Readthrough of premature termination codons in the adenomatous polyposis coli gene restores its biological activity in human cancer cells

Abstract

The APC tumor suppressor gene is frequently mutated in human colorectal cancer, with nonsense mutations accounting for 30% of all mutations in this gene. Reintroduction of the WT APC gene into cancer cells generally reduces tumorigenicity or induces apoptosis. In this study, we explored the possibility of using drugs to induce premature termination codon (PTC) readthrough (aminoglycosides, negamycin), as a means of reactivating endogenous APC. By quantifying the readthrough of 11 nonsense mutations in APC, we were able to identify those giving the highest levels of readthrough after treatment. For these mutations, we demonstrated that aminoglycoside or negamycin treatment led to a recovery of the biological activity of APC in cancer cell lines, and showed that the level of APC activity was proportional to the level of induced readthrough. These findings show that treatment with readthrough inducers should be considered as a potential strategy for treating cancers caused by nonsense mutations APC gene. They also provide a rational basis for identifying mutations responsive to readthrough inducers.

Figure 1. Identification of APC nonsense mutations responsive to aminoglycoside treatment.

Readthrough efficiencies for 11 nonsense mutations in the APC gene were assessed in NIH3T3 cells with and without gentamicin (800 µg/ml) treatment for 24 h. Two nonsense mutations (L360X and R1114X) displayed levels of gentamicin-induced readthrough of more than 0.5%. Means values are presented, together with the standard error of the mean (SEM) (n = 5).

Figure 2. APC biological activity is correlated with antibiotic-induced readthrough level in human colorectal cancer cells.

(A) Readthrough efficiencies for APC L360X nonsense mutations were determined in DLD-1 cells in the presence of G418 (10, 25, 50, 100 and 200 µg/ml), negamycin (1 mg/ml), amikacin (2 mg/ml) or gentamicin (800 µg/ml). (B) Readthrough efficiencies for APC R1114X nonsense mutations were determined in NIH3T3 cells in the presence of G418 (50, 100 and 200 µg/ml) or negamycin (1 mg/ml). The readthrough efficiencies in DLD-1 cells were consistent with those in NIH3T3 cells (data not shown). (C) Aminoglycoside treatment restored APC activity. APC binding to beta-catenin (reppression of the pTOPGlow reporter) is restored by the treatment of DLD-1 cells transiently transfected with cDNA APC L360X with G418 (10, 25, 50, 100 and 200 µg/ml), negamycin (1 mg/ml), amikacin (2 mg/ml) or gentamicin (800 µg/ml). (D) APC binding to beta-catenin (repression of pTOPGlow reporter) is restored by the treatment of LoVo cells carrying APC R1114X with G418 (50, 100 and 200 µg/ml) or negamycin (1 mg/ml). As a control, LoVo cells were cotransfected with the pTOPGlow reporter plasmid and either the APC-targeting siRNA (siRNA APC) or a non-targeting siRNA (siRNA NT) and treated with G418 (200 µg/ml). (E) Effect of the siRNA targeting APC mRNA. We transiently transfected LoVo cells with an siRNA targeting (siRNA APC) or not targeting (siRNA NT) APC mRNA. Quantitative PCR was used to determine mRNA levels. Results are expressed relative to the amount of mRNA in the presence of siRNA NT. Mean values are presented, together with the SEM (n = 3).

Figure 3. Western blot.

LoVo cells were left untreated or were treated with G418 (200 µg/ml) for 72 hours. Western blots were probed with the FE-9 antibody directed against the N-terminus of APC. The truncated forms corresponding to the mutated alleles present in LoVo cells are indicated by arrows. An extract from HeLa cells (APC WT) was used as a control; a band was detected at 311 kDa.