Neuroendocrine cells: Potential cells of origin for small cell lung carcinoma

Small cell lung carcinoma (SCLC) is the most aggressive type of lung cancer, accounting for at least 14% of new lung cancer cases.¹ SCLC is characterized by expression of neuroendocrine markers, such as synaptophysin and calcitonin gene-related peptide. However, about 10% of SCLCs have areas of other lung cancer types, such as squamous cell carcinoma and adenocarcinoma. This observation suggests that, in addition to neuroendocrine cells, SCLC may originate from some not-yet-identified, multi-/bipotent stem/progenitor cells, which are able to differentiate into several cell types, including neuroendocrine cells. Alternatively, SCLC may be a result of phenotypical plasticity of other pulmonary non-neuroendocrine cells, such as Clara cells, which can generate ciliated cells of the lower airways, alveolar type 2 (AT2) cells, which are able to differentiate toward alveolar type 1 (AT1) cells of the alveolar ducts and acini or bronchio-alveolar stem cells (BASCs), which can give rise to Clara, AT1 and AT2 cells (Fig. 1).

Figure 1.

Models of potential origin of SCLC. Model 1 proposes that unipotent stem/progenitor cells or mature cells of neuroendocrine cell lineage give rise to SCLC. This model anticipates that cells lacking p53 and Rb may differentiate toward other types of lung cancer, such as adenocarcinoma and squamous cell carcinoma in some cases. Model 2 proposes that SCLC arises from stem/progenitor cells able to differentiate toward several cell types, including neuroendocrine cells. Model 3 suggests that at least some of SCLCs could arise from non-neuroendocrine cell lineages due to differentiation plasticity of cells deficient for p53 and Rb.

In their study, groups led by Julien Sage and Carla Kim² addressed the issue of the cell of SCLC origin by taking advantage of previous observations that Cre-loxP-mediated inactivation of the tumor suppressor genes p53 and Rb in mouse lungs resulted in neuroendocrine malignancies closely resembling human SCLC according to their morphological and molecular properties.³ Since p53 and Rb alterations are among the most common mutations in human SCLC,¹ this mouse model is likely to accurately reflect many facets of the disease pathogenesis.

By expressing Cre recombinase in different cell types of the lung, the authors determined that neither Clara cells nor AT2 cells nor BASCs gave rise to neuroendocrine neoplasms. The authors concluded that neuroendocrine cells were the most likely cells of origin of SCLC. A similar conclusion was recently reached by the group of Anton Berns.⁴ However, at variance with the study by Park et al.,² Sutherland et al.⁴ observed that deletion of p53 and Rb in AT2 cells does result in neuroendocrine neoplasms, albeit at lower rate (45 vs. 83%) and after increased median latency (452 vs. 362 d), as compared with neuroendocrine cell-specific deletion of these genes. As Park et al. note, the difference in observations can be explained by different efficiency of adenoviruses used by the two groups. However, it is less clear why mice expressing Cre under the control of AT2 and BASC-specific surfactant protein C (SPC) promoter have not developed neuroendocrine neoplasms similar to those initiated by SPC-Cre adenovirus used by sutherland et al. and Park et al. propose that constitutive deletion of p53 and Rb in transgenic mice may affect lung cells differently than acute deletion of these genes after administration of SPC-Cre adenovirus. it is also possible that unlike the 4.8 kb fragment of mouse SPC promoter used in the adenoviral vector, the 3.7 kb human SPC promoter fragment used in transgenic mice does not drive Cre expression in cells capable of neuroendocrine differentiation. Another possibility is that the high levels of adenoviral infection may affect the degree of cell susceptibility to increased differentiation plasticity after p53 and Rb inactivation, either directly or by initiating microenvironmental changes. Finally, a difference in genetic backgrounds of mice used by different groups could be a possible reason. Future studies based on additional, preferably not adenovirus-based, systems should clarify this discrepancy.

Taken together, the studies by Park et al. and sutherland et al. make an important step toward better understanding of SCLC pathogenesis by strengthening the possibility that neuroendocrine cell lineage is the primary source of SCLC while convincingly excluding Clara cells as a potential cell of origin. At the same time, these studies also highlight the need for better tools allowing characterization of pulmonary neuroendocrine cell lineage during ontogenesis. A recent report has indicated the existence of mouse embryonic stem/progenitor cells, which are able to differentiate toward all epithelial cell types, including neuroendocrine cells.⁵ Unfortunately, there has been no direct evidence that neuroendocrine cells are among the progeny of recently identified putative mouse adult lung stem/progenitor cells.^6,7 it is likely that further progress in these areas will firmly establish the cell of origin for SCLC, thereby facilitating development of new rationally designed diagnostics and therapeutics.