Targeting protein arginine methyltransferase 5 inhibits colorectal cancer growth by decreasing arginine methylation of eIF4E and FGFR3

Abstract

Protein arginine methyltransferases (PRMTs) plays critical roles in cancer. PRMT5 has been implicated in several types of tumors. However, the role of PRMT5 in cancer development remains to be fully elucidated. Here, we provide evidence that PRMT5 is overexpressed in colorectal cancer (CRC) cells and patient-derived primary tumors, correlated with increased cell growth and decreased overall patient survival. Arginine methyltransferase inhibitor 1 (AMI-1)strongly inhibited tumor growth, increased the ratio of Bax/Bcl-2, and induced apoptosis in mouse CRC xenograt model. AMI-1 also induced apoptosis and decreased the migratory activity in several CRC cells. In CRC xenografts AMI-1 significantly decreased symmetric dimethylation of histone 4 (H4R3me2s), a histone mark of type II PRMT5, but not the expression of H4R3me2a, a histone mark of type I PRMTs. These results suggest that the inhibition of PRMT5 contributes to the antitumor efficacy of AMI-1. Chromatin immunoprecipitation (ChIP) identified FGFR3 and eIF4E as two key genes regulated by PRMT5. PRMT5 knockdown reduced the levels of H4R3me2s and H3R8me2s methylation on FGFR3 and eIF4E promoters, leading to decreased expressions of FGFR3 and eIF4E. Collectively, our findings provide new evidence that PRMT5 plays an important role in CRC pathogenesis through epigenetically regulating arginine methylation of oncogenes such as eIF4E and FGFR3.

Keywords: AMI-1; FGFR3; PRMT5; colorectal cancer; eIF4E.

Figure 1. PRMT5 expression in colorectal tumor and cell lines

A. Western blot analysis of PRMT5 expression in human normal colonic mucosal FHC cells and CRC cell lines. B. Western blot analysis of PRMT5 expression in CRC tissues C. and NATs (N).

Figure 2. PRMT5 is overexpressed in CRCs and negatively correlated with patient survival

A. Immunohistochemical staining of PRMT5 in CRC tissues and corresponding NATs. Positive cells were stained brown. Magnification, (a, c) 100×; (b, d,) 400×. B. Plot representation of scores based on the nuclear or cytoplasmic expression of PRMT5 in 90 CRC patients compared with matched normal tissues. C. Total PRMT5 score in adjacent normal and tumor tissues. D. Correlation analysis of PRMT5 expression status in the nucleus and patient survival.

Figure 3. AMI-1 inhibits CRC cell growth in vitro and in vivo

A. The inhibitory effects of AMI-1 (50 μM) on PRMTs activity. B. The effects of AMI-1 on cell proliferation of human CRC cell lines. C. The effects of AMI-1 on colony formation of CRC cell lines. D. The effects of AMI-1 on tumor formation in a nude mouse xenograft model. SW480 cells (1 × 10⁶) were injected s.c. into the right flank of nude mice. There were 10-11 mice each in treated and control groups. After 11 d, when tumor nodules were clearly visible, sites were injected 3 times per week with AMI-1 (0.5 mg in 100 μL of vehicle) or 100 μL vehicle (0.9% NaCl). Control animals and animals treated with AMI-1 were euthanized after 28 d and tumor were excised, measured and analyzed. E. The expression of pro-, anti-apoptotic protein, H4R3me2s and H4R3me2a in SW480 tumor exnograft was detected by Western blot analysis. F. Densitometric analysis of band intensities (H4R3me2s and H4R3me2a) using Image J software (normalized to β-actin). Controls were vehicle-treated group.

Figure 4. AMI-1 promotes apoptosis and decreases migratory activity of CRC cells

A. SW480 and HCT116 cells were treated with vehicle alone or AMI-1 and stained by Annexin V-fluorescein isothicyanate (FITC) and propidium iodide (PI), followed by flow cytometry. B. AMI-1 decreased migratory activity of CRC cells measured by Transwell assay. Representative photos of stained cells are shown with original magnification 200×. Controls were vehicle-treated group.

Figure 5. PRMT5 knockdown decreases CRC cell growth in vitro and in vivo

A. The effects of si-PRMT5 on cell proliferation of CRC HCT116, SW620, SW480, and LS-174T cells. B. The effects of si-PRMT5 on colony formation of CRC cells (HCT116 and SW480). Representative results of colony formation of only vehicle (left), si-NC (middle), and si-PRMT5 (right) SW480 and HCT116 cells. C. The effects of si-PRMT5 on CRC cell cycle. SW480 and HCT116 cells were transfected with si-NC or si-PRMT5 and subjected to cell cycle analysis. D. and E. The effects of si-PRMT5 on tumor formation in nude mouse xenograft model. The tumor volume and weight of si-PRMT5 group was significantly decreased compared with that of control group.

Figure 6. PRMT5 knockdown represses FGFR3 and eIF4E expression and decreases H3R8 and H4R3 methylation on their promoters

A. HCT116, SW620, and SW480 cells were transfected with si-NC or si-PRMT5. FGFR3 and eIF4E protein were analyzed by immunoblot 72 h later. B. FGFR3 and eIF4E mRNA expression in CRC tissues was analyzed by qRT-PCR. C. ChIP assays were performed in SW480 cells with antibodies (PRMT5, H3R8, H4R3 and normal IgG), and qRT-PCR with primers targeting control region or part of FGFR3 or eIF4E gene promoter. D. SW480 cell were transfected with si-NC or si-PRMT5 and protein expression of H4R3me2s and H3R8me2s was analyzed by Western blot 72 h later. E. SW480 cell were transfected with si-NC or si-PRMT5 and the expression of indicated proteins was analyzed by Western blot 72 h later. F. Proposed molecular mechanisms by which PRMT5 promotes CRC. PRMT5 overexpression promotes the activation of FGFR3, AKT, mTOR, ERK and eIF4E, leading to CRC cell growth, survival and migration. PRMT5 knockdown or inhibition by AMI-1 results in the downregulation of FGFR3, AKT, mTOR, ERK and eIF4E, leading to the inhibition of CRC cell growth, survival and migration.