Cross-kingdom inhibition of breast cancer growth by plant miR159

Abstract

MicroRNAs (miRNAs) are critical regulators of gene expression, and exert extensive impacts on development, physiology, and disease of eukaryotes. A high degree of parallelism is found in the molecular basis of miRNA biogenesis and action in plants and animals. Recent studies interestingly suggest a potential cross-kingdom action of plant-derived miRNAs, through dietary intake, in regulating mammalian gene expression. Although the source and scope of plant miRNAs detected in mammalian specimens remain controversial, these initial studies inspired us to determine whether plant miRNAs can be detected in Western human sera and whether these plant miRNAs are able to influence gene expression and cellular processes related to human diseases such as cancer. Here we found that Western donor sera contained the plant miRNA miR159, whose abundance in the serum was inversely correlated with breast cancer incidence and progression in patients. In human sera, miR159 was predominantly detected in the extracellular vesicles, and was resistant to sodium periodate oxidation suggesting the plant-originated 2'-O-methylation on the 3' terminal ribose. In breast cancer cells but not non-cancerous mammary epithelial cells, a synthetic mimic of miR159 was capable of inhibiting proliferation by targeting TCF7 that encodes a Wnt signaling transcription factor, leading to a decrease in MYC protein levels. Oral administration of miR159 mimic significantly suppressed the growth of xenograft breast tumors in mice. These results demonstrate for the first time that a plant miRNA can inhibit cancer growth in mammals.

Figure 1

miR159 is detectable in human serum and breast tumor tissue. (A) qPCR for miR159 in the serum of healthy human donors (n = 6) and breast cancer patients who did (n = 10) or did not (n = 20) progress to a higher stage of disease after chemotherapy. Data were first normalized to the level of human miR-16 and then plotted relative to the average level of miR159 in healthy human donors. (B) Enrichment of miR-16 and miR159 in EVs isolated from healthy human donor serum by ultracentrifugation (n = 3). (C) qPCR for the indicated miRNAs with (oxidized) or without (unoxidized) treatment with sodium periodate. RNAs were extracted from EVs isolated from human sera or from human sera depleted of EVs by ultracentrifugation. Data were normalized to miRNA levels of unoxidized samples (n = 3). (D, E) Representative ISH images of two human breast tumors (D) without or (E) with detectable miR159. (F) Scatter plot of ISH scores for miR159 expression in the tumor and the relative serum miR159 abundance shown in A. Scale bar, 100 μm. ^*P < 0.05. ND, not detectable.

Figure 2

miR159 decreases BC cell proliferation. (A-D) Cell proliferation of breast cell lines transfected with a miR159 mimic or scrambled control oligo. Cell proliferation was determined by cell counting (n = 3). (E) Cell counting of MDA-MB-231 cells transfected with the indicated dose of miR159. Cell were counted 4 days post-seeding (n = 3). (F) Cell counting of MDA-MB-231 cells transfected with anti-miR159 LNAs or control LNAs and treated with human serum EVs. Cells were treated with EVs isolated from 4 ml of pooled human sera or PBS twice over 4 days (n = 3). ^*P < 0.05.

Figure 3

miR159 directly targets human TCF7. (A) Overlap of putative miR159 targets in humans identified by RNA-Seq of MDA-MB-231 cells transfected with miR159 mimic, Diana microT v3.0 target prediction software, and the RISCTRAP miR159 association assay. (B) Fold change and statistical significance of the three target genes identified by all three methods. (C-D) qPCR validations of the regulation of the mRNA levels of putative miR159 targets by miR159 in MDA-MB-231 cells (C) and the association of miR159 with TCF7 mRNA in RISCTRAP assay (D). (E) The two putative miR159-binding sites in TCF7 3′UTR identified by Diana microT v3.0. (F) Schematic of the TCF7 miR159-binding site reporter constructs. (G) Dual-luciferase assay in MD-MB-231 cells co-transfected with miR159 or scrambled control oligos and the reporter constructs from F. (H-J) Dual-luciferase assays in the indicated breast cells transfected with miR159 or scrambled control oligos for TOP/FOP (H), E-box (I) or NFκB reporter (J) activity. In C, D, G, and H-J, data shown represent mean ± SD (n = 3). (K) Western blot assays to evaluate TCF7 or MYC expression in breast cell lines transfected with miR159 or scrambled control oligos. Band intensities were measured using ImageJ and were normalized to ACTIN. ^*P < 0.05. ND, not detectable.

Figure 4

miR159 inhibits cell proliferation through targeting TCF7. (A, C) Cell counting of MDA-MB-231 cells co-transfected with miR159 and TCF7HA (A) or MYC (C). Cells were counted 4 days post-seeding. (B) Dual-luciferase assay for TCF reporter activity. MDA-MB-231 cells were co-transfected with miR159, TCF7HA and either TOPFLASH or FOPFLASH plasmids. (D) Dual-luciferase assay for E-box reporter activity. MDA-MB-231 cells were co-transfected with miR159, MYC and either E-box reporter or vector plasmids. In B and D, the luciferase assay was performed 3 days after transfection. (E) Working Model for miR159 inhibition of breast cancer growth. Data shown represent mean ± SD (n = 3). ^*P < 0.05.

Figure 5

miR159 level is inversely correlated with levels of TCF7 and MYC in breast tumors. (A, B) Selected images showing ISH for miR159 and IHC for TCF7 and MYC in two BC samples with (A) or without (B) detectable miR159 from a human breast cancer array. (C) Statistical analysis of correlation between miR159 and TCF7 or MYC levels from the human breast cancer array BC08118. (D) Histogram of miR159 scores from the human breast cancer array. Scale bar, 50 μm.

Figure 6

miR159 decreases tumor growth through targeting TCF7 in vivo. (A) MDA-MB-231 tumor growth measured by calipers in mice gavage fed with miR159 or scrambled control oligos. (B) Final tumor weight of MDA-MB-231 tumors. (C, D) qPCR analysis of TCF7 (C) and MYC (D) expression in MDA-MB-231 tumors. (E, F) Quantification of IHC staining signals for Ki67 (E) and cleaved caspase-3 (F) in MDA-MB-231 tumor slides. (G) Representative IHC/ISH images for Ki67, cleaved caspase-3, and miR159 in MDA-MB-231 tumors. In A-G, MDA-MB-231 xenograft mice were fed with miR159 (n = 13) or scrambled control oligos (n = 11). (H) MDA-MB-231-TCF7HA tumor growth measured by calipers in mice gavage fed with miR159 or scrambled control oligos. (I) Final tumor weight of MDA-MB-231-TCF7HA tumors. (J, K) qPCR analysis of TCF7 (J) and MYC (K) expression in MDA-MB-231-TCF7HA tumors. (L, M) Quantification of IHC staining signals for Ki67 (L) and cleaved caspase-3 (M) in MDA-MB-231-TCF7HA tumor slides. (N) Representative IHC/ISH images for Ki67, cleaved caspase-3, and miR159 in MDA-MB-231-TCF7HA tumors. In H-N, MDA-MB-231-TCF7HA xenograft mice were fed with miR159 (n = 12) or scrambled control oligos (n = 12). Scale bar, 100 μm. ^*P < 0.05.