Figure 1.
CgA coding regions, putative functional domains and targets.
The primary transcript, on chromosome 14, is derived from 8 exons and includes exon 1 which is untranslated but contains a signal peptide region for protein processing. In this study, PCR was performed using intron spanning primers to examine exons I-VI. Mature CgA mRNA includes 439 coding base pairs which are translated into a primary peptide of 431 amino acids. Processing of CgA following cleavage at dibasic and monobasic residues e.g. by PC1/3 and CPE results in production of a range of intermediate peptides as well biologically active peptides [36].
Figure 2.
Chromogranin A expression in normal mucosa, EC cells, small intestinal NENs (SI-NENs) and primary and metastatic SI-NEN cell lines.
CgA mRNA expression in normal human mucosa (NML), EC cell preparations (EC), localized NENs (PRIM), primaries with metastasis (MET PRIM) and liver metastases (METS) demonstrated that all NENs expressed higher CgA levels (Kruskal-Wallis p<0.0001) compared to normal mucosa (*p<0.001) or normal EC cells (#p<0.001) (a). Levels of CgA protein expression, measured by ELISA, showed a similar pattern (2C, Kruskal-Wallis p<0.0001) and was increased in PRIM (*p<0.01), MET PRIM (*p<0.001) and METS (*p<0.05) compared to normal mucosa. CgA western blot in normal mucosa, normal EC cells and SI-NENs identified a mature CgA band of 75-80 kDa in all NENs but not in normal mucosa or EC cells (2E). Fragment sizes included peptides ranging in size from ~30-60kDa, consistent with CgA processing intermediates [36]. In cell lines, CgA mRNA was expressed in higher levels in the two primary cell lines in comparison to metastatic cell lines (2B, Kruskal-Wallis p<0.0001), particularly in P-STS CgA was over-expressed compared to H-STS (* p<0.001) and L-STS cells (#p<0.01). In KRJ-1 CgA was also elevated in comparison to H-STS cells (*p<0.01, 2B). Protein level (ELISA) followed similar pattern (Kruskal-Wallis p=0.0273, *p<0.05, 2D). Using western blot, total CgA (75-80 kDa) was identified in all cell lines, highest in the primary cell lines KRJ-1 and P-STS (2F). Band sizes consistent with CgA processing were evident and exhibited different patterns of expression consistent with alterations in translational modifications (2F). No external receptor was identified for CgA, but Cy5-labeled immunofluorescence (IF) was identified within KRJ-I and H-STS cells. We interpret this dot-like signal to reflect intracellular uptake of this CgA peptide (2G).
Figure 3.
CgA processing enzyme prohormone convertase expression and the effect of tumor growth on CgA and processing.
PCSK1 mRNA expression was increased in SI-NEN metastases (METS) and primaries with metastasis (MET PRIM) compared to normal mucosa (NML, *p<0.05) and normal EC cells (EC, #p<0.05) (3A, Kruskal-Wallis p=0.0003). Western blot analysis confirmed that protein levels of prohormone convertase 1-3 were elevated in metastases compared to normal mucosa (3C,*p<0.05).
CgA mRNA (3B, 6 exons) and protein (3D) were elevated at the plateau growth phase (day 7) compared to logarithmic growth (day 2) in H-STS cells (*p<0.05). PC1-3 proteins were decreased at day 7 (3D), which can be discussed as one reason for the elevation of total intracellular CgA at this time point.
Mean±SEM. PCSK1: prohormone convertase 1.
Figure 4.
CgA silencing, processing enzyme inhibition and functional analysis of CgA peptides in SI-NEN cell lines.
After successfully silencing CgA in H-STS cells (data not shown), proliferation was significantly decreased (4A, *p<0.05). Secretion of CgA (p<0.01) and 5-HT (4B, *p<0.05) was significantly reduced following CgA antisense. Inhibition of the CgA processing enzyme prohormone convertase using Decanoyl-Arg-Val-Lys-Arg-CMK also decreased proliferation of H-STS cells (25 µM [data not shown] and 50 µM, 4D, *p<0.05). Additionally, secretion of CgA and 5-HT (4E, *p<0.05) was also significantly reduced. Decreases in CgA and its fragments (Vasostatin II and Pancreastatin) after treatment with the prohormone convertase inhibitor were confirmed with western blot (4C and F). Chromostatin (<20 kDa) was too small to appear on this WB. The fragments Vasostatin I and II significantly stimulated proliferation (up to 60%, *p<0.02) in both metastatic cell lines (L-STS and H-STS, 4H and I, square) but had no effects on the primary tumor cell lines. Chromostatin inhibited the well-differentiated localized NEN cell line proliferation (P-STS, 4G, square, ~50%, *p<0.05) but not proliferation of the less well-differentiated cell line, KRJ-I. Mean±SEM; n=6, CON: control, KD: knockdown, SCR: scrambled, P: P-STS, K:KRJ-1, L: L-STS, H: H-STS. 5-HT: Serotonin.
Figure 5.
Effect of Vasostatin I and Chromostatin on AKT phosphorylation in metastatic and localized NEN cell lines.
Vasostatin I stimulated AKT phosphorylation in the liver metastasis (H-STS) (CASE ELISA: 50%, *p<0.04, western blot: 25%) and could be completely reversed by pre-incubation with RAD001 (5A/C, #p<0.01). AKT antisense reversed vasostatin-mediated proliferation (BrdU uptake) (5E). In contrast, chromostatin, inhibited AKT signaling in the primary cell line (P-STS) (5B/D, ~25%, *p<0.05). AKT antisense reversed chromostatin-mediated inhibition of proliferation (BrdU uptake) (5F). Mean±SD; AS = antisense, CON: control, SCR: scrambled, V-I: vasostatin I, R: RAD001, CST: chromostatin.