Son Is Essential for Nuclear Speckle Organization and Cell Cycle Progression

cDNA Constructs

Son cDNA sequence (KIAA1019) was obtained from the HUGE consortium, Japan (Saitoh et al., 2004). The original KIAA1019 (Saitoh et al., 2004) was truncated at the 5′ end but is complete in the updated KIAA1019 cDNA now available from the HUGE consortium. Because the 3′ end of the updated KIAA1019 cDNA does not encode the G-patch or double-stranded RNA-binding domain (DS RBD) motifs that are present in cellular Son transcripts, we substituted the correct 3′ end from a separate partial Son cDNA clone (clone fh21047 obtained from T. Nagase, HUGE consortium). The full-length Son cDNA open reading frame was then subcloned into pEYFP-C1 (Takara, Kyoto, Japan). To produce a yellow fluorescent protein (YFP)-tagged small interfering RNA (siRNA)-refractory Son mRNA (YFP-siR-Son), the open reading frame near the 5′ end of YFP-Son cDNA that corresponds to the nucleotides targeted by Son siRNA 4 was changed at seven nucleotide positions by site-directed mutagenesis using forward oligonucleotide (oligo) 5′-gccagttgtaacCatgAGCgtCgaAtaCcagatgaagtctg-3′ and reverse oligo 5′-cagacttcatctgGtaTtcGacGCTcatGgttacaactggc-3′ (the siRNA-targeted region is underlined and all mutagenized nucleotides are indicated in uppercase letters). Convenient restriction sites were used for subcloning Son deletion mutants into pEYFP-C1. The sequence of all polymerase chain reaction (PCR) products was confirmed by DNA sequencing. YFP-Magoh and YFP-MLN51 cDNAs were a kind gift from Catherine Tomasetto (Institute of Genetics and Molecular and Cellular Biology, Strasbourg, France).

Cell Culture

HeLa cells and HeLa S3 cells were maintained in DMEM supplemented with penicillin/streptomycin and 10% fetal bovine serum. HeLa cells stably expressing YFP-SF2/ASF and SC35-YFP were described previously (Prasanth et al., 2003; Bubulya et al., 2004). For transcription inhibition, α-amanitin (50 μg/ml) was added for 6 h before harvesting cells.

Preparation of the Nuclear Insoluble Fraction

S3 cells (5.0 × 10⁶) were incubated at 4°C for 3 min in cytoskeleton (CSK) buffer [10 mM piperazine-N,N′-bis(2-ethanesulfonic acid) (PIPES)-NaOH, pH 6.8, 100 mM NaCl, 300 mM sucrose, 3 mM MgCl₂, 1 mM EGTA, 0.5% Triton X-100, 1 μg/ml leupeptin, 1 μg/ml aprotinin, 1 μg/ml pepstatin, and 1 mM phenylmethylsulfonyl fluoride (PMSF)]. For extracts of Son-depleted cells, cells were first transfected with 100 nM Son siGENOME1 or siGENOME4 48 h before preparation of the nuclear insoluble fraction. After centrifugation at 600 × g for 3 min, the pellets were treated with DNase I at 200 U/ml at 25°C for 30 min in digestion buffer (10 mM PIPES-NaOH, pH 6.8, 50 mM NaCl, 300 mM sucrose, 3 mM MgCl₂, 1 mM EGTA, 0.5% Triton X-100, and 1.2 mM PMSF). Ammonium sulfate was then added to the digested mixtures from a 1 M stock solution to a final concentration of 0.25 M, and the mixtures were incubated at 25°C for 5 min, followed by centrifugation at 600 × g for 3 min. For further extraction, the pellets were extracted in the digestion buffer containing 2 M NaCl by incubation at 4°C for 20 min, followed by centrifugation at 600 × g for 3 min. To remove RNA related components, the pellet was incubated with 0.2 mg/ml RNase A in the digestion buffer at 25°C for 30 min, and the mixtures were pulled down by centrifugation at 600 × g for 3 min to obtain a final fraction (high salt nuclear insoluble fraction). The fraction was directly dissolved in SDS sample buffer (50 mM Tris-HCl, pH 6.8, 2% SDS, 10% glycerol, and 1% β-mercaptoethanol) for one-dimensional-SDS-polyacrylamide gel electrophoresis (PAGE).

Antibodies, Immunofluorescence, and Immunoblots

We have generated affinity-purified rabbit polyclonal anti-Son antibodies at Covance (Denver, PA). Two polyclonal antisera (WU 14 and WU15) were directed against the peptide CEESESKTKSH located near the amino terminus of Son. For immunofluorescence, HeLa cells were fixed with 2% formaldehyde, washed three times for 5 min each in phosphate-buffered saline (PBS), and then permeabilized with 0.2% Triton X-100 in PBS. After 3 × 5 min washes in PBS 0.5% normal goat serum (NGS), cells were incubated with primary antibodies WU14 (1:1000), WU15 (1:15,000), monoclonal anti-SF2/ASF AK103 (1:2500; provided by A. Krainer, Cold Spring Harbor Laboratory, Cold Spring Harbor, NY), monoclonal anti-U1-70K 2.73 (1:100; provided by S. Hoch, Agouron Institute, La Jolla, CA) and polyclonal anti-pinin A301-022A (1:1000; Bethyl Laboratories, Montgomery, TX), human anti-nucleolar antibody (anti-fibrillarin, 1:20; Sigma-Aldrich, St. Louis, MO), and monoclonal anti-PML 5E10 (1:20; provided by R. van Driel, University of Amsterdam, Amsterdam, The Netherlands) for 1 h at room temperature. Cells were washed three times for 5 min with PBS, 0.5% NGS and then incubated in fluorescently conjugated secondary antibodies (1:500; Jackson ImmunoResearch Laboratories, West Grove, PA). Peptide blocking experiments were performed by preincubation of anti-Son antibodies with 10 μM N-terminal Son peptide on a rotating platform for 1 h at room temperature before addition of the antibody to the cells. 5-Bromouridine 5′-triphosphate (bromo-UTP) labeling and oligo(dT) RNA-fluorescence in situ hybridization (FISH) were performed as described previously (Sacco-Bubulya and Spector, 2002). For imaging of fixed material, z-stacks spanning the entire nuclei were acquired on a DeltaVision RT using a 60× objective (1.4 numerical aperture; Olympus, Tokyo, Japan), and raw images were displayed as volume projections.

For immunoblotting of nuclear insoluble fractions, 10 μg of protein from each fraction were separated on a 4–20% gradient polyacrylamide gel and transferred into polyvinylidene difluoride membrane at 70 V for 3 h. Signals were detected using SuperSignal West Femto Maximum Sensitivity Substrate (Thermo Fisher Scientific, Waltham, MA). For all other immunoblots, cells were extracted in Laemmli buffer, and equal amounts of protein were applied to 7% SDS-PAGE. Proteins were transferred to nitrocellulose membrane, blocked in PBS containing 0.1% Tween 20 and 5% nonfat dry milk and probed with primary antibodies (anti-Son WU15, anti-Son WU14, and anti-Lamin B1 were used at 1:1000; anti U1-70K and anti-SF2/ASF were used at 1:100; anti-β-actin [Sigma-Aldrich] was used at 1:10,000; and anti-histone H3 phospho S10 (Abcam, Cambridge, United Kingdom) was used at 1:500). Anti-rabbit immunoglobulin Ig (IgG) horseradish peroxidase (HRP)-conjugated antibody was used at 1:40,000 for blotting of the nuclear insoluble fraction, and anti-mouse and anti-rabbit IgG HRP-conjugated antibodies (Jackson ImmunoResearch Laboratories) were used at 1:50,000 for all other immunoblots.

RNA Interference (RNAi)

Predesigned siGENOME siRNAs directed against Son were purchased from Dharmacon RNA Technologies (Lafayette, CO) (catalog no. D-012983-01, target sequence GCUGAGCGCUCUAUGAUGU and catalog no. D-012983-04, target sequence CAAUGUCAGUGGAGUAUCA). HeLa cells (3 × 10⁴) were plated in antibiotic free medium and transfected by using Oligofectamine (Invitrogen, Carlsbad, CA) alone (mock) or mixed with siRNAs against luciferase (60 pmol; control; catalog no. D-001210-02; Dharmacon RNA Technologies) or Son siRNAs (60 pmol). For siRNA rescue experiments, 10 μg each of YFP-siR-Son or YFP-Son plasmid was transfected using the calcium phosphate method. After a 48-h recovery period, the cells were transfected with siRNAs and harvested for analysis 48 h later. To assess Son mRNA knockdown by semiquantitative reverse transcription (RT)-PCR, reverse transcriptase reactions were performed on 100 ng of total RNA (SuperScript One-Step RT-PCR System; Invitrogen). The linear range for amplification was verified, after which 32 cycles were used to amplify both Son and glyceraldehyde-3-phosphate dehydrogenase (GAPDH) mRNAs.

Live Cell Microscopy

HeLa cells (3 × 10⁵) stably expressing YFP-SF2/ASF were seeded onto 60-mm coverslips 24 h before RNAi. siRNA-mediated depletion of Son was done by transfection with siGENOME4. Thirteen hours after transfection, the cells were transferred to a live imaging chamber (Bioptechs, Butler, PA), perfused with L-15 medium (without phenol red) containing 20% fetal bovine serum, and placed into a 37°C environmental chamber on the stage of a DeltaVision RT imaging system. Nuclear speckle dynamics were observed by acquiring z-stacks through the entire nucleus every 5 min for at least 6 h.

Cell Growth and Flow Cytometry Analysis

For cell growth/cell cycle analysis, HeLa cells were plated in quadruplicate for a time course of both ViCell counts and flow cytometry analysis at the time of siRNA treatment and every 24 h thereafter. Cells treated in parallel were harvested 48 h after siRNA transfection for Western blots or to confirm Son knockdown by RT-PCR. A small coverslip was included in each set of cell growth and flow cytometry experiments and processed for immunofluorescence to confirm change in nuclear speckle morphology.

Son Is a Bona Fide Nuclear Speckle Protein

When Son was first identified in the nuclear speckle proteome, a 5′-truncated cDNA (KIAA1019) was subcloned into pEYFP, and the YFP-KIAA1019 fusion protein was expressed in HeLa cells (Sacco-Bubulya and Spector, 2002; Saitoh et al., 2004). YFP-KIAA1019 showed a precise overlay with splicing factor SC35, confirming its localization in nuclear speckles (Saitoh et al., 2004). The full-length sequence of Son has not previously been assembled. Here, we have generated a full-length Son that is fused with YFP, and expression of this fusion in HeLa cells shows precise localization with the SR splicing factor SF2/ASF in nuclear speckles (Figure 2A), consistent with previous reports of Son localization (Wynn et al., 2000; Saitoh et al., 2004). Furthermore, to ensure that endogenous Son is localized in nuclear speckles, we generated two rabbit polyclonal antibodies, designated as WU14 and WU15, against peptides corresponding to sequence near the amino terminus of Son. Immunolabeling of HeLa cells with affinity purified anti-Son antibody WU14 and WU15 shows colocalization with SF2/ASF in nuclear speckles (Figure 2B). Antibody specificity was shown by peptide competition assay in an immunofluorescence experiment (Figure 2C). In addition, the labeling of nuclear speckles with anti-Son WU14 antibody was consistently diminished in cells treated with siRNAs that target Son mRNA (Figures 3 and 4; data for WU15 were identical and are not shown). Together, the results of these experiments demonstrate specificity of our WU14 and WU15 anti-Son antibodies. Interestingly, proteomic analysis of insoluble proteins from HeLa-S3 cell nuclei revealed Son as one of the largest proteins in this fraction (Takata et al., 2009). An immunoblot with anti-Son antibody WU15 shows specific reactivity with a large protein that is enriched in a nuclear insoluble fraction of HeLa cells (Figure 2D, lane 3) compared with whole cell extract (Figure 2D, lane 1) or nuclear extract (Figure 2D, lane 2). As expected, there was enrichment of the nuclear lamina protein lamin B1 in the insoluble nuclear fraction as shown by immunoblot. Treatment of HeLa cells with Son siRNAs causes disappearance of the band that corresponds to Son (Figure 2D, lane 4), which is further evidence that the WU15 anti-Son antibody reactivity is specific to Son (similar results were obtained using WU14 anti-Son antibody; data not shown). Preincubation of the anti-Son antibodies with the peptide used to generate WU14 and WU15 blocks reactivity with Son by immunoblot, demonstrating antibody specificity (data not shown).

Nuclear Speckle Organization of pre-mRNA Splicing Factors Requires Son

Immunofluorescence experiments and YFP-Son expression show the predominant localization of Son in nuclear speckles, with little detectable diffuse nuclear signal (Figure 2, A and B). We did not detect Son in other nuclear domains such as PML bodies, nucleoli, or Cajal bodies (Supplemental Figure 1). Based on motif analysis and localization studies, we hypothesized that Son may be a structural component of the nuclear speckles. If Son is essential for nuclear speckle organization, we would expect knockdown of Son to significantly alter the distribution of pre-mRNA processing factors. We used RNAi to knock down the expression of Son in HeLa cells, and we evaluated changes in the distribution of pre-mRNA processing factors. U1-70K small nuclear ribonuclear protein (snRNP) protein localization was dramatically altered in the nuclear speckles in cells that showed significant reduction of Son immunostaining after Son depletion compared with controls (Figure 3). Son depletion by two separate siRNAs caused a significant reduction of Son immunostaining (Figure 3, i and m) as well as a complete reorganization of U1-70K to round, doughnut-shaped nuclear structures (Figure 3, j and n, arrows). Importantly, Son depletion did not alter the U1-70K distribution in Cajal bodies that are morphologically distinct from the nuclear speckles (Figure 3n, arrowhead). Similar data for nuclear speckle reorganization was obtained using human anti-Sm autoimmune serum and Y12 anti-snRNP antibody (data not shown). These data demonstrate that although snRNPs do not require Son to maintain their localization in Cajal bodies, their organization in nuclear speckles requires that Son maintain nuclear speckle structure.

Nuclear distribution of SR proteins SF2/ASF and SC35 was also altered by Son depletion. Figure 4A shows that SF2/ASF localization in nuclear speckles was altered after depletion of Son with siRNAs (Figure 4, j and n, arrows) but not with mock transfection (Figure 4b, arrow) or control siRNA (Figure 4f, arrow). We also observed similar results in other cell lines including IMR90 and NIH 3T3 (data not shown). Identical results were seen with HeLa cells expressing YFP-SF2/ASF (see Figure 9). The Son depletion phenotype was identical for two additional siRNAs (data not shown). The similar outcome for Son depletion using four separate siRNAs, each of which target a different region of Son mRNA, suggests that the alteration in nuclear speckle phenotype is not an off-target effect of RNA interference. HeLa cells stably expressing SC35-YFP were also treated with Son siRNAs and processed for immunolocalization of Son. SC35-YFP localization in nuclear speckles was altered to a doughnut-shaped pattern after depletion of Son (Supplemental Figure 2). Identical results were observed for the splicing factor pinin (see Supplemental Figure 2). A representative example of semiquantitative RT-PCR used to verify Son knockdown is shown in Figure 4B. Son mRNA is significantly reduced only after treatment with Son siRNAs (Figure 4B, top, lanes 4 and 8) but not after mock transfection (Figure 4B, top, lanes 2 and 6) or treatment with control siRNA (Figure 4B, top, lanes 3 and 7). Although partially reduced expression of Son at 24 h after RNAi was always sufficient to observe changes in nuclear speckle organization, continued decrease in Son expression was typically observed over 48 h. Notably, the reorganization of nuclear speckle components is not limited to splicing factors. We expressed YFP-tagged exon junction complex (EJC) central core proteins Magoh and MLN51 (Degot et al., 2004; Baguet et al., 2007) in HeLa cells to observe their response to Son depletion. Both proteins localized in nuclear speckles (Figure 5, a–h and m–t), and became redistributed into doughnut-shaped nuclear speckles after Son depletion (Figure 5, i–l and u–x).

Nuclear speckles and associated nuclear domains such as paraspeckles contain polyadenylated RNA species whose identities and functions are only beginning to be characterized. Oligonucleotide [oligo(dT)] probe has been used to detect stable polyA⁺ RNA in nuclear speckles (Huang et al., 1994) and was used here to monitor the response of nuclear speckle polyA⁺ RNA after Son knockdown. Although there was approximately equal labeling of polyA⁺ RNA in all conditions, Son knockdown caused dramatic redistribution of nuclear speckle polyA⁺ RNA (Figure 6, e–h). The similar effects of Son depletion on polyA⁺ RNA, snRNPs, SR proteins, pinin, and EJC core proteins suggest that Son may have general organizational functions for pre-mRNA processing factors as well as other constituents of the nuclear speckles.

Because indirect effects on nuclear speckle organization might result if Son depletion alters transcription, it is important to determine whether the onset of pre-mRNA processing factor redistribution is a result of decreased global gene activity in Son-depleted cells. To address this, we performed bromo-UTP incorporation in living HeLa cells treated with Son siRNAs for 24 h. Labeling of transcripts was equivalent in Son depleted and control cells, indicating that global transcription proceeds robustly despite nuclear speckle reorganization (Figure 7A). Furthermore, if transcription were inhibited under conditions of Son depletion, we would expect to see pre-mRNA machinery in large rounded speckles and not as doughnut shapes. We confirmed that transcription inhibition of RNA polymerase II by α-amanitin treatment does not cause Son or SF2/ASF to exhibit a doughnut-like morphology. Rather, it leads to the expected larger round nuclear speckle organization of these factors (Figure 7B). We conclude that transcription activity is not reduced at the time that nuclear speckles reorganize after Son depletion, ruling out transcription-related secondary effects.

Immunofluorescence labeling of splicing factors in the nucleoplasm sometimes seemed reduced when the doughnut-shaped nuclear speckles were observed after Son depletion. To determine whether splicing factors are destabilized by disruption of nuclear speckle organization, HeLa cells treated with siRNAs against Son for 48 h were extracted for immunoblot analysis of splicing factor abundance. Expression of splicing factors U1-70K and SF2/ASF was not significantly different in extracts from Son-depleted cells (Figure 8A, lanes 3 and 4) compared with cell treated with mock transfection (Figure 8, lane 1) or control siRNA (Figure 8, lane 2). Son mRNA was nearly absent after treatment of cells with Son siRNAs (Figure 8, lanes 8 and 9), whereas it was not reduced in cells treated with mock transfection (Figure 8, lane 6) or control siRNA (Figure 8, lane 7). These data suggest that the decrease in immunofluorescence signal of splicing factors observed after morphological alteration of Son-depleted nuclear speckles is due to general redistribution of these proteins and not to decreased protein levels.

Nuclear Speckle Disorganization Occurs Quickly and Directly upon Son Depletion

Immunofluorescence experiments performed on Son-depleted cells indicated that the redistribution of pre-mRNA splicing factors into a doughnut-shaped structure began reproducibly across the entire coverslip of cells at ∼14–16 h after siRNA treatment when Son expression begins to decline (data not shown). We performed time-lapse microscopy to determine whether the nuclear speckles completely disassembled and then reorganized into these structures, or whether each nuclear speckle remained intact but simply reorganized directly to a different shape. Live imaging of HeLa cells stably expressing YFP-SF2/ASF showed that morphologically normal nuclear speckles change their organization directly into a “doughnut” over time (Figure 9). At 13 h post-siRNA treatment, the YFP-SF2/ASF is distributed in nuclear speckles (Figure 9a) and then begin to reorganize into “doughnuts” at ∼15–16 h (Figure 9, c and d), which are completely formed by 18 h after Son depletion is initiated (Figure 9f). A time-lapse video of YFP-SF2/ASF localization showed direct transition from nuclear speckles to doughnut-shaped structures (see Supplemental Movie S1). Importantly, this morphological change over time was not seen when YFP-SF2/ASF HeLa cells were treated with control siRNA (data not shown). Time-lapse imaging has therefore ruled out a disassembly/reassembly model for the structural alteration observed after Son depletion. Furthermore, it indicates that the onset of morphological change in nuclear speckle structure occurs within a few hours of the onset of Son depletion.

The Unique Son Repeat Domain Rescues Nuclear Speckle Organization in Cells Depleted of Endogenous Son

Seven silent mutations were introduced into the nucleotide sequence of YFP-tagged Son cDNA (see Materials and Methods) in the 5′ coding region that corresponds to the position in Son mRNA targeted by Son siRNA 4 to produce a siRNA-refractory YFP-tagged Son (referred to here as YFP-siR-Son). Expression of YFP-Son-siR in HeLa cells showed that it colocalizes with SF2/ASF in nuclear speckles (Figure 10A), which is expected because it has the same amino acid sequence as YFP-Son. Experiments were performed to analyze SF2/ASF distribution in Son-depleted cells that are protected by expression of YFP-siR-Son (Figure 10A). As expected, treatment of cells with control siRNA (Figure 10, a–d) did not affect distribution of SF2/ASF. Treatment of cells with Son siRNA 4 yielded doughnut-like redistribution of SF2/ASF in cells that were not protected by siRNA-resistant Son (Figure 10, e–h, arrow), while yielding normal nuclear speckle localization for SF2/ASF in cells that were protected by expression of YFP-siR-Son (Figure 10, e–h, the top cell is transfected with YFP-siR-Son). In addition, YFP-siR-Son expression did not rescue SF2/ASF distribution in cells treated with siRNA 1, because the region of Son targeted by this oligo has not been modified to resist its effects (data not shown). Likewise, YFP-Son could not rescue SF2/ASF distribution because it was targeted by siRNAs (Figure 10B, lanes 4 and 5, top; immunofluorescence data not shown). Semiquantitative RT-PCR was used to determine the levels of exogenous and endogenous Son expression in this experiment. RT-PCR to detect the YFP portion of exogenously expressed constructs verified siRNA-resistant expression of YFP-siR-Son (Figure 10, lane 10, top), whereas RT-PCR against endogenous Son shows depletion of endogenous Son (Figure 10B, lane 10, middle). YFP-siR-Son resisted depletion by siRNA 4 but not siRNA 1 (Figure 10B, compare lanes 9 and 10 in upper panel). Both endogenous Son and nonresistant YFP-Son were reduced by both siRNAs (Figure 10B, lanes 4 and 5). This set of experiments demonstrates that siRNA-resistant Son expression prevents alterations in nuclear speckle phenotype in Son-depleted cells.

To address our hypothesis that the unique repeats of Son mediate proper nuclear speckle organization, similar experiments were performed to determine whether the repeat region of Son could rescue nuclear speckle organization in Son-depleted cells. Several C-terminal deletion constructs were made in YFP-siR-Son at amino acid positions indicated on Figure 1. These constructs were transiently expressed before Son depletion to determine whether they could protect nuclear speckles from reorganizing. Importantly, cells that expressed YFP-siR-Son amino acids 1-1493 containing the unique repeat region (Rpt) of Son showed unaltered nuclear speckle localization for SF2/ASF after Son depletion (Figure 10, m–p, the cell on the right is transfected with YFP-siR-Son repeats) compared with controls (Figure 10, i–l). RT-PCR confirmed that the YFP-siR Son resisted depletion by siRNA 4 (Figure 10, lane 15, top), whereas endogenous Son was depleted (Figure 10, lane 15, middle). An intermediate deletion construct encoding amino acids 1-2008 includes the RS domain but lacks the G-patch and DS RBD and was also able to rescue nuclear speckle organization (Supplemental Figure 3); however, because amino acids 1-1493 are sufficient for rescue, the RS domain, G-patch, and DS RBD of Son are clearly dispensable. Finally, a construct comprised of the extreme amino terminus of Son (amino acids 1-334) did not rescue nuclear speckle organization (Supplemental Figure 3). We therefore conclude that the unique repeat region of Son mediates proper nuclear speckle organization.

Son Is Essential for Cell Viability and Cell Cycle Progression

Treatment of HeLa cells with four separate siGENOME oligos (Dharmacon RNA Technologies) directed against Son leads to a significant reduction of Son mRNA and protein by 24 h posttransfection. To determine longer term effects of Son depletion, we performed a time course of siRNA treatment. Visual observations by phase contrast microscopy indicated that longer siRNA treatments correlated with reduced cell numbers, suggesting that Son depletion interferes with cell growth. This is consistent with a previous report that also noted Son is required for cell growth (Ahn et al., 2008). However, exactly which phase of the cell cycle is affected has not been determined. To begin to understand the effects of Son depletion on cell viability and cell cycle, we performed cell counts in untreated cells and mock-treated cells versus cells treated with control siRNA or Son siRNAs (Figure 11A). For these experiments, a small coverslip was inserted into each culture dish and processed for immunofluorescence to confirm the expected change in nuclear speckle morphology in Son-depleted cells (data not shown). We counted the total numbers of cells in mock, control siRNA oligo and Son siRNA-treated samples at different time points after siRNA treatment, and automated trypan blue staining was used to score cell viability in the same experiment. Growth curves indicate that cell growth was stalled in Son-depleted cells by 48 h, compared with cells treated with mock transfection and control siRNA that continued to grow with the normal doubling time of ∼24 h. The graph depicts the average numbers of viable cells counted at each time point from three separate groups of experiments.

To determine whether the stalled growth after 48 h of Son depletion was due to changes in the cell cycle, we performed flow cytometry analysis that indicated the cell cycle distribution of control siRNA-treated cells was similar to that of untreated cells. siRNA experiments from the 48-h time point in part A were analyzed by flow cytometry; data shown are representative of one of the four experimental groups (Figure 11B). A corresponding table of the percent of cells in each cell cycle stage indicates a 9–12% decrease in G1 stage cells after Son depletion with two separate siRNAs (compared with control siRNA), with a concomitant increase (8–12%) of cells in G2/M stages. Flow cytometry analysis of cells at 72 h after Son depletion showed similar results, also with a similar decrease in G1 and increase in G2/M phase cells (data not shown).

Immunoblot analysis of Son-depleted cells was performed to distinguish between growth arrest at G2 versus M phase after Son depletion. We assessed whether there was an increase in phosphorylated histone H3 after Son depletion, as this would indicate an increase in mitotic cells. Cells harvested from 48 h after Son depletion in the same growth curve experiment shown in Figure 11, A and B, were extracted in Laemmli buffer for immunoblot analysis of histone H3Ser(10)P (Figure 11C). A significant increase in phosphorylation of histone H3Ser(10) was detected in extracts from cells treated with Son siRNAs (Figure 11C, lanes 3 and 4) compared with extracts from mock-treated cells (Figure 11C, lane 1) and cells treated with control siRNA oligos (Figure 11C, lane 2). Extract from nocodazole arrested HeLa cells (Figure 11C, lane 5) was used as a positive control for elevated H3Ser(10)P in mitosis. These results suggest that Son depletion leads to interrupted cell growth patterns presumably by interfering with progression through mitosis.

The increase in mitotic cells was also observed directly by microscopy, and the number of cells in each phase of mitosis was scored in Son-depleted cells (Figure 12). HeLa cells expressing H2B-GFP were treated with mock transfection, a mix of siRNAs against Son, or Son siRNA oligo 4. An increase of cells in mitosis was readily observed in Son-depleted cells after 36–48 h (Figure 12A, i–j and n–o) compared with mock-treated cells (Figure 12A, d and e). The mitotic index of cells scored over time after siRNA treatment (Figure 12B) also indicates a 9% increase in mitotic cells by 48 h that is consistent with the flow cytometry data shown in Figure 10B. Finally, the percentage of cells in each mitotic phase was determined (Figure 12C). After 48 h of Son depletion, the number of prometaphase/metaphase cells nearly doubles (increases from 54 to 96% of the mitotic cells), whereas the number of cells in anaphase and telophase decreases dramatically. This demonstrates that Son is necessary for cell cycle progression and that the absence of Son causes mitotic arrest at metaphase.