EPO Mediates Neurotrophic, Neuroprotective, Anti-Oxidant, and Anti-Apoptotic Effects via Downregulation of miR-451 and miR-885-5p in SH-SY5Y Neuron-Like Cells

Abstract

Erythropoietin (EPO) is a neuroprotective cytokine, which has been applied in several animal models presenting neurological disorders. One of the proposed modes of action resulting in neuroprotection is post-transcriptional gene expression regulation. This directly brings to mind microRNAs (miRNAs), which are small non-coding RNAs that regulate gene expression at the post-transcriptional level. It has not yet been evaluated whether miRNAs participate in the biological effects of EPO or whether it, inversely, modulates specific miRNAs in neuronal cells. In this study, we employed miRNA and mRNA arrays to identify how EPO exerts its biological function. Notably, miR-451 and miR-885-5p are downregulated in EPO-treated SH-SY5Y neuronal-like cells. Accordingly, target prediction and transcriptome analysis of cells treated with EPO revealed an alteration of the expression of genes involved in apoptosis, cell survival, proliferation, and migration. Low expression of miRNAs in SH-SY5Y was correlated with high expression of their target genes, vascular endothelial growth factor A, matrix metallo peptidase 9 (MMP9), cyclin-dependent kinase 2 (CDK2), erythropoietin receptor, Mini chromosome maintenance complex 5 (MCM5), B-cell lymphoma 2 (BCL2), and Galanin (GAL). Cell viability, apoptosis, proliferation, and migration assays were carried out for functional analysis after transfection with miRNA mimics, which inhibited some biological actions of EPO such as neuroprotection, anti-oxidation, anti-apoptosis, and migratory effects. In this study, we report for the first time that EPO downregulates the expression of miRNAs (miR-451 and miR-885-5p) in SH-SY5Y neuronal-like cells. The correlation between the over-expression of miRNAs and the decrease in EPO-mediated biological effects suggests that miR-451 and miR-885-5p may play a key role in the mediation of biological function.

Keywords: apoptosis; cell death; erythropoietin; miRNAs; migration; neurite outgrowth; oxidative stress; proliferation.

Figure 1

Erythropoietin decreases miR-451 and miR-885-5p expression in SH-SY5Y cells. Cells were treated with 1 U/ml EPO during 24 and 48 h. Expression levels of miR-451 (A) and miR-885-5p (B) in both treated and non-treated cells were quantified by qRT-PCR. EPO; erythropoietin. The data are presented as mean ± SE, n = 5 (*p < 0.05).

Figure 2

Mir-451 comes from a link RNA (not shown in graph). Mir-451 has 10 known targets (green). These targets are part of a number of reactome pathways (blue). The number in the blue nodes can be used to access the associated reactome pathway, and Table S3 in Supplementary Material provides further information for clarity. Size of blue nodes reflects the number of mir-885 targets it contains. Edges ending in circles (gray) indicate a part of relationship (e.g., FBP1 is a part of the “glycogen storage diseases” pathway, which has the reactome accession number 3229121). Edges ending in straight lines (red) indicate inhibition of the target (e.g., mir-451 inhibits AKT1).

Figure 3

Mir-885 comes from an intron in ATP2B2 (both red). Mir-885 has two known targets (green). These targets are part of a number of reactome pathways (blue). The Number in the blue nodes can be used to access the associated reactome pathway and Table S2 in Supplementary Material provides further information for clarity. Size of blue nodes reflects the number of mir-885 targets it contains. Edges ending in circles (gray) indicate a part of relationship (e.g., CDK2 is a part of the “Activation of the pre-replicative complex” pathway, which has the reactome accession number 68962). Edges ending in straight lines (red) indicate inhibition of the target (e.g., mir-885 inhibits MCM5).

Figure 4

qRT-PCR array validation of the microarray data for the selected genes. The RT-PCR data were normalized to the housekeeping gene β-actin. Pearson’s coefficient, r = 0.95, p = 0.000.

Figure 5

Time-dependent CDK2 gene expression changes following EPO treatment in SH-SY5Y cells. Cells were treated with 1 U/ml EPO during 0–24 h time intervals. Expression levels of CDK2 were quantified by qRT-PCR. CDK2; cell cycle dependent kinase 2. The data are presented as mean ± SE, n = 5 (*p < 0.05).

Figure 6

miR-451 and miR-885-5p significantly downregulate expression of target genes in SH-SY5Y cells. qRT-PCR was performed to determine the mRNA levels of target genes after transfection of SH-SY5Y cells with miR-451, miR-885-5p mimics, and negative control (NC) miRNAs without EPO treatment (A), with 24 h EPO treatment (B) and 48 h EPO treatment (C). Control: cells without any transfection. The data are presented as mean ± SE, n = 5 (*p < 0.05).

Figure 7

miR-451 and miR-885-5p may mediate the neuroprotective effect of EPO. Transfected and non-transfected cells were treated with EPO and/or CoCl₂. (A) Cell viability was analyzed by WST-8 assay, (B) representative phase-contrast microcopy images of SH-SY5Y cells, and (C) representative light microscopy images showing trypan blue stained dead cells. Over-expression of miR-451 and miR-885-5p reverse the neuroprotective effect of EPO induced by CoCl₂. CoCl₂; cobalt chloride, EPO; erythropoietin. The data are presented as mean ± SE, n = 5 (*p < 0.05 compared to control and ^#p < 0.05 compared to CoCl₂ treated cells).

Figure 8

Anti-apoptotic and anti-oxidant effects of EPO in SH-SY5Y cells may be regulated by miR-451 and miR-885-5p. Transfected and non-transfected cells were treated by staurosporine (STS) (25 nM) and 1 U/ml EPO. (A) DNA fragmentation was evaluated by cell death ELISA assay. (B) Apoptotic phosphatidylserine (PS) positive cells were stained by Annexin-V-FITC dye and visualized using immunofluorescence microscopy. Cells were also treated with CoCl₂ (250 μM) and 1 U/ml EPO (C) and DNA fragmentation was analyzed by cell death ELISA assay. Transfection of miR-451 and miR-885-5p mimics reduced the anti-apoptotic effect of EPO in both STS and CoCl₂ induced apoptosis. (D) Intracellular ROS production was quantified by using the CM-H₂DCFDA method. Over-expression of miR-451 and miR-885-5p reversed anti-oxidant effect of EPO induced by CoCl_2. STS; staurosporin, EPO; erythropoietin, CM-H₂DCFDA; 2′,7′-dichlorodihydrofluorescein, acetyl ester. The data are presented as mean ± SE, n = 5 [*p < 0.05 compared to control and ^#p < 0.05 compared to toxic agents (STS or CoCl₂) treated cells].

Figure 9

Upregulation of miR-451 and miR-885-5p reduces proliferation of SH-SY5Y cells enhanced by EPO. In miRNA mimic transfected and non-transfected cells, cell proliferation was observed by counting cells using phase-contrast microscopy following either 24 or 48 h EPO treatment. (A) Cell counts and (B) phase-contrast microscopy images. Transfection of miR-451 and miR-885-5p mimics reduces the proliferative effect of EPO. EPO; erythropoietin. The data are presented as mean ± SE, n = 5 (*p < 0.05).

Figure 10

EPO induces migration in SH-SY5Y cells, which can be reversed by mir-451 and mir-885-5p. Transfected and non-transfected cells were treated with 1 U/ml EPO and migration assay was conducted for either condition. (A) Cells were counted by using ImageJ 1.42 (http://imagej.nih.gov/ij/). (B) Representative images of migrated SH-SY5Y cells. Over-expression of miR-451 and miR-885-5p reverse the migratory effect of EPO. EPO, erythropoietin. The data are presented as mean ± SE, n = 3 (*p < 0.05).

Figure 11

EPO may enhance neurite outgrowth and length via regulation of miR-451 and miR-885-5p. Transfected and non-transfected cells were treated with 1 U/ml EPO. (A) Neurite cell number was counted using ImageJ 1.42 (http://imagej.nih.gov/ij/). Randomly selected fields in microscopic analysis with an average of 35 cells per field were analyzed. (B) Representative phase-contrast microscopy images are shown in order to indicate the increased neurite numbers induced by EPO and reversed by the transfection of mimics of miR-451 and miR-885-5p. (C) The average maximal neurite length for both, the transfected and non-transfected cells were analyzed by ImageJ 1.42 (http://imagej.nih.gov/ij/). At least 30 neurites per condition were measured. (D) Neurite length distribution. Arrowheads indicate the neurites arising from cells. Scale bar, 100 μm. EPO, erythropoietin. The data are presented as mean ± SE, n = 5 (*p < 0.05).