The putative RNA helicase DDX1 associates with the nuclear RNA exosome and modulates RNA/DNA hybrids (R-loops)

Abstract

The RNA exosome is a ribonuclease complex that mediates both RNA processing and degradation. This complex is evolutionarily conserved, ubiquitously expressed, and required for fundamental cellular functions, including rRNA processing. The RNA exosome plays roles in regulating gene expression and protecting the genome, including modulating the accumulation of RNA-DNA hybrids (R-loops). The function of the RNA exosome is facilitated by cofactors, such as the RNA helicase MTR4, which binds/remodels RNAs. Recently, missense mutations in RNA exosome subunit genes have been linked to neurological diseases. One possibility to explain why missense mutations in genes encoding RNA exosome subunits lead to neurological diseases is that the complex may interact with cell- or tissue-specific cofactors that are impacted by these changes. To begin addressing this question, we performed immunoprecipitation of the RNA exosome subunit, EXOSC3, in a neuronal cell line (N2A), followed by proteomic analyses to identify novel interactors. We identified the putative RNA helicase, DDX1, as an interactor. DDX1 plays roles in double-strand break repair, rRNA processing, and R-loop modulation. To explore the functional connections between EXOSC3 and DDX1, we examined the interaction following double-strand breaks and analyzed changes in R-loops in N2A cells depleted for EXOSC3 or DDX1 by DNA/RNA immunoprecipitation followed by sequencing. We find that EXOSC3 interaction with DDX1 is decreased in the presence of DNA damage and that loss of EXOSC3 or DDX1 alters R-loops. These results suggest EXOSC3 and DDX1 interact during events of cellular homeostasis and potentially suppress unscrupulous expression of genes promoting neuronal projection.

Keywords: DDX1; DRIP sequencing; R-loop; RNA degradation; RNA exosome; RNA helicase; RNA processing; RNA sequencing.

Figure 1

RNA exosome subunits coimmunoprecipitate with tagged EXOSC3.A, the RNA exosome is a conserved exo/endoribonuclease complex that comprises ten subunits. Nine of the ten subunits are structural and termed exosome components. EXOSC1, EXOSC2, and EXOSC3 make up the cap and EXOSC4, EXOSC5, EXOSC6, EXOSC7, EXOSC8, and EXOSC9 make up a barrel-shaped core. In this graphic, EXOSC3 (navy) and EXOSC9 (green) are highlighted. EXOSC6, EXOSC7, and EXOSC8 are positioned behind subunits EXOSC4, EXOSC5, and EXOSC9, and consequently are not visible. The catalytic subunit, DIS3 or DIS3L, sits at the base of the complex [PDB 6H25 (6)]. B, EXOSC9 core subunit coprecipitates with myc-EXOSC3 from murine neuronal N2A cell line. Cells were transfected with a plasmid encoding Vector control or myc-EXOSC3, followed by immunoprecipitation using anti-myc magnetic beads. Input for Vector control and myc-EXOSC3 was probed by an anti-myc antibody and a band corresponding to the molecular weight is detected in the input but not Vector control for myc-EXOSC3. Input for Vector control and myc-EXOSC3 is probed by an anti-EXOSC9 antibody and a band at the corresponding molecular weight is present in both lanes. Bound for Vector control and myc-EXOSC3 is probed by an anti-myc and an anti-EXOSC9 antibody, and a band corresponding to the molecular weight is detected in the bound fraction for myc-EXOSC3 but not for Vector control. Stain-free blot indicates the loading of total protein in the input. Immunoprecipitation of the myc-tagged EXOSC3 copurifies with the endogenous EXOSC9 subunit. C, eluates of the bound myc-EXOSC3 immunoprecipitation were analyzed by LC-tandem mass spectrometry. A table shows all RNA exosome subunits detected, listing the peptide-spectrum matches (PSM) and the peptide numbers for each subunit. Vector IP serves as a control. IP, immunoprecipitation.

Figure 2

Novel EXOSC3/RNA exosome interactors identified using LC-tandem mass spectrometry.A, pie chart that organizes proteins that coprecipitate with myc-EXSOSC3 from murine neuronal N2A cell line by protein class. Peptide-spectrum matches (PSM) were analyzed using a log₂ ratio, so that any result above 0 indicates binding in myc-EXOSC3 IP over vector IP. Results that are less than or equal to 0 were excluded from the analyses. Proteins that contained a PSM beyond the log₂ cut-off of 0 were analyzed by Panther gene ontology terms. The number of proteins within a class is indicated inside the pie slices. B, the table (left) shows selected RNA exosome interactors detected and lists the PSM and Peptide number for each protein. Vector IP serves as the control. The gene list corresponding to the proteins is provided as Table S1. Note that MPP6 is the protein name and MPHOSPH6 is the gene name. MTR4 is the most common name for this helicase but the name appears as SKIV2L2 in Table S1. A short list of nucleic acid metabolism/binding proteins (right) is provided. The RNA exosome subunits (bold, navy) reside in the nucleic acid metabolism/binding subcategory, together with known RNA exosome cofactors (green). Candidates investigated as novel RNA exosome interactors are listed in blue, including the putative RNA helicase, DDX1. IP, immunoprecipitation; MMP6, M-phase phosphoprotein 6.

Figure 3

DDX1 coimmunoprecipitates with EXOSC3.A, a graphical representation of the domain structure of human putative DEAD-box helicase DDX1. The SPRY protein–interacting domain in DDX1 is located between a phosphate-binding P-loop motif and an ssDNA binding Ia motif, separating the motifs by 240 residues instead of the usual 20 to 40 residues seen in other DEAD-box proteins (54). The catalytic ATP-binding helicase and C-terminal helicase domains lie downstream of the SPRY domain. B, DDX1 coimmunoprecipitates with EXOSC3 in the nuclear, but not cytoplasmic, fraction of N2A cell lysate. EXOSC3 was immunoprecipitated from the cytoplasmic or nuclear fraction, followed by immunoblotting using EXOSC3, EXOSC9, and DDX1 antibodies. The input, unbound and bound fractions from the EXOSC3 and control nonspecific rabbit IgG (Ctrl IgG) immunoprecipitation are shown. EXOSC9 serves as a representative of the copurified RNA exosome subunits. Stain-free blot serves as the loading control. C, immunoprecipitation of EXOSC3 from the nuclear fraction was performed with RNase A treatment. EXOSC3 antibody described previously is used in the No treatment and +RNase A immunoprecipitation. Nonspecific rabbit IgG (Ctrl IgG) was used as a control. EXOSC3 and DDX1 were analyzed for the input and bound fractions (IP: EXOSC3). The IgG light-chain band is visible (asterisk) in the bound fractions probed with EXOSC3, just below the EXOSC3 band. Stain-free blot serves as a loading control for the input. The experiment was performed in biological replicates (n = 5) and bands in the bound fractions were quantified relative to the No treatment control. The values below the lanes correspond to the amount of protein quantified from the bands. The asterisk (∗) below the values indicate the p value <0.05. IgG, immunoglobulin G; IP, immunoprecipitation.

Figure 4

The interaction between EXOSC3 and DDX1 is sensitive to DNA damage.A, N2A cells were treated with camptothecin (CPT) or PBS (control), fixed, and analyzed by immunofluorescence using an antibody that detects the DNA damage marker, γH2AX. B, the input and immunoprecipitated samples from nuclear fractions (bound) treated with either CPT or PBS (control) for both EXOSC3 and control IgG (Ctrl IgG) are shown. DDX1, EXOSC9, and EXOSC3 are detected. C, the immunoprecipitation experiment in Figure 4B was performed in biological triplicate and DDX1 bands in EXOSC3 bound fractions were quantified. Statistical significance was calculated by a student’s t test. Asterisk (∗) represents p value <0.05. IB, immunoblot; IgG, immunoglobulin G; IP, immunoprecipitation.

Figure 5

EXOSC3 and DDX1 are robustly depleted by siRNA-mediated knockdown in N2A cells.A, N2A cells were transfected with scramble control or EXOSC3 siRNA (siEXOSC3) in biological triplicate. The steady-state level of EXOSC3 was assessed by immunoblotting. The asterisk (∗) indicates a nonspecific band. Stain-free blot serves as a loading control. B, quantification of immunoblot in (A) shows that EXOSC3 is depleted to 13.1%. The percent depletion of EXOSC3 was determined by quantifying the immunoblot in (A) relative to scramble and averaging the values. This result is significant across three biological replicates. C, N2A cells were transfected with scramble or DDX1 siRNA (siDDX1) in biological triplicate. The steady-state level of DDX1 was assessed by immunoblotting. Stain-free blot serves as a loading control. D, quantification of immunoblot in (C) shows that DDX1 is depleted to 12.4%. The percent depletion of DDX1 was determined by quantifying the immunoblot in (C) relative to scramble and averaging the values. This result is significant across three biological replicates. The statistical analyses for (C and D) were calculated using a student’s t test. Asterisks (∗∗∗∗) represent a p value <0.0001 and (∗∗∗) represent a p value <0.001.

Figure 6

Depletion of EXOSC3 or DDX1 results in misprocessing of rRNA precursors.A, a graphical schematic of murine rRNA processing, adapted from Henras et al., 2015 (71) that includes up and down arrows to summarize the results obtained in this study. B, Northern blots of total RNA from N2A cells depleted of EXOSC3 or DDX1 using rRNA probes show that levels of 12S rRNA precursors and 5.8S rRNA are altered. EXOSC3 or DDX1 was depleted from cells by siRNA knockdown and total RNA was isolated for Northern blotting. An ITS2 probe was used to detect 32S and 12S rRNA precursors. Additionally, we employed a probe specific for 5.8S rRNA. The 7SL transcript serves as a loading control. Lanes 1 to 3 indicate the scramble control; lanes 4 to 6 indicate siEXOSC3; lanes 7 to 8 indicate siDDX1. C, Northern blots from Figure 6B were quantified relative to 7SL in biological triplicates. An asterisk (∗) indicates a significant difference using a p value cut-off of < 0.05.

Figure 7

DRIP-seq reveals that depletion of EXOSC3 or DDX1 alters R-loop regions. DNA/RNA immunoprecipitation followed by sequencing (DRIP-seq) was performed on N2A cells depleted of EXOSC3 or DDX1 by siRNA. A, the volcano plots show the number of R-loop regions that statistically increase and decrease in either EXOSC3 or DDX1 depletions compared to scramble control. The plot is graphed using a log₂ fold change across a -log₁₀ false discovery rate (FDR). The R-loop regions that did not achieve the FDR cut-off of <0.05 are indicated in black and fall under the horizontal lines. The numbers on the left of 0 on the x-axis indicate significantly decreased R-loop regions and the numbers on the right of 0 indicate significantly increased R-loop regions. B, the statistically increased and decreased R-loop regions in cells siRNA depleted of EXOSC3 or DDX1 identified in the DRIP-seq dataset are compared using a Venn diagram. The dark blue circles indicate the number of increased (n = 722) or decreased (n = 935) R-loop regions upon siRNA-mediated EXOSC3 depletion. The yellow circles represent the number of increased (n = 638) or decreased (n = 1058) R-loop regions that were affected upon siRNA-mediated DDX1 depletion. The overlap in green indicates the number of increased (n = 140) or decreased (n = 425) R-loop regions that siEXOSC3 or siDDX1 have in common (p value < 2.2 × 10⁻¹⁶). C, the increased and decreased R-loop regions from both siEXOSC3 and siDDX1 samples were analyzed using Panther gene ontology terms and categorized by biological process. These analyses are statistically significant with a cut-off at p value <0.05. D, the classes of RNA that were affected upon depletion of either EXOSC3 or DDX1 are organized into a pie chart. Protein-coding genes were excluded. The numbers to the right of the class of RNA are the number of R-loop regions that correspond to that class.

Figure 8

RNA-seq shows that depletion of EXOSC3 or DDX1 results in more shared decreased transcripts than shared increased transcripts.A, RNA-seq was performed on N2A cells siRNA depleted of EXOSC3 or DDX1. The volcano plots show the number of mRNA transcripts that increased and decreased in either siEXOSC3 or siDDX1 compared to scramble control. The plot is graphed using a log₂ fold change across a -log₁₀ false discovery rate (FDR). The differential transcripts that did not met the FDR cut-off <0.05 are indicated in black and fall under the horizontal lines. The numbers on the left of 0 of the x-axis indicate significantly decreased differential transcripts and the numbers on the right of 0 indicate significantly increased differential transcripts. B, the increased and decreased transcripts are compared using a Venn diagram. The dark blue circles indicate the number of increased (n = 1757) or decreased (n = 2192) mRNA transcripts upon EXOSC3 depletion. The yellow circles represent the number of increased (n = 734) or decreased (n = 968) mRNA transcripts that are affected upon DDX1 depletion. The overlap (p-value < 2.2 × 10^-16) represented in green indicates the number of increased (n = 322) or decreased (n = 599) transcripts that are common to depletion of both EXOSC3 and DDX1.

Figure 9

Filtering DRIP reads through RNA-seq revealed genes that are simultaneously affected by depletions of EXOSC3 or DDX1.A, a graphical representation of the pipeline used to focus on specific genes using the analysis funnel. N2A cells siRNA depleted of either EXOSC3 or DDX1 were subjected to both DRIP- and RNA-seq. We filtered results from the DRIP-seq through the RNA-seq reads, using only mRNA transcripts. We were then able to produce heatmaps and examine transcriptomic regions using an integrated genomics viewer (IGV). B, using the pipeline described, we created heatmaps showing the landscape of increased and decreased R-loop regions upon depletion of either EXOSC3 (blue) or DDX1 (green). C, the IGV images of Ints6 and Celf4. The chromosome is displayed at the top of the window. The span lists the number of bases currently displayed. The tick marks indicate the chromosome locations. The red line marks the regions in which R-loops are significantly changed. The top track displays the Mus musculus reference genome (NCBI37/mm9) in orange. The following three tracks display the R-loop regions of interest corresponding to scramble, siEXOSC3, and siDDX1, respectively. The middle three tracks display the RNase H-treated R-loop regions of interest corresponding to scramble, siEXOSC3, and siDDX1, respectively. The last three tracks display the transcript regions of interest corresponding to scramble, siEXOSC3, and siDDX1, respectively. Celf4 exhibited three changed regions in the genes and is displayed by separate panels. D, quantification of Ints6 and Celf4 transcripts in scramble, siEXOSC3, and siDDX1 by fragment per kilo million reads (FPKM). DNA/RNA-immunoprecipitation followed by sequencing.