The Functional Characterization of Normal and Neoplastic Human Enterochromaffin Cells

Abstract

Neuroendocrine regulation of small intestinal (SI) function is poorly understood because pure neuroendocrine cells are unavailable, whereas the biological basis of SI carcinoid tumors is unknown because neoplastic human enterochromaffin (EC) cells are unavailable. The objective of this study was to define the secretory regulation and transcriptome of naive and neoplastic SI neuroendocrine cells. EC cells from human ilea were isolated and purified, and a malignant EC cell carcinoid cell line (KRJ-I) was characterized. Human ilea from right hemicolectomies were pronase/collagenase digested and Nycodenz gradient centrifuged, and EC cells were fluorescence-activated cell sorting (FACS) sorted after acridine orange labeling. Enrichment was defined by immunostaining, gene expression, serotonin (5-HT) content, and real-time RT-PCR. Naive FACS-sorted EC and KRJ-I cells were cultured, and 5-HT secretion was measured after stimulation with forskolin, isoproterenol, acetylcholine, gamma-aminobutyric acid A (GABA(A)), pituitary adenylate cyclase-activating polypeptide (PACAP)-38, and gastrin. Normal and neoplastic EC cell transcriptomes were acquired by Affymetrix profiling (U133A). FACS produced 100 +/- 0.3% (chromogranin A staining) and 99 +/- 0.7% pure EC cells by immunostaining for tryptophan hydroxylase with greater than 67-fold enrichment and a 5-HT content of 180 +/- 18 ng/mg protein (mucosa, 3.5 +/- 0.9). Forskolin- and isoproterenol-stimulated 5-HT secretion was 10-100 times more potent for naive cells (EC(50), 1.8 x 10(-9) m; 5.1 x 10(-9) m) than neoplastic cells (EC(50), 2.1 x 10(-7) m; 8.1 x 10(-8) m), but the effect of PACAP-38 was similar (EC(50), 1 x 10(-7) m). Isoproterenol stimulated cAMP levels 1.6 +/- 0.1-fold vs. basal (EC(50), 2.7 x 10(-9) m). Acetylcholine inhibited naive EC cell 5-HT secretion more potently than neoplastic (IC(50), 3.2 x 10(-9) vs. 1.6 x 10(-7) m), whereas GABA(A) was more potent in neoplastic cells (IC(50), 3.9 x 10(-10) vs. 4.4 x 10(-9) m). Octreotide inhibited naive, but not neoplastic, basal 5-HT secretion. Gastrin had no effect on 5-HT secretion. Comparison of naive and neoplastic transcriptomes revealed shared neuroendocrine and EC cell-specific marker genes. Real-time PCR confirmed that expression of adrenergic (beta1), somatostatinergic (SST(R)2), and neural (VPAC(1) and GABA(A)) receptors occurred on both cell types, but PACAP type 1 (PAC(1)) and cholecystokinin type 2 (CCK(2)) were undetectable. The putative carcinoid malignancy genes (MTA1 and MAGE-D2) were unique to the neoplastic EC cell transcriptome. These data support novel methodology to purify live human EC cells for functional characterization and transcriptome assessment, which will allow identification of new targets to control the secretion and proliferation of SI carcinoids.