Small molecule amiloride modulates oncogenic RNA alternative splicing to devitalize human cancer cells

Abstract

Alternative splicing involves differential exon selection of a gene transcript to generate mRNA and protein isoforms with structural and functional diversity. Abnormal alternative splicing has been shown to be associated with malignant phenotypes of cancer cells, such as chemo-resistance and invasive activity. Screening small molecules and drugs for modulating RNA splicing in human hepatocellular carcinoma cell line Huh-7, we discovered that amiloride, distinct from four pH-affecting amiloride analogues, could "normalize" the splicing of BCL-X, HIPK3 and RON/MISTR1 transcripts. Our proteomic analyses of amiloride-treated cells detected hypo-phosphorylation of splicing factor SF2/ASF, and decreased levels of SRp20 and two un-identified SR proteins. We further observed decreased phosphorylation of AKT, ERK1/2 and PP1, and increased phosphorylation of p38 and JNK, suggesting that amiloride treatment down-regulates kinases and up-regulates phosphatases in the signal pathways known to affect splicing factor protein phosphorylation. These amiloride effects of "normalized" oncogenic RNA splicing and splicing factor hypo-phosphorylation were both abrogated by pre-treatment with a PP1 inhibitor. Global exon array of amiloride-treated Huh-7 cells detected splicing pattern changes involving 584 exons in 551 gene transcripts, many of which encode proteins playing key roles in ion transport, cellular matrix formation, cytoskeleton remodeling, and genome maintenance. Cellular functional analyses revealed subsequent invasion and migration defects, cell cycle disruption, cytokinesis impairment, and lethal DNA degradation in amiloride-treated Huh-7 cells. Other human solid tumor and leukemic cells, but not a few normal cells, showed similar amiloride-altered RNA splicing with devitalized consequence. This study thus provides mechanistic underpinnings for exploiting small molecule modulation of RNA splicing for cancer therapeutics.

Figure 1. Modification of splicing patterns of BCL-X, HIPK3 and RON gene transcripts by amiloride.

RNA was extracted from Huh-7 cells exposed to growth medium containing amiloride or other pHi modulator amiloride derivatives and examined for the oncogenic RNA alternative splicing by RT-PCR, as described in Materials and Methods. Panel (A) shows that amiloride increased the use of upstream alternative 5′-splice site within exon 2 that yields the apoptotic BCL-XS isoform (upper row), increased the exon 11 (U exon) skipping of HIPK3 that increases the apoptotic isoform (middle row), and slightly increased the exon 11 skipping of RON at 0.5 mM concentration (lower row). Panel (B) shows that splicing patterns of the tested genes were not significantly affected by four amiloride derivatives, including 5-(N-ethyl-N-isopropyl)-amiloride or EIPA; 5-(N-methyl-N-isobutyl)amiloride or MIBA; 5-(N,N-dimethyl)amiloride or DMA; and 5-(N,N-hexamethylene)amiloride or HMA, at equivalent concentrations of pHi modulation.

Figure 2. Amiloride effects on cellular phospho-serine proteins.

2D-gel electrophoresis analysis of proteins extracted from Huh-7 cells without (left) and with (right) 0.5mM amiloride treatment for 24 hours showed significant changes for at least ten spots, of which seven (indicated by arrows) were isolated for nano-LC-MS/MS spectrometry analysis and protein identification, as shown in Table 1.

Figure 3. Effects of amiloride treatment on phosphorylation of SR-related proteins.

Western blot analysis of subcellular fractions of Huh-7 cells with and without 0.5 mM amiloride 24-hour treatment was performed using specific antibodies against various splicing protein factors and β-tubulin. (Panel A) The amiloride treatment increased the dephosphorylated SF2/ASF forms in both cytoplasmic and nuclear fractions. (Panel B) The amiloride treatment decreased the expression of SRp20 and two un-identified phosphorylated SR proteins in both C (cytoplasmic) and N (nuclear) cellular fractions.

Figure 4. Phosphorylation of Akt kinase (activation) and PP1 phosphatase (inactivation) in relation to amiloride-altered AS mechanisms.

(Panel A). Western blots show significant decreases of cytoplasmic p-Akt, nuclear p-Akt and nuclear p-PP1 in Huh-7 cells treated with amiloride at 0.3 mM and above. (Panel B) RT-PCR and western blot analyses show that pre-treatment with a PP1 inhibitor, okadaic acid, abrogated the amiloride effects on the BCL-X and HIPK3 oncogenic RNA splicing, the phosphorylation of ASF2/ASF splicing factor and the level of SRp20 in Huh-7 cells.

Figure 5. Gene ontology distribution of 495 protein-encoding gene transcripts with amiloride-modulated AS patterns in Huh-7 cells.

Bioinformatic analyses of the 495 gene transcripts with amiloride-modified alternative splicing, detected by the genome-wide exon-array measurements, were performed according to gene ontology categories of Biological Process (A), Molecular Function (B) and Cellular Component (C). Amiloride-altered AS patterns of APAF1, CRK,MBNL2, MIZF,WAC, PAPD5 and SURVIVIN specific protein-encoding gene transcripts detected by the exon-array analysis were validated by RT-PCR of spliced RNA isoforms (Panel D).

Figure 6. Cell migration and post-mitotic cytokinesis disorders subsequent to amiloride-altered AS of related functional gene transcripts.

(Panel A). Cell migration was determined in confluent Huh-7 cell monolayers scraped with a plastic scraper, exposed to growth medium with or without 0.45 mM amiloride for 48 hours, rinsed 2x with PBS and onward fed growth medium without amiloride. Serial microscopic photographs of same areas (marked green or red on the bottom side of dishes) taken at 48 and 96 hours show complete inhibition of cell migration from the edge of amiloride-treated monolayer. (Panel B) Cells exposed to the indicated concentrations of amiloride in growth medium for 36 hours were fixed with paraformaldehyde, permealized and stained with DAPI and FITC-phalloidin for observation by confocal microscopy. DAPI-stained nuclei are magenta-pseudocolored in merged confocal photographs. (Panel C) Inhibitory effect of amiloride on cytokinesis was observed with mitotic Huh-7 cells shaken off from the dish, re-plated in growth medium containing different concentrations of amiloride. The dividing daughter cells were fixed at 24, 48 and 72 hours, stained and photographed microscopically. (Panel D) Measurement of distances between the two nuclei of dividing or divided daughter cells shows that amiloride inhibits post-mitotic cytokinesis and separation of daughter cells. The diameter of the cell nucleus is set as 10. The average and standard deviation of each sample point are from 15 to 20 daughter cell pairs.

Figure 7. Growth inhibitory and apoptotic/cytotoxic consequences of amiloride-induced AS alterations.

(Panel A) Huh-7 cells plated in 6-well dishes at 3×10⁵ cells per well overnight were changed to growth medium containing various concentrations of amiloride; cell numbers were determined one and three days later. Cell counts at 24 hours show growth inhibition by amiloride at 0.4 mM and above. Further cell count decreases were evident at 72 hours due to cell death and detachment in medium containing amiloride at 0.3 mM and above. (Panel B) Cytofluorometric analysis shows amiloride inhibition of cell cycle progression with accumulation of G2/M phase cells. (Panel C) Apoptosis is evident by cytofluorometric detection of annexinV binding in Huh-7 cells treated with 0.5 mM of amiloride. (Panel D) Marked nuclear DNA degradation was observed in Huh-7 cells treated with 0.4 mM amiloride but not with 0.2 mM amiloride, particularly noted at 48 and 72 hours.