Depletion of METTL3 alters cellular and extracellular levels of miRNAs containing m6A consensus sequences

Abstract

Extracellular vesicles (EVs) are capable of transferring cargo from donor to recipient cells, but precisely how cargo content is regulated for export is mostly unknown. For miRNA cargo, we previously showed that when compared to isogenic colorectal cancer (CRC) cells expressing wild-type KRAS, a distinct subset of miRNAs are differentially enriched in EVs from KRAS mutant active CRC cells, with miR-100 being one of the most enriched. The mechanisms that could explain how miR-100 and other miRNAs are differentially exported into EVs have not been fully elucidated. Here, we tested the effect of N⁶-methyladenosine (m⁶A) modification on miRNA export into EVs by depletion of METTL3 and ALKBH5, a writer and eraser of m⁶A modification, respectively. While the effects of ALKBH5 knockdown were quite modest, decreased levels of METTL3 led to reduced cellular and extracellular levels of a subset of miRNAs that contain consensus sequences for m⁶A modification. Functional testing of EVs prepared from cells expressing shRNAs against METTL3 showed that they were less capable of conferring colony growth in 3D to wild-type KRAS cells and were also largely incapable of conferring the spread of cetuximab resistance. Our data support a role for METTL3 modification on cellular miRNA levels and export of specific miRNAs.

Keywords: Base modification; Extracellular vesicle; RNA; Tumor microenvironment; m6A; miRNA.

Figure 1

Reduced extracellular transfer of miR-100 after knockdown of readers, writers and erasers of RNA modification. Stable shRNA knockdown cell lines were created using mutant KRAS (DKO-1) cells transfected with either an empty shRNA vector, a scrambled control shRNA vector, or vectors encoding shRNAs targeting 14 different proteins (see Table S1). A) Schematic of luciferase reporter assay. B) Luciferase reporter assay. Wild-type KRAS DKs-8 recipient cells were seeded in the bottom of Transwell dishes and co-transfected with vectors expressing β-galactosidase and a luciferase reporter containing either a control 3′UTR or a modified 3′UTR with three perfect miR-100 binding sites. Recipient cells were co-cultured the indicated donor cells for 24 h before cell lysates were collected. Luciferase expression was quantified with decreased expression indicating increased transfer of miR-100. Significance was determined by one-way ANOVA. Data represent mean ± SE, n = 3. ∗p < 0.05, ∗∗p < 0.01, ∗∗∗p = 0.0002, ∗∗∗∗p < 0.0001.

Figure 2

Decreased levels of METTL3 and ALKBH5 alters cellular and EV small RNA levels. A) Western blots were performed on cell lysates from stable lines expressing an empty shRNA vector (Empty), a scrambled control shRNA vector, or two independent shRNAs targeting METTL3 or ALKBH5. Representative immunoblots using antibodies against METTL3, ALKBH5, Tubulin, and GAPDH are shown with the average decrease in METTL3 and ALKBH5 being 60.4 ± 1.1% and 68.4 ± 1.04%, respectively. B) qRT/PCR analysis of miR-100 levels in cellular (black) and EV (gray) samples. Significance was determined by one-way ANOVA. Data represent mean ± SE, n = 3, and ∗p < 0.05.

Figure 3

Knockdown of METTL3 reduces EV-mediated anchorage-independent growth. A) Gene ontology analysis of significant biological processes affected in METTL3 knockdown cells compared to DKO-1 control cells after total RNAseq. B) Immunoblots of EVs from DKO-1 or METTL3 knockdown cells using antibodies against CD63, TSG101, CD81, and H3. C,D) DKO-1 and METTL3 knockdown cells (500 cells/well) were grown in soft agar for 2 weeks in the presence or absence of EVs derived from either DKO-1 cells or METTL3 knockdown cells. Representative images (C) and quantification of colony counts (D) are shown. Significance was determined by one-way ANOVA. Equivalent amounts of EVs were added. Data represent mean ± SE, n = 3, ∗p < 0.05.

Figure 4

Knockdown of METTL3 reduces EV-mediated transfer of cetuximab resistance. Wild-type KRAS DKs-8 cells (∼2000 cells/well) were grown in soft agar for 2 weeks in the presence or absence of EVs derived from either control DKO-1 cells or METTL3 knockdown cells. Representative images (A) and quantification of colony counts (B) are shown. Significance was determined by one-way ANOVA. Equivalent amounts of EVs were added. Data represent mean ± SE, n = 3, ∗∗p < 0.01, ∗∗∗∗p < 0.0001.

Fig S5