hUPF2 silencing identifies physiologic substrates of mammalian nonsense-mediated mRNA decay

Abstract

Nonsense-mediated mRNA decay (NMD) is a conserved eukaryotic surveillance pathway that selectively degrades aberrant mRNAs with premature termination codons (PTCs). Although a small number of cases exist in mammals, where NMD controls levels of physiologic PTC transcripts, it is still unclear whether the engagement of NMD in posttranscriptional control of gene expression is a more prevalent phenomenon. To identify physiologic NMD substrates and to study how NMD silencing affects the overall dynamics of a cell, we stably down-regulated hUPF2, the human homolog of the yeast NMD factor UPF2, by RNA interference. As expected, hUPF2-silenced HeLa cells were impaired in their ability to recognize ectopically expressed aberrant PTC transcripts. Surprisingly, hUPF2 silencing did not affect cell growth and viability but clearly diminished phosphorylation of hUPF1, suggesting a role of hUPF2 in modulating NMD activity through phosphorylation of hUPF1. Genome-wide DNA microarray expression profiling identified 37 novel up-regulated and 57 down-regulated transcripts in hUPF2-silenced cells. About 60% of the up-regulated mRNAs carry typical NMD motifs. Hence, NMD is important not only for maintaining the transcriptome integrity by removing nonfunctional and aberrant PTC-bearing transcripts but also for posttranscriptional control of selected physiologic transcripts with NMD features.

FIG. 1.

Stable silencing of hUPF2 in HeLa cells. (A) Schematic presentation of the pSUPER-hUPF2 vector. A DNA fragment consisting of the 19-nt hUPF2 sequence followed by a 9-nt spacer and the 19-nt reverse hUPF2 sequence was cloned downstream of the H1 RNA promoter. The predicted structure of the hUPF2 shRNA is shown above with the 19-nt hUPF2 sequence underlined. (B) Western blot analysis of hUPF2 in HeLa (WT) cells and a clone stably transfected with an hUPF2 shRNA-encoding plasmid (H2KD.2). Cellular proteins were reduced, separated by 8% SDS-PAGE, and transferred to a nitrocellulose membrane, which was first stained with polyclonal rabbit anti-hUPF2 antibodies (hUPF2) and developed with an appropriate HRP-conjugated secondary antibody and the ECL method. Loading of equal cell equivalents was assessed with rabbit antibodies against β-actin. Molecular masses of standard proteins are indicated in kDa on the left. (C) Western blot analysis of hUPF2 in serially diluted cell lysates from HeLa (WT) and H2KD.2 cells. Analysis was performed as described in panel B with polyclonal rabbit antibodies against hUPF2 and β-actin. The lanes represent threefold serial dilutions of protein. (D) RT-PCR analysis of hUPF2 mRNA in WT and H2KD.2 cells. PCR products were separated on a 1.5% agarose gel and stained with ethidium bromide. Loading of same cell equivalents was assessed by analyzing β-actin mRNA levels (lower panel). Results represent one of two independent experiments. (E) Analysis of hUPF2 mRNA in WT and H2KD.2 cells by qRT-PCR. Relative hUPF2 mRNA levels were determined and normalized to β-actin mRNA levels, which served as an internal control. The amount of hUPF2 mRNA in WT cells was set to 100%. Average values and standard deviations of three independent qRT-PCR runs (triplicates for each) are shown.

FIG. 2.

Analysis of ectopic nonsense mRNA levels in hUPF2-silenced cells. WT and H2KD.2 cells were transfected with plasmids encoding either sense or PTC-containing nonsense β-globin (A) or μHC transcripts (B) and, in order to control for transfection efficiency, with a plasmid encoding κLC. Total RNA was isolated 48 h after transfection, separated on a 1.2% agarose-formaldehyde gel, and subjected to Northern blot analysis. Fifty percent more RNA from cells transfected with nonsense vectors was loaded. Radioactive signals were quantified by phosphorimaging. Relative transcript levels were calculated by dividing β-globin and μHC signals by corresponding κLC signals; relative intensities of sense transcripts in WT cells were arbitrarily set to 100%. Mean values and standard deviations (SD) were calculated from three independent experiments.

FIG. 3.

Analysis of hUPF protein levels in hUPF2-silenced cells. Cellular lysates from 4.5 × 10⁵, 1.5 × 10⁵, and 0.5 × 10⁵ WT or H2KD.2 cells were reduced, separated by 8% SDS-PAGE, and transferred to a nitrocellulose membrane. The membrane was stained stepwise with anti-hUPF2, anti-hUPF1, anti-hUPF3a, and anti-hUPF3b antibodies and developed with an appropriate HRP-conjugated secondary antibody and the ECL method. Loading of same cell equivalents was assessed with rabbit antibodies against β-actin. Results correspond to one of two independent experiments. The lanes represent threefold serial dilutions of protein and demonstrate that the Western blotting is semiquantitative. Molecular masses of standard proteins are indicated in kDa on the left. The Western blots for hUPF2 and β-actin are identical to the ones shown in Fig. 1C and have been included in Fig. 3 for clarity.

FIG. 4.

Analysis of hUPF1 phosphorylation in hUPF2- and hSMG-6-silenced cells. (A) Detection of P-(S/T) ATM/ATR motifs in hUPF1 by a combined immunoprecipitation (IP) and Western blot (WB) analysis. hUPF1 was precipitated from cleared lysates of 5 × 10⁶ WT or H2KD.2 cells with specific rabbit anti-hUPF1 antibodies (IP: anti-hUPF1), electrophoretically separated, and transferred to a nitrocellulose membrane. The membrane was first developed with P-(S/T) ATM/ATR substrate-specific antibodies (WB, left blot) and, after stripping, with anti-hUPF1 antibodies (right blot). (B) Effect of hUPF1 down-regulation on the phosphorylation status of (S/T) ATM/ATR substrates. Cellular proteins from 4 × 10⁵, 2 × 10⁵, and 0.5 × 10⁵ HeLa cells transfected without siRNA (no), with hUPF1 siRNA (hUPF1), and with lamin A/C siRNA (lamin A/C) were electrophoretically separated and analyzed by stepwise immunoblotting with antibodies against P-(S/T) ATM/ATR substrates (P-hUPF1), hUPF1, and hUPF2. Equal protein loading was assessed with rabbit antibodies against β-actin. The lanes represent two- and fourfold serial dilutions of protein and demonstrate that the Western blotting is semiquantitative. Molecular masses of standard proteins are indicated in kDa on the left. (C) Effect of hSMG-6 down-regulation on the phosphorylation status of hUPF1. Cellular proteins from 2 × 10⁵, 1 × 10⁵, and 0.5 × 10⁵ HeLa cells transiently transfected with the empty pSUPER-puro vector (no) or pSUPER-puro encoding hSMG-6-specific shRNAs (hSMG-6) were electrophoretically separated and analyzed by stepwise immunoblotting with antibodies against P-(S/T) ATM/ATR substrates (P-hUPF1) and hUPF2. BiP stainings served as a control for protein loading. The lanes represent twofold serial dilutions of total lysates. Cellular lysates in panels A, B, and C were reduced and separated on 8% SDS-PAA gels, and signals were developed with appropriate HRP-conjugated secondary antibodies and the ECL method. Molecular masses of standard proteins and their positions are indicated in kDa on the left, and the positions of the respective protein bands are indicated on the right of the blots. Results represent one of at least two independent experiments.

FIG. 5.

Effect of hUPF2 down-regulation on cell growth and cell cycle progression. (A) Growth rates of WT (circles) and H2KD cells (boxes). Cells (1 × 10⁵) were plated in triplicate in three six-well plates and cultured at 37°C for 3 days. Viable cells were counted every 24 h using a hemocytometer after being stained with trypan blue. Measurements were performed in triplicate in a series of three independent experiments. Standard deviations are indicated. (B) Cell cycle analysis of WT (a) and H2KD cells (b). Cells were harvested, permeabilized, and stained with PI. PI fluorescence was measured by flow cytometric analysis and analyzed with the ModFit LT software. The bar diagrams on the left illustrate the mean distributions from at least seven independent experiments of WT and H2KD cells in the sub-G₁ (gate G in the right diagram), the G₁ (gate D), the S (gate E), and the G₂/M (gate F) phases of the cell cycle; standard deviations are indicated. The diagrams on the right with either a linear (Lin) or logarithmic representation of the PI fluorescence show the distribution of WT and H2KD cells in the individual phases of the cell cycle. Gates were set to distinguish the different cell populations according to their DNA content.

FIG. 6.

Verification of DNA microarray results by semiquantitative and quantitative real-time RT-PCR analyses. Total RNA from WT and hUPF2-silenced H2KD.2 cells was analyzed by RT-PCR and qRT-PCR for the presence of selected transcripts (indicated on the right of each gel picture) that were either (A) down-regulated or (B) up-regulated in the DNA microarray analyses. RT-PCR transcripts were detected after agarose gel electrophoresis by ethidium bromide staining. β-Actin signals served to assess the loading of equal amounts of cDNA. Results represent one of two independent experiments. Differences in mRNA abundance (n-fold) obtained in microarray (array) and qRT-PCR analyses are indicated on the right side of the gel pictures. qRT-PCR values were normalized to β-actin mRNA levels and represent the means from at least three independent experiments, which were run in triplicate. To demonstrate a linear relationship between the amounts of input cDNA and RT-PCR products, serial dilutions of cDNA were analyzed by semiquantitative RT-PCR in panel B. (C) Schematic depiction of mRNA transcripts with possible NMD-causing features.

FIG. 7.

Analysis of alternatively spliced TfR2 isoforms for the presence of premature termination codons. (A) Total RNA from WT and H2KD cells was subjected to RT-PCR analysis with primers specific for exons 6 and 18 of TfR2. PCR products were separated on a 1.0% agarose gel and detected with ethidium bromide. Size standards of 100 bp and 1 kb are shown on the left. RT-PCR for β-actin served as a control for equal loading of cDNA. (B) Detection of alternative TfR2 splice isoforms by sequence analysis of cloned bands A to F in panel A. RT-PCR products from WT and H2KD cDNAs (bands A to F) from the gel shown in panel A were excised, purified, and cloned. Restriction digest of recombinant plasmids revealed different insert sizes. Several sequences were obtained from each cloned band. An overview of the observed TfR2 splicing patterns is schematically shown. Exon sizes are shown at the bottom, while the presence (+) or absence (−) of a PTC and the size of the PCR product are indicated on the right. Retained introns (the B2, D1, D2, D3, and D5 transcripts) are shown as gray boxes. The D4, D5, E1, and E4 transcripts use alternative splice acceptor sites, while the E1 transcript also uses an alternative splice donor site. Note that exons and introns are not drawn to scale. (C) Detection of alternative TfR2 splice products in band A. Serial dilutions of excised band A from WT and H2KD cells were subjected to PCR analysis with TfR2 primers specific for exons 15 and 16. PCR products were separated on a 1.5% agarose gel and visualized by ethidium bromide staining. A 100-bp size standard is shown on the left, while an overview of the expected TfR2 splicing patterns is schematically shown on the right.

FIG. 8.

Comparative qRT-PCR analysis of selected mRNA levels in transiently hUPF1- and hUPF2-silenced HeLa cells. (A) Western blot analysis. Threefold serially diluted cellular lysates starting from 4.5 × 10⁵ HeLa cells transiently transfected with hUPF2 siRNA, hUPF1 siRNA, and lamin A/C siRNA were analyzed as described in the legend to Fig. 4B by stepwise immunoblotting the same membrane with antibodies against hUPF2, hUPF1, and lamin A/C. Equal protein loading was assessed with antibodies against BiP. Molecular masses of standard proteins are indicated in kDa on the left, and the positions of the respective protein bands are indicated on the right of the blots. Results represent one of at least two independent experiments. (B) Quantitative RT-PCR analysis. qRT-PCR values were normalized to β-actin mRNA levels. Relative amounts of respective mRNAs in lamin A/C siRNA-transfected HeLa cells were set to 1, and mRNA changes (n-fold) were determined by dividing relative signals of the respective gene in hUPF1- and hUPF2-silenced cells by that in lamin A/C-silenced cells. Average mRNA changes (n-fold) and standard deviations of at least four independent qRT-PCR assays (triplicates for each) are shown.