A Dominant Negative Antisense Approach Targeting β-Catenin

Abstract

There have been many attempts to unveil the therapeutic potential of antisense molecules during the last decade. Due to its specific role in canonical Wnt signalling, β-catenin is a potential target for an antisense-based antitumour therapy. In order to establish such a strategy with peptide nucleic acids, we developed a reporter assay for quantification of antisense effects. The luciferase-based assay detects splice blocking with high sensitivity. Using this assay, we show that the splice donor of exon 13 of β-catenin is particularly suitable for an antisense strategy, as it results in a truncated protein which lacks transactivating functions. Since the truncated proteins retain the interactions with Tcf/Lef proteins, they act in a dominant negative fashion competing with wild-type proteins and thus blocking the transcriptional activity of β-catenin. Furthermore, we show that the truncation does not interfere with binding of cadherin and α-catenin, both essential for its function in cell adhesion. Therefore, the antisense strategy blocks Wnt signalling with high efficiency but retains other important functions of β-catenin.

Keywords: Antisense; Morpholino; PNA; Wnt signalling; β-Catenin.

Fig. 1

Splice-based reporter assay. a Schematic view of the reporter assay. AS (antisense reagent), SD (splice donor), SA (splice acceptor), P1 and P2 (primers for PCR). b HeLa pSplice3basis1 cells were electroporated with the indicated amount of PNA1. After 24 h luciferase activity was determined. Nluc values were divided through internal Fluc reference and normalised to control cells (0) electroporated without PNA. Data represent mean values of at least 3 independent experiments, except for 32 µM (single determination), and error bars indicate SEM. c PCR from cDNA of antisense-treated pSplice3basis1 cells treated with 16 µM MO1 compared to equally treated control cells without reagent (ctrl). Scraping was used for delivery, and cells were harvested after 24 h

Fig. 2

Targeting β-catenin with PNA and morpholino oligos. a Schematic view of the β-catenin gene including exon/intron map with positions of exon 13 SD (SD13) and splice-specific PCR primers (P13, P14), numbering of exons according to NM001904.3. The positions of the exons relative to the armadillo repeats of the protein are indicated. Abbreviations: amino acid (aa), armadillo (ARM), N-terminal domain (NTD), C-terminal domain (CTD). b Luciferase assay with different PNAs targeting β-catenin at the SD of exon 13. HeLa pSplice3βcat Ex13 SD cells were electroporated with 128 µM of the indicated PNAs. Nluc values were divided through the internal reference Fluc and normalised to control cells (ctrl) electroporated without PNA. Data represent mean values of at least 3 independent experiments, and error bars indicate SEM

Fig. 3

Effect of splice block at exon 13 of β-catenin in SW480 cells. a SW480 cells were electroporated with the indicated amounts of MO3 and PNA18. PCR from cDNA with primers P13 and P14, specific for the presence of intron 13 (304 bp band) as a result of the splice block and 50 bp resulting from normal splicing. Quantification of the band ratio normalised to MO3 is shown below (quant.). b Effect on β-catenin protein level in SW480 cells electroporated with 32 µM MO3, extracts prepared after 48 h. Bands for β-catenin and GAPDH are indicated. Quantification is shown for two independent experiments (resulting in a mean protein level of 40% for the MO3-treated samples compared to the control). c mRNA levels of the β-catenin target genes axin, VEGF and c-myc determined by qPCR normalised to GAPDH and mock-treated control cells (ctrl). SW480 cells were electroporated with MO3 (16 µM) and PNA18 (256 µM). Data represent mean values of duplicates. Error bars indicate SEM

Fig. 4

Protein–protein interactions of truncated β-catenin. a Schematic presentation of truncated β-catenin versions. Position of armadillo (ARM) repeats is indicated as well as the interaction sites of known binding partners [45, 46, 48]. aa (amino acids), NTD (N-terminal domain), CTD (C-terminal domain), ARM (armadillo). b Mammalian two-hybrid assay with the baits Lef1 (pMCLef1mZFb6), Tcf3 (pMCZFb6Tcf3), cadherin (pMCZFghe-cadherin) or β-catenin (pMCZFg alpha-catenin) together with the preys β-catenin full length (full; pMC65gDPβcat), β-catenin(+I13) (pMC65gDPβcat[1-694]), β-catenin(E13skip) (pMC65gDPβcat[1-653]), β-catenin(+I6) (pMC65gDPβcat[plusintron6]), β-catenin(E6skip) (pMC65gDPβcat[exon6skip]) and no prey (ctrl). All Fluc values were divided through the internal reference Gluc and then normalised to samples lacking bait and prey. Data represent mean values of at least 3 independent experiments, and error bars indicate SEM

Fig. 5

Dominant negative effect of truncated β-catenin versions. a 70 ng pMlucF6lefcons, 2 ng pMCGlucS and 10 ng β-catenin full length [pKCDPβcat] were co-transfected into HeLa cells with 30 ng truncated β-catenin versions: pKCDPβcat(E6skip), pKCDPβcat(+I6), pKCDPβcat(1-694)(+I13), pKCDPβcat(1-653)(E13skip) and empty pKC vector (vector ctrl). Measurements were taken 24 h after transfection. All values were normalised to samples with empty expression vector pKC. Data represent mean values of at least 3 independent experiments, and error bars indicate SEM. b Concentration-dependent dominant negative effect of β-catenin(E13skip) (pKCDPβcat[1-653]) compared to β-catenin full length (full; pKCDPβcat). 70 ng pMlucF6lefcons, 2 ng pMCGlucS and 10 ng pKCDPβcat were co-transfected with 3, 10 and 30 ng pKCDPβcat or pKCDPβcat(1-653) into HeLa cells. Values were normalised to controls without expression vector for β-catenin. All Fluc reporter values were divided through Gluc internal reference. Data represent mean values of at least 3 independent experiments, and error bars indicate SEM