Evaluation of EGFR and COX pathway inhibition in human colon organoids of serrated polyposis and other hereditary cancer syndromes

Abstract

Serrated polyposis syndrome (SPS) presents with multiple sessile serrated lesions (SSL) in the large intestine and confers increased colorectal cancer (CRC) risk. However, the etiology of SPS is not known. SSL-derived organoids have not been previously studied but may help provide insights into SPS pathogenesis and identify novel biomarkers and chemopreventive strategies. This study examined effects of EGFR and COX pathway inhibition in organoid cultures derived from uninvolved colon and polyps of SPS patients. We also compared with organoids representing the hereditary gastrointestinal syndromes, Familial Adenomatous Polyposis (FAP) and Lynch syndrome (LS). Eighteen total organoid colon cultures were generated from uninvolved colon and polyps in SPS, FAP, LS, and non-syndromic screening colonoscopy patients. BRAF and KRAS mutation status was determined for each culture. Erlotinib (EGFR inhibitor) and sulindac (COX inhibitor) were applied individually and in combination. A 44-target gene custom mRNA panel (including WNT and COX pathway genes) and a 798-gene microRNA gene panel were used to quantitate organoid RNA expression by NanoString analysis. Erlotinib treatment significantly decreased levels of mRNAs associated with WNT and MAPK kinase signaling in organoids from uninvolved colon from all four patient categories and from all SSL and adenomatous polyps. Sulindac did not change the mRNA profile in any culture. Our findings suggest that EGFR inhibitors may contribute to the chemopreventive treatment of SSLs. These findings may also facilitate clinical trial design using these agents in SPS patients. Differentially expressed genes identified in our study (MYC, FOSL1, EGR1, IL33, LGR5 and FOXQ1) may be used to identify other new molecular targets for chemoprevention of SSLs.

Keywords: Colon organoid; Colon polyp; Erlotinib; Familial adenomatous polyposis; Gene expression; Lynch syndrome; Serrated polyposis; Sulindac.

Fig. 1

Photomicrographs of colon organoids derived from patients with Serrated Polyposis Syndrome (SPS) and Familial Adenomatous Polyposis (FAP). Left, colon organoids derived from uninvolved colon from an SPS patient by whole mount light microscopy (20× , top row), hematoxylin and eosin-stained tissue slide (10× , 2nd row), alcian blue stained tissue slide (3rd row, 20×) and fluorescence microscopy of mucin 2 (20× , bottom row) Center, colon organoids from a sessile serrated lesion (SSL). Right, colon organoids from an adenomatous polyp (AP) from FAP patient

Fig. 2

Hierarchical clustering of colon organoid RNA expression using a custom 48 gene NanoString panel. The log2 ratios for organoids derived from polyps (SSLs and APs) were determined by comparing each polyp to its paired uninvolved control. The log2 ratio for each uninvolved colon sample (CTRL, SPS, FAP and Lynch) was determined by comparing each uninvolved colon samples to the mean of all uninvolved colon samples. Red denotes increased expression, blue reduced expression, and white no change in expression. Organoid samples with red labeled text were positive for BRAF V600E mutation

Fig. 3

MicroRNA expression in colon organoids (A) Principal component analysis of microRNA expression in organoids derived from four SSLs from SPS patients and three APs from FAP patients. 3D figure shows first three components (PC1-PC3) accounting for approximately 70% of the variation in the data. Log2 ratios were calculated by comparing each polyp to its uninvolved control. Red denotes two SSLs and one AP with BRAF V600E mutation (n = 3) and blue denotes three SSLs and one AP without BRAF mutation (n = 4). (B) Relative expression of seven microRNAs (miRNAs) differentially expressed in three BRAF V600E mutant, two SSLs and one AP, compared to four BRAF wildtype, three SSLs and one AP, organoids. Bar graphs show the mean and standard error of normalized read counts for each miRNA. Statistical significance determined by DESeq2, FDR < 0.05

Fig. 4

RNA expression in control and erlotinib treated colon organoids (A) RNA expression of MYC, EGR1, IL33, FOSL1, LGR5 and FOXQ1 in control (C) and erlotinib (E) treated organoids derived from uninvolved colon from SPS, FAP, Lynch and control (non-syndromic patients with average cancer risk) patient cohorts. Bar graphs show the mean and standard error (n = 4–6) of normalized read counts for each gene. Statistical significance determined by DESeq2, FDR < 0.05. (B) RNA expression of MYC, EGR1, IL33, FOSL1, LGR5 and FOXQ1 in control (C) and erlotinib (E) treated organoids derived from SSLs from SPS patients and APs from FAP patients. Bar graphs show the mean and standard error (n = 4–6) of normalized read counts for each gene. Statistical significance determined by DESeq2, FDR < 0.05

Fig. 5

RNA expression in sulindac and erlotinib treated colon organoids (A) RNA expression of MYC, EGR1, IL33, FOSL1, LGR5 and FOXQ1 in control (C) and sulindac (S) treated organoids derived from uninvolved colon from SPS, FAP, Lynch and control patient cohorts. Bar graphs show the mean and standard error (n = 4–6) of normalized read counts for each gene. (B) RNA expression of MYC, EGR1, IL33, FOSL1, LGR5 and FOXQ1 in control (C) and erlotinib + sulindac combination (ES) treated organoids derived from SSLs from SPS patients and APs from FAP patients. Bar graphs show the mean and standard error (n = 4–6) of normalized read counts for each gene. Statistical significance determined by DESeq2, FDR < 0.05