RNase L downmodulation of the RNA-binding protein, HuR, and cellular growth

Abstract

Ribonuclease L (RNase L) is an intracellular enzyme that is vital in innate immunity, but also is a tumor suppressor candidate. Here, we show that overexpression of RNase L decreases cellular growth and downmodulates the RNA-binding protein, HuR, a regulator of cell-cycle progression and tumorigenesis. The effect is temporal, occurring in specific cell-cycle phases and correlated with the cytoplasmic localization of RNase L. Both cellular growth and HuR were increased in RNASEL-null mouse fibroblast lines when compared to wild-type cells. Moreover, the stability of HuR mRNA was enhanced in RNASEL-null cells. The HuR 3' untranslated region (UTR), which harbors U-rich and adenylate-uridylate-rich elements, was potently responsive to RNase L when compared to control 3' UTR. Our results may offer a new explanation to the tumor suppressor function of RNase L.

Figure 1. RNase L effect on cellular growth and cellular mRNAs

(A, B) Growth curves of RNase L-expressing and empty vector polyclonal cells. RNase L/neomycin or vector/neomycin cells were seeded in two different densities (5000 cells, A or 2500 cells, B) per well in 96-well plates and then monitored for growth by continuous real time RT-CES biosensor each 20 min for a period of five days. Data are from four replicates (shaded grey areas are SEM). Curve fit and doubling time was assessed using the exponential growth equation Y=Start*exp(K*X); Y=Start and increases exponentially with rate constant K; the doubling time equals 0.69/K. Curve comparison (K parameter difference) was used using F test (p values are indicated). (C) Microarray profiling of AU-rich mRNA expression as a result of RNase L and HuR action. Colored list shows the genes that are related in this study which are significant and statistically reproducible between control and either of RNase L-expressing or HuR-over-expressing (HeLa) lines. The color legend shows gradation of color proportional to the magnitude of induction (red) or repression (green) of expression. Inside image is a representative Western blot for the HuR over-expressing cell line. (D) Total RNA was extracted from confluent cells and used for RT-PCR at semi-quantitative conditions as described in Materials and Methods, with primers specific to HuR and β-actin mRNA.

Figure 2. Effect of RNase L over-expression and knockout on HuR and cellular growth

(A) Equal amount of proteins extracted from confluent cells that stably express neomycin vector or RNase L/neomycin were analyzed by Western Blotting using anti-HuR and anti-β-actin. Data shown are two representative experiments of three. Lower panel showed quantitation of β-actin normalized signals (Mean ±SEM) of HuR from three independent experiments. ** denotes p<0.01 using student t-test statistic. (B) Equal amount of proteins extracted from confluent MEFs generated from wild type or RNASEL-knockout mice were used for Western blotting (anti-mouse RNase L, anti-HuR and anti-β-actin). Right panel showed quantitation of β-actin normalized signals (Mean ± SEM) of HuR from three independent experiments. *** denotes p<0.001 using student t-test statistics. (D) Growth curve of RNASEL-knockout and wild type MEFs. Cells were seeded in a 96-well plate with 5000 cells (upper panel) or 2500 cells per well (lower panel). Cellular growth monitoring and statistics were performed as described in (Fig. 1A).

Figure 3. Effect of RNase L on HuR mRNA

(A) RNase L^-/-and RNase L^+/+ MEFs were treated with medium control (0 time point) or actinomycin D (5 ug/ml) for various times, as indicated. Semi-quantitative PCRs were performed with primers specific to HuR and β-actin cDNA. PCR, with cycle number that allows at least semi-quantitative comparison, was used as described Materials and Methods. Gels shown are from two independent experiments. (B) Signal intensities on gels from Fig. 3A were quantified. One-phase exponential decay curves for relative mRNA half-life measurements in the two cell types are shown. Density units (minus background) are plotted as a function of time after the addition of actinomycin D. Detail of the model is given in Materials and Methods.

Figure 4. Response of HuR 3′UTR-mediated reporter activity to RNase L

(A) The positions of U-rich/ARE-like regions and ARE in the 3′UTR of HuR mRNA reference record. B) Schematic diagram of reporter constructs fused with control EEF1A1 3′UTR (Construct 1), HuR 3′UTR U-rich/ARE like region (Construct 2), HuR ARE-containing 3′UTR region (Construct 3), and uPA 3′UTR (Construct 4). The shaded boxes show recognizable ARE of HuR 3′UTR and the known uPA ARE. (C) Cell lines stably expressing RNase L/neo and vector/neo in 96-well plates were transfected with the different 3′UTR constructs as indicated. Numbers in x-axis refer to the constructs as outlined in B. Readings are mean ± SEM of fluorescence intensities from quadruplicate wells. *** p value <0.001 with Bonferroni posttests and two-away ANOVA.

Figure 5. RNase L activity during cellular growth and confluence

(A) RNase L/neomycin or vector/neomycin stably expressing cells were grown into two different densities in 6-well plates, so that confluent (100% of total well area) and sub confluent (∼40%) monolayers were achieved the next day. Western blots were performed using anti-HuR and anti-β-actin. The β-actin normalized densitometry of blots (Mean ± SEM) from three independent experiments are shown (lower panel). *** denotes p<0.001 using student-t test. (B) Cells that stably express neomycin/vector control or RNase L/neomycin were seeded on cover slips. After 20 hrs, the cells were serum starved then treated with aphidocholin for additional 20 hrs to arrest cells in G1/S, i.e., late G1 (upper panels). Subsequently, cells were released from the arrest to enter S phase (lower panels) by removing the drug and sub-culturing in complete medium with 15% serum for 8 hours-These conditions were first optimized as shown in Supplementary Fig.3. The fluorescently-labeled secondary antibody was used to reveal anti-HuR antibody using confocal microscopy. Flow cytometry was performed by propidium iodide staining.

Figure 6. Nuclear/cytoplasmic distribution of RNase L

(A) Huh-7 cells, which constitutively express immunofluorescently detectable levels of RNase L, were seeded on cover slips with two different densities to allow cells to reach either sub-confluent (∼40%) or confluent stage the next day. Cells were stained with anti-RNase L or anti-HuR followed by secondary antibody that is either FITC-conjugated (green color, HuR) or TRITC-conjugated (red color, RNase L) for confocal visualization. (B) Huh-7 cells were seeded with two different densities to allow cells to reach either sub-confluent (∼40%) or confluent stage the next day. Nuclear and cytoplasmic extracts were subjected to Western blotting using antibodies to RNase L, HuR, and tubulin (cytoplasmic control) to confirm the findings in A. The blot is one of two (RNase L) and three (HuR) independent experiments.