Alu RNA Modulates the Expression of Cell Cycle Genes in Human Fibroblasts

Abstract

Alu retroelements, whose retrotransposition requires prior transcription by RNA polymerase III to generate Alu RNAs, represent the most numerous non-coding RNA (ncRNA) gene family in the human genome. Alu transcription is generally kept to extremely low levels by tight epigenetic silencing, but it has been reported to increase under different types of cell perturbation, such as viral infection and cancer. Alu RNAs, being able to act as gene expression modulators, may be directly involved in the mechanisms determining cellular behavior in such perturbed states. To directly address the regulatory potential of Alu RNAs, we generated IMR90 fibroblasts and HeLa cell lines stably overexpressing two slightly different Alu RNAs, and analyzed genome-wide the expression changes of protein-coding genes through RNA-sequencing. Among the genes that were upregulated or downregulated in response to Alu overexpression in IMR90, but not in HeLa cells, we found a highly significant enrichment of pathways involved in cell cycle progression and mitotic entry. Accordingly, Alu overexpression was found to promote transition from G1 to S phase, as revealed by flow cytometry. Therefore, increased Alu RNA may contribute to sustained cell proliferation, which is an important factor of cancer development and progression.

Keywords: Alu retrotransposons; cell cycle; non-coding RNA.

Figure 1

Sequence features of a consensus Alu element. The left arm harbors the internal Pol III promoter, composed of the A Box and B Box. The right arm has a 31-bp insertion. The two arms are separated by an A-rich region and the entire Alu element ends with a poly(A) tail. Grey indicates the genomic repeats that originate from the retrotransposition event and the 3′ trailer sequence between the poly(A) tail and the canonical terminator (at least four Ts).

Figure 2

Schematic representation of the DNA inserted by the lentivirus vector used to generate IMR90 and HeLa cells that stably overexpress Alu sequences. The Pol III H1 promoter (H1p) controls the expression of AluSq2, AluSx, or a Control RNA. An empty vector was used as negative control. The eGFP gene and the gene coding for puromycin resistance are inserted under the control of the PGK gene promoter (PGKp).

Figure 3

Alu overexpression from lentivirus vector stably inserted in the genome. The bar plots report the fold increase of AluSq2 and AluSx expression in either (left panel) IMR90 or (right panel) HeLa cells that were stably transformed with a lentivirus vector carrying the corresponding Alu. The fold increase was derived from comparisons with cells transformed with the empty vector (background Alu expression). Alu RNA levels were quantified by RT-qPCR analysis, conducted with primers chosen to target unique sequence tracts within the Alu 3′ trailer region. In all measurements, gene expression levels were normalized to U1 gene expression, used as an internal standard.

Figure 4

Differentially expressed genes in Alu/Control RNA-overexpressing cells. (a) Volcano plots showing the distribution of the differentially expressed genes in human fibroblasts overexpressing AluSq2, AluSx, or the Control RNA. Blue spots represent downregulated genes, red spots represent upregulated genes. (b) Venn diagram showing the intersection of genes that are differentially expressed in AluSq2, AluSx, and Control RNA-overexpressing cells.

Figure 5

Dysregulation of cell cycle genes in Alu-overexpressing fibroblasts. (a) Pathways enriched in differential transcriptomes of AluSq2, AluSx, and Control RNA transformed fibroblasts as revealed by IPA analysis. Only pathways with p-value (BH correction) <0.01 in at least one comparison are shown and are ordered by AluSx vs. empty vector p-value. Z-score for activated or inhibited pathways is shown in red or blue, respectively. Grey bars: no predictions can be made. (b) Prediction of the upstream regulators that could modulate the expression of differentially expressed genes in AluSx-overexpressing fibroblasts and their effect on cell functions. Dysregulated genes are shown in the middle row as upregulated genes (red symbols) and downregulated genes (green symbols). Regulators are shown in the upper part of the figure. Blue indicates a predicted inhibition of the protein activity, while orange indicates a predicted activation. Blue lines indicate an inhibitory relationship, orange lines show an activating relationship, yellow lines indicate inconsistent relationship, while gray lines stand for no predicted effect. Continuous lines show direct interactions and dashed lines show indirect interactions (less than three passages). (c) Heatmap of differentially expressed genes shown in (b).

Figure 5

Dysregulation of cell cycle genes in Alu-overexpressing fibroblasts. (a) Pathways enriched in differential transcriptomes of AluSq2, AluSx, and Control RNA transformed fibroblasts as revealed by IPA analysis. Only pathways with p-value (BH correction) <0.01 in at least one comparison are shown and are ordered by AluSx vs. empty vector p-value. Z-score for activated or inhibited pathways is shown in red or blue, respectively. Grey bars: no predictions can be made. (b) Prediction of the upstream regulators that could modulate the expression of differentially expressed genes in AluSx-overexpressing fibroblasts and their effect on cell functions. Dysregulated genes are shown in the middle row as upregulated genes (red symbols) and downregulated genes (green symbols). Regulators are shown in the upper part of the figure. Blue indicates a predicted inhibition of the protein activity, while orange indicates a predicted activation. Blue lines indicate an inhibitory relationship, orange lines show an activating relationship, yellow lines indicate inconsistent relationship, while gray lines stand for no predicted effect. Continuous lines show direct interactions and dashed lines show indirect interactions (less than three passages). (c) Heatmap of differentially expressed genes shown in (b).

Figure 6

AluSq2 and AluSx stimulate cell cycle progression of IMR90 cells. (a) FACS analysis of synchronized IMR90 cells after 24 h of serum re-feeding. ModFit LT analyses revealed an accumulation of AluSq2- and AluSx-overexpressing IMR90 cells in S phase. (b) Bar plot data are derived from at least four independent experiments obtained from two IMR90 cell cultures transformed with a lentivirus vector. * p < 0.05, ** p < 0.01, *** p < 0.0 01 compared to the Control.