Endothelin-converting enzyme-1 (ECE-1) is post-transcriptionally regulated by alternative polyadenylation

Abstract

Endothelin-converting enzyme-1 (ECE-1) is the enzyme predominantly responsible for producing active endothelin-1 (ET-1), a mitogenic peptide implicated in the aetiology of a number of diseases, including cancer. Elevated levels of ECE-1 have been observed in a range of malignancies, with high expression conferring poor prognosis and aiding the acquisition of androgen independence in prostate cancer. The mechanisms regulating the expression of ECE-1 in cancer cells are poorly understood, hampering the development of novel therapies targeting the endothelin axis. Here we provide evidence that the expression of ECE-1 is markedly inhibited by its 3'UTR, and that alternative polyadenylation (APA) results in the production of ECE-1 transcripts with truncated 3'UTRs which promote elevated protein expression. Abolition of the ECE-1 APA sites reduced protein expression from a reporter vector in prostate cancer cells, suggesting these sites are functional. This is the first study to identify ECE-1 as a target for APA, a regulatory mechanism aberrantly activated in cancer cells, and provides novel information about the mechanisms leading to ECE-1 overexpression in malignant cells.

Figure 1. The full-length ECE-1 3′UTR significantly reduces luciferase activity.

The full-length ECE-1 3′UTR was cloned into pMIR-REPORT luciferase vector (ECE UTR) and transfected into the cell lines indicated. Activities of Firefly and Renilla luciferase (endogenous control) were assayed 48 h post transfection. Data represent the ratio of Firefly/Renilla luciferase activities normalised to empty miReport vector (miReport) ± S.E.M. (n = 9, * = P<0.001 by Student t-test).

Figure 2. The 3′-UTR of ECE-1 represses ECE-1 protein expression.

PC-3, RWPE-1, Huh7 and CHO cells were transiently transfected with empty vector (1), V5-tagged ECE-1c cDNA (2) or V5-tagged ECE-1c cDNA with full-length 3′ UTR (3) and immunoblotted for V5 (A). β-actin was used as a loading control. The PC-3 blot was reprobed for ECE-1 expression. Data represent the ratio of V5:actin expression ± S.E.M. (B; n = 3, * = P<0.05, ** = P<0.01, *** = P<0.001 by Student t-test). Real-time PCR analysis of transcript abundance in transfected PC-3 cells shown in A. V5 expression was normalised to neomycin expression (transfection control) and U6 expression (endogenous control). n = 3, * = P<0.005 by Student t-test (C).

Figure 3. The effect of the ECE-1 3′UTR on luciferase activity is due to length not specific sequence.

Schematic representation of the truncated UTRs identified by 3′RACE which were then cloned downstream of the Firefly luciferase gene in pMIR-REPORT (A). Empty pMIR-REPORT, full-length ECE UTR, the six truncated UTRs and the equivalent six truncated UTRs controlled for length (RC) were transfected into PC-3 cells. Activities of Firefly and Renilla luciferase (endogenous control) were assayed 48 h post transfection. Data represents the ratio of Firefly/Renilla activities normalised to empty pMIR-REPORT vector (miReport) ± S.E.M. (n = 9) (B).

Figure 4. The effect of the ECE-1 3′UTR on ECE protein expression depends on UTR length.

PC-3 (A) or RWPE-1 (B) cells were transfected with V5-tagged ECE-1c cDNA only or with the truncated or full-length UTR cloned immediately downstream for 48 h. Total lysates (30 mg protein) were immunoblotted for V5; b-actin was used as a loading control. Data represent the ratio of V5:actin expression ± S.E.M. n = 3, * = P<0.05 by Student t-test.

Figure 5. The effect of the UTR on ECE protein expression is due to length not specific sequence.

PC-3 cells were transfected with V5-tagged cDNA, V5 cDNA plus 184 bp UTR, V5 cDNA plus 184 bp+RC, or V5 cDNA plus full-length UTR for 48 h. Total lysates (30 mg) were immunoblotted for V5; b-actin was used as a loading control (A). Data represent the ratio of V5:actin expression normalised to cDNA ± S.E.M. n = 3, *** = P<0.001 by Student t-test, n.s = not significant (B).

Figure 6. Removal of the APA sites reduces luciferase activity.

The SV40poly(A) tail sequence was deleted from the wild-type ECE-UTR miReport construct (ECE SV40). The six APA sites were then all abolished by site-directed mutagenesis (APA SV40). Constructs were transfected into PC-3 cells for 48 h. Data represent the ratio of Firefly/Renilla luciferase activities normalised to ECE SV40± S.E.M. (n = 9, * = P<0.05 by Student t-test).