Generation and characterization of hESC-derived lung cells and formation of xenografts with RPM cells.

A) Schematic of the protocol used to generate PNECs by stepwise differentiation of hPSCs to form: definitive endoderm (DE), day 3; anterior foregut endoderm (AFE), day 6; and lung progenitor cells (LPs) days 15 -25. LPs were further differentiated into the types of lung cells (LCs) found in mature human lung parenchyma and airway epithelium, days 25 -55. DAPT (10 μM) encourages the formation of PNECs, and addition of doxycycline (1 μM; DOX) induces expression of shRNAs against RB1 and TP53 mRNAs, as well as expression of cMYC or cMYC (T58A), as described in the text. B) Western blot of extracts of RUES2 LCs at day 25 of differentiation protocol treated with DOX (1 μM for 72hrs); cells unexposed to DOX served as negative expression controls. Apparent differences in cMYC protein levels may be attributable to the HA-tagged version of cMYC (T58A), which migrates slightly slower that wildtype cMYC protein. C) Schematic representation of tumorigenesis experiments comparing injection sites (renal capsule or subcutaneous), DOX treatment (+/-DOX diet), and genotypes (see Methods for additional details). Total numbers of animals are 6-7 per experimental arm with 2 injection sites per mouse (right and left flank). Renal capsule injections were performed on a single kidney. Transgenic lines of RUES2 hESCs were differentiated and grown in DAPT (10 μM) from days 25 -55. At day 55, PNECs were separated from other LCs by sorting for PE+ CGRP-expressing cells (see Methods). PNECs were then injected either subcutaneously or into the renal capsular space in NOG mice, half of which then received DOX in their feed as described in Materials and Methods. D) Table summary of experiments with xenografted mice, indicating the number of animals that developed visible tumors (≥250 mm3 in volume) at the site of injection or the number of visible metastases in the liver or lung. *, P < 0.05; **, P < 0.01 by Fisher’s test to denote significant differences between mice that did and did not receive DOX diet. As before, abbreviations for cell lines are: RP = shRB1 + shTP53; RPM = shRB1 + shTP53 + WT cMYC; RPM (T58A) = shRB1 + shTP53 + cMYC (T58A).

Similar growth of RUES2-derived RPM and RPM (T58A) subcutaneous tumors and DOX-dependent growth.

A) Subcutaneous engraftment of 106 viable cells at ∼ day 50 of RUES2 lung differentiation protocol from indicated genotypes. Cells were cultured in the presence of media containing 1 μM DOX and 10 μM DAPT from days 25 – 55. All immunocompromised mice were maintained on DOX diet throughout this study; n=5/arm. B) Subcutaneous engraftment of 106 viable RPM tumor cells from the first passage mouse-to-mouse into immunocompromised mice maintained on DOX or normal chow; n=5 animals per arm with single flank engaftments, +/-SEM. One month following no DOX, mice were placed on DOX chow (DOX “add-back”); data were obtained one month following DOX chow add-back (2 months on study). C) Representative H&E tumor histology from RPM or RPM (T58A) engraftments; 2X (scalebar 1mm) and 20X (scalebar 100um). Representative whole lung sagittal sections showed no evidence of gross metastasis from flank injections at study endpoints; n=3 analyzed per genotype. Dashed lines shown for boundaries of lung lobes.

Effects of cMYC on the gross and histopathological appearance of xenografts grown in immunodeficient mice fed with DOX for 3-4 months.

A) Subcutaneous xenografts formed with hESC-derived PNECs with RP, RPM or RPM (T58A) genotypes. Photographs of representative tumors formed with cells of the indicated genotypes are shown for RP and RPM; indicated scale of 1cm. B) Quantification of the tumor sizes and paired comparisons for a secondary in vivo experiment; n=5 animals per arm with single subcutaneous engraftmetns; **P<0.01. C) H&E staining of the indicated tumors from panel B. D) Gross and histologic pathology of renal capsular xenografts and liver metastases formed with the RPM or RPM (T58A) cells. Left panels, gross appearance of representative tumors within the liver (metastasis) or kidney (primary); right panels, H&E staining of primary and metastatic tumors (T) formed in kidney (K) and liver (L) in the RPM (left) or RPM (T58A) model. E) Immunostaining of tumor samples from panel D stained with antisera for Ki67 (left) or MYC (right). F) Immunofluorescence staining for neuroendocrine markers NEUROD1 and CD56 from sections in E.

Single cell and bulk RNA profiling of RUES2-derived RPM tumors and comparison with primary human SCLC.

A) UMAP projection of RP samples from Chen et al (2019) and RPM samples [this study] with 3 major cellular lineages annotated by color. B) Expression of SCLC subtype markers across dataset in A. C) Neuroendocrine fraction of cells by sample ID. D) Subclustering of the “neuroendocrine differentiation” cluster from A. E) Dot plot of differential cluster markers in the subclustering analysis from E. F) Cell cycle evaluation of NE and NE-Variant clusters indicated by color and fraction of cells. G) SCLC subtype enrichment scores from Chan et al (2021); cluster markers in NE and NE-Variant cells from RPM tumors. H) Bulk RNA-sequencing subtype estimation based on Chan et al (2021) SCLC subtypes. Published labels were obtained from Rudin et al (2019). Bulk patient RNA-sequencing data (reads per kilobase per million mapped reads (RPKM)) were compared to bulk RNA-sequencing of select RPM tumors and metastatic samples. Primary tumor and liver metastases were obtained from two pairs (animals #1 and #4) of mice engrafted with RUES2-derived RPM tumor cells into the renal capsule and grossly macro-dissected during necropsy.

Gene expression and characterization of RUES2-derived tumors.

A) EGFP expression across the full dataset integrating RP and RPM cells in addition to RPM tumor cells. B) Individual sample IDs and relative contribution (weighted; red) to initial clustering distribution for the 6 sample sets used in this combined analysis. C) Differential gene expression (DEG) between NE-Variant and NE cluster. Select genes are called out in red. D) Select gene set enrichment analysis for top 3 differentially expressed programs between NEUROD1-high (NE-variant) and ASCL1-high (NE) subsets of cells from only RPM tumor cells; shown as normalized enrichment scores (NES). Positive enrichment is greater activation in the NE-Variant subset (i.e. the NEUROD1+ subset), whereas negative enrichment is greater in the NE subset. E) Cellular density of co-expression patterns for ASCL1 and NEUROD1 across the neuroendocrine differentiation subset. F) Scatter density plot of cMYC and NEUROD1 expression across all cells in the neuroendocrine differentiation cluster. G) Plot of cMYC and NEUROD1 expression in bulk expression data derived from RPM tumors; data are shown as log2 of transcripts per million (TPM). H) Enrichment score for SCLC subtype markers in NE and NE-Variant cells from RP cell cultures (2D), demonstrating preferably enrichment of the SCLC-A subtype in the absence of over-expression of a MYC transgene.