-like and most closely resembled a pulmonary basal cell.

Gardiner's experimental approach began by developing a genetically engineered murine model using recombination technology to introduce into the mouse genome a set of conditional alleles for oncogenic mutations in EGFR (EGFR^L858R) or Myc (Myc^T58A), tumor suppressor genes Rb and Trp53, and the Mapk kinase pathway regulator Pten. Conditional alleles allowed the experimenters to either turn on or off the expression of the oncogenes of interest through externally administered agents, principally doxycycline (DOX) or tamoxifen (T). By then inducing either LUAD or SCLC with an adenovirus vector targeting cell-specific promoters for genes characteristic of AT2 cells (surfactant) or PNEC cells (Ascl1) they were able to assess the relative sensitivity to one or more oncogenic drivers measured through mouse survival, tumor histopathology correlated with gene expression. Using the Cre-Lox system of DNA recombination to tag tumor lineage of origin allowed the investigators to trace the lineage over time of either cell type during experimental oncogene manipulations.

In experiments focusing first on LUAD, in late-stage disease, if the investigators induced oncogenic EGFR expression by DOX and then sustained that expression, the animals would rapidly succumb. If they turned on EGFR by DOX but then withdrew DOX for a period and then reintroduced DOX, animals would experience better survival, confirming a gene dose-response relationship to establish causality.  If DOX exposure was followed by sustained removal of DOX, suppressing oncogenic EGFR influence, the animals would be left with a small population of residual disease. By then activating oncogenic Myc in those residual cells they observed a small population of cells in an intermediate state exhibiting basal stem-like features representing a state of dedifferentiation. Their small number suggests they represent a population 'bottleneck',  Importanly these undifferentiated cells are highly sensitive to the Myc oncogene,  a phenotypic charactersitic quite different from  of theoriginal cell lineage. That new found Myc sensitivity then drives the subsequent histologic transformed into SCLC. That transformation efficiency could then be further enhanced if experimenters also deleted the tumor suppressor gene, Rb. In clinical oncology, human SCLC is consistently marked by Rb deletion.  Figure 2 is taken from an accompanying summary of the experiments by Anton Berns of the Netherlands Cancer Institute to illustrates those transition findings.

 

Throughout the experiments beginning prior to, during and then after DOX exposure to control oncogenic EGFR, the investigators assessed the accompanying transcriptional programs active at those times employing single-cell RNA sequencing. By doing so they observed a continuous transition in cell state over the time course of On-Off DOX sequences so that as oncogenic EGFR in LUAD was suppressed, a reciprocal emergence of genetic markers associated with SCLC (synaptophysin, Ascl1) was observed. This experiment replicates what might be occurring in humans with EGFR^L858R positive LUAD following prolonged targeted exposure to therapeutic EGFR antagonists.

The experiments also informed our understanding of cellular context in determining oncogene sensitivity. They found that PNEC cells were uniformly sensitive to the transforming capability of Myc. AT2 cells on the other hand are intolerant to Myc, experiencing senescence following oncogenic exposure. LUAD generally resists the effects of oncogenic Myc unless there is co-expression of oncogenic EGFR. Oncogenic EGFR exerts no effect on SCLC.

Gardiner summarizes their findings concerning histologic transformation from LUAD, sensitive to EGFR transforming to SCLC, sensitive to Myc as representing a shift in the dominant oncogenic program operative from the initial lung cancer subtype to account for the transformation process. Using lineage tracing they identified a key intermediate transition state exhibiting genetic markers associated with a stem-like basal cell, an indication of epigenetic reprogramming of the LUAD cells, ‘uncovering’ an emergent Myc sensitivity. That bottleneck, slowing the escape from the selection pressure of targeted therapy, can then be ‘decompressed’ when the basal cells adopt Myc sensitivity permitting relief from the transformation holdup. By adopting a new phenotype, the cancer may now thrive in its new environment. Gardiner concludes by pointing to the need for therapeutic inhibitors for Myc to address this end around ploy by lung cancer, an ability that oncologists presently do not possess.

In an upcoming post, the Commentary will conclude a survey of epigenetic regulation and carcinogenesis by stepping back to take a big picture view of cancer as a model for disorder, an inescapable characteristic of our universe as described by the laws of thermodynamics. We will examine how life forms extract usable energy from their environment to delay this inevitability using a mechanism which physicist Brian Greene has described as the “entropic 2-step”.