A role for the Cajal-body-associated SUMO isopeptidase USPL1 in snRNA transcription mediated by RNA polymerase II

Abstract

Cajal bodies are nuclear structures that are involved in biogenesis of snRNPs and snoRNPs, maintenance of telomeres and processing of histone mRNA. Recently, the SUMO isopeptidase USPL1 was identified as a component of Cajal bodies that is essential for cellular growth and Cajal body integrity. However, a cellular function for USPL1 is so far unknown. Here, we use RNAi-mediated knockdown in human cells in combination with biochemical and fluorescence microscopy approaches to investigate the function of USPL1 and its link to Cajal bodies. We demonstrate that levels of snRNAs transcribed by RNA polymerase (RNAP) II are reduced upon knockdown of USPL1 and that downstream processes such as snRNP assembly and pre-mRNA splicing are compromised. Importantly, we find that USPL1 associates directly with U snRNA loci and that it interacts and colocalises with components of the Little Elongation Complex, which is involved in RNAPII-mediated snRNA transcription. Thus, our data indicate that USPL1 plays a key role in RNAPII-mediated snRNA transcription.

Keywords: Cajal body; SUMO isopeptidase; snRNA transcription.

Fig. 1.

Consequences of USPL1 knockdown on nuclear architecture. (A) Immunofluorescence of U2OS cells transfected with siRNA as indicated were stained against USPL1, SMN and coilin. Closed arrows in top row indicate Cajal bodies with USPL1, SMN and coilin present. Open arrows in middle row highlight USPL1 nuclear foci in the absence of coilin; arrowheads (middle and bottom row) indicate SMN nuclear foci upon treatment with siRNA against coilin or USPL1. A maximum intensity projection in the right column demonstrates the increase in SMN nuclear foci (foci indicated by arrowheads, nucleus indicated by the dotted line). Similar effects were observed with HeLa cells (data not shown). (B) Immunofluorescence of siRNA transfected U2OS cells stained against USPL1, coilin and UBF. Similar effects were observed with HeLa cells. (C) A maximum intensity projection of siRNA treated HeLa cells stained with antibodies against PML and coilin. Similar effects were observed with U2OS cells (data not shown). (D) Immunofluorescence of HeLa cells transfected with siRNA as indicated were stained against ASF and coilin. Arrows highlight splicing speckles in control and coilin-depleted cells (top and middle row, respectively), arrowheads indicate enlarged splicing speckles upon USPL1 knockdown (bottom row). Similar effects were observed with U2OS cells (Fig. 7). Scale bars: 10 µm.

Fig. 2.

Knockdown of USPL1 affects pre-mRNA splicing. (A) An excess of HeLa (top panel) or U2OS (bottom panel) cells treated with siRNA against USPL1 were mixed with control cells and analysed by 5-ethynyl uridine pulse labelling to monitor nascent RNA transcription. siControl- and siUSPL1-treated cells can be distinguished by parallel immunostaining for coilin: arrows indicate control cells with coilin in Cajal bodies, whereas arrowheads highlight USPL1-knockdown cells with coilin in nucleoli. Note, to allow visibility of nucleolar coilin, Cajal-body-localised coilin was allowed to saturate the camera during image acquisition. Scale bar: 10 µm. (B) Changes in pattern of pre-mRNA splicing detected by qualitative RT-PCR on total RNA isolated from siRNA-transfected U2OS cells and subsequent ethidium bromide agarose gel electrophoresis. Similar effects were observed with HeLa cells (data not shown). Splicing products from either HeLa or U2OS USPL1-knockdown cells were sequenced and their positions are indicated next to the gel. White boxes indicate exons, filled black boxes introns, with the position of respective primers corresponding to the first and last box, respectively (size of box not to scale). An asterisk indicates a nonspecific band.

Fig. 3.

SILAC MS-based analysis reveals impact of USPL1 knockdown on localisation and mobility of spliceosome-associated proteins. (A) Experimental workflow of the cellular fractionation process and 1D SDS-PAGE for subsequent SILAC-MS analysis upon USPL1 knockdown. (B–D) SILAC analysis of cellular fractionation of HeLa cells upon USPL1 knockdown. Only hits identified with >1 peptide in an individual fraction are displayed here. The log₂ ratio heavy (H; siUSPL1)/light (L; siControl) of each cellular fraction is displayed as frequency histogram for cytoplasm (B) and nucleolus (C). Protein groups above/below the arbitrary threshold are indicated by a red lined box (B,C). The log₂ ratios [H (siUSPL1)/L (siControl)] of the cytoplasmic versus nucleolar fraction are displayed as a scatter plot with each dot representing an individual protein (D). Spliceosome proteins (green, based on Hegele et al., 2012) and ribosomal proteins (purple) are highlighted as indicated. The red lined box highlights protein groups enriched in the cytoplasmic fraction. Note, for clarity, the three major environmental contaminants (keratins) are not displayed in D.

Fig. 4.

Defects in snRNP production upon USPL1 knockdown. (A) Total cell lysates of siRNA-treated U2OS cells were analysed by immunoblotting using antibodies as indicated. Different amounts of control lysate (100%, 50%, 25%) provide an internal standard for protein levels. Similar effects were observed with HeLa cells (data not shown). (B) Immunoblot of total cell lysate of U2OS GFP–SmB cells without (−dox) or 18 hours after (+dox) induction using the Y12 or an anti-GFP antibody showing expression of GFP–SmB at its expected molecular mass. Alpha-tubulin (TUB) stain serves as loading control. (C) U2OS GFP–SmB cells were analysed without (−dox) or 18 hours after (+dox) induction for expression and localisation of GFP-SmB by immunostaining using the Y12 and anti coilin antibody as marker for nuclear speckles (arrowheads) and Cajal bodies (arrows), respectively. Scale bar: 10 µm. (D) Formation of nascent snRNPs was analysed by TMG immunoprecipitation in siRNA-treated U2OS GFP–SmB cells. Empty Protein-G–agarose beads (-) serve as a control. Bound proteins were eluted from the beads using 25 mM 7-methylguanosine (7-mG) and analysed for the presence of GFP–SmB by immunoblotting for GFP. Input corresponds to 0.75% of the material used in the immunoprecipitation.

Fig. 5.

Cellular snRNA levels are reduced upon knockdown of USPL1. (A) qRT-PCR for different major U snRNA species in HeLa cells treated with siRNA against USPL1. U2pre represents the unprocessed U2 snRNA. Respective expression levels were normalised for levels of β-actin and snRNA levels of siControl-treated cells were set to 1. Bars represent the s.e.m. of four independent experiments, each measured in technical replicates of 2. (B) Schematic representation of the initial U2 snRNA transcript (adapted from Broome and Hebert, 2012) depicting additional 634 bp after the end of the U2 snRNA coding sequence and the position of the respective primers for qRT-PCR used for U2 snRNA in A. (C) HeLa cells treated with control siRNA or siRNA against USPL1 were subjected to RNA-FISH for U2, U5 and U6 snRNA using Alexa-Fluor-488-labelled probes. The immunostaining against coilin was used to monitor the efficiency of the siRNA treatment against USPL1. Cajal bodies are indicated by arrows, splicing speckles by arrowheads and nucleolar coilin by an open arrowhead. Scale bar: 10 µm.

Fig. 6.

USPL1 colocalises and interacts with members of the LEC complex. (A) Total cell lysate obtained from HEK293 cells 24 hours after transfection with the indicated plasmids were subjected to immunoprecipitation using an anti-FLAG antibody. 1.5% of the total lysate was loaded as input (left). Bound proteins were analysed by immunoblotting using an anti-HA or anti-FLAG antibody (right). (B) Total cell lysate from either HEK293 (293) or HeLa cells was subjected to immunoprecipitation using the USPL1 antibody. Input corresponds to 1% of the total cell lysate (left) and a cell lysate sample of HEK293 cells transfected with either HA–USPL1 or FLAG–ELL was loaded onto the same gel as positive control for USPL1 or ELL, respectively. Empty Protein-G–agarose beads served as immunoprecipitation control. Bound proteins were analysed using either the USPL1 antibody or ELL antibody (right). The arrow indicates endogenous ELL, non-specific bands recognised by the ELL antibody are marked with asterisks. (C) Colocalisation of either ELL (top) or Ice1 (bottom) with USPL1 and coilin in Cajal bodies (arrows) shown by immunostaining of HeLa cells. Scale bar: 10 µm.

Fig. 7.

Knockdown of Ice1 has comparable consequences for nuclear architecture as USPL1 knockdown. (A) Immunofluorescence of HeLa cells transfected with siRNA as indicated were stained against USPL1, Ice1 and coilin. The arrowhead indicates a Cajal body in control cells with all three proteins present (top). Nucleolar coilin upon Ice1 or USPL1 knockdown is indicated by open arrowheads (middle and bottom, respectively). Arrows in the bottom panel highlight nuclear foci containing Ice1 in the absence of USPL1. (B) Immunofluorescence of siRNA treated U2OS cells against coilin and ASF. Arrows highlight enlarged, rounded nuclear speckles for ASF upon siIce1 (middle) or siUSPL1 (bottom) transfection. (C) Immunofluorescence of siRNA treated U2OS cells against coilin and SMN. A maximum-intensity projection illustrates the total number of SMN-containing nuclear foci (arrows in right panel). Scale bars, 10 µm.

Fig. 8.

USPL1 is associated with U snRNA gene loci by DNA-FISH and ChIP. (A) Specificity of the probe for the U2 gene array (Cy5, shown in green; arrows) demonstrated on a metaphase spread from normal human male lymphocytes. (B) HeLa cells were subjected to DNA-FISH for the U2 gene locus (Cy5, shown in red) in combination with immunostaining for USPL1 (Alexa Fluor 488, shown in green). Arrows indicate nuclear foci containing both USPL1 and the U2 gene locus. Scale bars: 10 µm. (C) ChIP from HeLa cells using the USPL1 antibody is displayed in comparison to IgG control for the respective gene locus (U1, U2 or GAPDH). Bars represent the s.e.m. of six independent experiments, each analysed as technical replicate of two in the qPCR-reaction. Statistical significance was determined using an unpaired, heteroscedastic Student's t-test. (D) RNAPII occupancy detected by ChIP against the CTD of RNAPII (4H8 antibody) in comparison to IgG control for the respective gene region (U1, U2, U2 −2 kb) upon USPL1 knockdown. Bars represent the s.e.m. of four independent experiments, each analysed as technical replicate of two in the qPCR reaction. Statistical significance was determined using an unpaired, heteroscedastic Student's t-test. *P<0.05.