A TFIIH-Associated Mediator Head Is a Basal Factor of Small Nuclear Spliced Leader RNA Gene Transcription in Early-Diverged Trypanosomes

Plasmid constructs.For PTP tagging of Med-T5, 232 bp of the 3′-terminal coding sequence was inserted into the pC-PTP-NEO vector (41) utilizing the vector's ApaI and NotI restriction sites. For Med-T5 and Med-T4 silencing, the coding regions from position 3 to position 420 and from position 7 to position 506, respectively, were integrated into the pT7-stl construct (3). For C-terminal PTP tagging of SNAP2, the coding region of neomycin phosphotransferase in plasmid SNAP2-PTP-NEO (40) was replaced by the coding region of the blasticidin S deaminase. The resulting vector was termed pSNAP2-PTP-BLA.

Cells.All experiments were conducted with procyclic cells of the T. brucei brucei 427 strain which were cultured, transfected, and cloned as previously described (25). RNA interference (RNAi) cells were induced with 2 μg/ml doxycycline, and cells were counted and diluted to 2 × 10⁶ cells/ml daily. The cell line exclusively expressing Med-T5-PTP was generated by inserting the StuI-linearized construct pMed-T5-PTP-NEO into one Med-T5 allele and by knocking out the second allele with a PCR product in which 102 bp of Med-T5 5′ and 3′ gene flanks were fused to the hygromycin phosphotransferase-coding region. Correct DNA integration was confirmed by PCRs in which one oligonucleotide hybridized outside the cloned region. Transfected cells were selected and grown in the presence of 40 μg/ml G418 and 20 μg/ml hygromycin.

RNA analysis.Analysis of RNA prepared from cells before and after Med-T5 silencing by semiquantitative reverse transcription-PCR (RT-PCR), primer extension, and nascent RNA labeling using permeabilized cells was conducted exactly as described previously (28).

Protein analysis.Extract preparation and tandem affinity purification of Med-T5 were carried out according to the standard PTP purification protocol (41). Purified proteins were separated on a 10 to 20% SDS-polyacrylamide gradient gel and stained with Pierce Gelcode Coomassie stain (Pierce). Sedimentation of XPD-P complexes in 4-ml 10 to 40% linear sucrose gradients was carried out as described before (27), whereas for sedimentation of Med-T5-P complexes, ultracentrifugation at 200,000 × g was shortened from 19 to 16 h. After gradient fractionation, Med-T and TFIIH proteins were separated on 10 to 20% SDS-polyacrylamide gradient gels and stained with Sypro Ruby (Invitrogen) according to the manufacturer's specifications. Proteins were identified by protein band excision, trypsin digestion, liquid chromatography-tandem mass spectrometry, and data analysis using the TurboSEQUEST program of the BioworksBrowser 3.1 software package (Thermo Electron). Peptide identifications were considered relevant only when the PeptideProphet probability values were ≥0.9.

In vitro transcription analysis.Transcription assays were performed as described previously (24, 25). Newly synthesized RNAs of the GPEET-trm and SLins19 templates were detected by primer extension of ³²P-end-labeled oligonucleotides Tag-PE and SLtag, which hybridize to unrelated oligonucleotide tags of the GPEET-trm and SLins19 RNAs. Primer extension products were resolved on 6% polyacrylamide-50% urea gels and visualized by autoradiography.

Anti-Med-T5 immune serum.Med-T5 was expressed with an N-terminal glutathione S-transferase (GST) tag in Escherichia coli, purified by glutathione affinity chromatography, and used as a GST fusion protein for immunizing rats according to a previously published protocol (39).

Light microscopy.For immunolocalization, Med-T5-PTP was detected with a rabbit polyclonal anti-protein A immune serum (Sigma) followed by an Alexa 594-conjugated anti-rabbit secondary antibody (Invitrogen), and the hemagglutinin (HA)-tagged p52 ortholog was detected with a mouse monoclonal anti-HA antibody conjugated with fluorescein isothiocyanate (FITC) (Sigma). Images were acquired on a BX41 epifluorescence microscope (Olympus) equipped with DAPI (4′,6′-diamidino-2-phenylindole), FITC, and rhodamine filters; a 100× (1.4-numerical-aperture) oil immersion objective; and a Retiga EXi charge-coupled device (CCD) camera (Q Imaging). All images were acquired at a resolution of 696 by 520 pixels.

EM.For EM characterization of purified Med-T, TFIIH, and Med-T/TFIIH samples, 3 μl of protein solutions was applied to a carbon-coated Maxtaform, 300-mesh Cu/Rh EM specimen grid (Ted Pella, Inc., Redding, CA) and freshly glow discharged in the presence of amyl amine. Particles were preserved by staining with a 2.0% (wt/wt) uranyl acetate solution using the sandwich carbon layer technique (45). The samples were imaged under low-dose conditions using a Tecnai F20 microscope (Philips/FEI) equipped with a field emission gun and operating at an accelerating voltage of 120 kV. Tilted (55°)/untilted image pairs were recorded on Kodak SO163 film at a magnification of ×50,000 and with underfocus values of between −0.8 and −1.1 μm. Micrographs were digitized on a Nikon Coolscan 9000 using a 6.35-μm sampling step size. Digitized images were 4-fold pixel averaged, resulting in final sampling of 5.08 Å pixel⁻¹.

Image processing.For Med-T, tilt-pair images were selected from digitized micrograph pairs using the TiltPicker program (51), high- and low-pass filtered, normalized, boxed, and decimated to a size of 70 by 70 pixels. Med-T, TFIIH, and Med-T/TFIIH images were initially analyzed using the ml_align2d program, a multireference two-dimensional (2D) alignment routine with a maximum-likelihood target function (38) implemented in the Xmipp package (44). Averages were utilized to run iterative alternating rounds of supervised multireference alignment/classification and reference-free alignment using the SPIDER software package to improve the homogeneity of the image classes (4). Corresponding tilted images were used to 3D reconstruct Med-T by the random conical tilt method (36). In immunofluorescence microscopy, colocalization was quantified using both the Colocalization Finder and JoACop plugins of ImageJ (http://rsb.info.nih.gov/ij/ ).

ChIP analysis.For the chromatin immunoprecipitation (ChIP) assays, 0.5 × 10⁸ procyclic cells were fixed in 1% formaldehyde and lysed according to a published protocol (29). Chromatin was fragmented by 25 sonication cycles (30 s on/30 s off) using a Bioruptor (Diagenode Inc.) at high setting. Chromatin fragments were collected by centrifugation for 10 min at 25,000 × g and 4°C and immunoprecipitated overnight at 4°C with protein G-bound rat polyclonal antibodies directed against Med-T5, TFIIB (39), or TSP2 (27) or with a protein A-bound rabbit polyclonal antibody (Sigma) directed against the protein A domains of the PTP tag. Beads were washed once with buffer 1 (50 mM HEPES-KOH [pH 8.0], 1 mM EDTA, 0.1% Na-deoxycholate, 1% Triton X-100, 0.1% SDS, 10 μg/ml aprotinin, and 10 μg/ml leupeptin) containing 275 mM NaCl and three times with buffer 1 containing 500 mM NaCl. Following transfer to a new tube, beads were washed once more with buffer 2 (10 mM Tris-HCl [pH 8.0], 250 mM LiCl, 1 mM EDTA, 0.5% NP-40, and 0.5% Na-deoxycholate) and equilibrated in Tris-EDTA (TE) buffer. Chromatin was eluted from beads during an RNase digest in 10 mM Tris-HCl (pH 8.0), 1 mM EDTA, and 300 mM NaCl for 30 min at 65°C, and a further incubation for 10 min in 55 mM Tris-HCl (pH 8.0), 10 mM EDTA, 150 mM NaCl, and 1% SDS. After de-cross-linking with proteinase K overnight at 52°C, immunoprecipitated DNA was analyzed by semiquantitative and quantitative real-time PCRs of the SLRNA promoter region (positions −90 to +63 relative to the TIS), the SLRNA intergenic region (+407 to +540), the GPEET procyclin promoter region (−144 to +6), and the α-tubulin-coding region (+737 to +860 relative to the translation initiation codon) using, respectively, the primer pairs 5′-CTACCGACACATTTCTGGC-3′/5′-GGTATGAGAAGCTCCCAGTAGCAGC-3′, 5′-ATGGCTTATACGTGCTCGTTTCTCC-3′/5′-CACATATAGGCGCTTTAAAGTCTGCT-3′, 5′-TGATAGGTATCTCTTATTAGTATAG-3′/5′-GGGGTTATCGGGTGAGTAC-3′, and 5′-GTGCATTGAACGTGGATCTG-3′/5′-GCCTACCACGAGCAACTCTC-3′.

Trypanosomes harbor a TFIIH-associated complex of nine subunits.Previously, we tandem affinity purified TFIIH from trypanosome extract and showed that this protein complex fully retained its functionality in RNA pol II-mediated transcription from the SLRNA promoter (27). Biochemical and structural analyses of the purified complex revealed a complete core complex but not a CAK subcomplex (27). However, due to the multifunctional nature of TFIIH, it was possible that only a minor subset of the purified TFIIH represented the transcriptionally active form associated with a kinase complex. Accordingly, when we strongly overexposed Sypro Ruby-stained proteins which copurified and, importantly, cosedimented in a sucrose gradient with tagged XPD, we detected new protein bands of minor abundance with apparent molecular masses of between 12.5 and 27 kDa. These proteins exhibited a common sedimentation peak in fraction 18, one fraction further down than the TFIIH peak (Fig. 1A). Presumably, the minor proteins formed a TFIIH-associated complex which caused the large TFIIH complex to sediment even faster. Since the proteins were of low abundance, we were able to identify only one by tandem mass spectrometry in the 17-kDa band. This protein is encoded by gene Tb11.03.0430 (genome data are accessible through http://www.GeneDB.org ) and was termed Med-T5. The amino acid sequence of Med-T5 is well conserved among kinetoplastid organisms but is not similar to that of any other protein; it also does not harbor a sequence motif of known function (see Fig. S1 in the supplemental material). To further study this protein, we generated a clonal procyclic cell line which exclusively expressed the protein as a C-terminal fusion to the composite PTP tag and raised a specific polyclonal immune serum against the purified, recombinantly expressed protein (Fig. 1B and C). Coimmunoprecipitation confirmed that Med-T5 specifically interacted with XPD in extract (Fig. 1D). Since the sedimentation analysis suggested that Med-T5 is a subunit of a larger complex, we tandem affinity purified Med-T5 and detected nine protein bands of major abundance and nine bands of minor abundance (Fig. 1E). The latter group corresponded to the nine previously characterized subunits of trypanosome TFIIH, whereas the major bands consisted exclusively of proteins which had been annotated as “conserved hypothetical,” meaning that the genome-derived protein sequences were found to be conserved only among kinetoplastid organisms (see Fig. S2 to S9 and Table S10 in the supplemental material). Accordingly, our own bioinformatic analyses could not detect conserved amino acid motifs or sequence similarities to known proteins, including subunits of the mediator complex (data not shown). We termed the subunits Med-T1 to Med-T9 according to their apparent sizes and to discriminate them from characterized mediator subunits in other systems (Table 1). Interestingly, two of the subunits, Med-T2 and Med-T4, appear to be restricted to the genus Trypanosoma, because they were not found in the genomes of three Leishmania species, which belong to the same phylogenetic order (see Table S10 in the supplemental material).

• Open in new tab
• Download powerpoint

FIG. 1.

Biochemical characterization of the Med-T complex. (A) Overexposure of Sypro Ruby-stained proteins that copurified with tagged XPD and cosedimented in the bottom fractions of a 10 to 40% sucrose gradient. Newly detected proteins of minor abundance are marked by white asterisks on both sides of their peak fraction 18. TFIIH subunits and protein marker sizes are indicated on the left and right, respectively. (B) Schematic depiction of the Med-T5 locus in a clonal cell line exclusively expressing Med-T5-PTP. One Med-T5 allele was replaced by the hygromycin phosphotransferase gene, and the other allele was altered by targeted insertion of the plasmid Med-T5-PTP-NEO. The Med-T5 coding region, PTP tag sequence, and hygromycin (HYG-R) and neomycin (NEO-R) resistance marker coding regions, as well as introduced gene flanks for RNA processing signals, are represented by open boxes, a black box, striped boxes, and smaller gray boxes, respectively. (C) Anti-Med-T5 immunoblots of whole-cell lysates from wild-type cells and cells exclusively expressing Med-T5-PTP (EE). Blots were probed with preimmune sera (pre-IS) or immune sera (IS). Detection of the spliceosomal protein U2-40K served as a loading control. (D) Coimmunoprecipitation of Med-T5 with XPD. XPD-PTP was precipitated from extract with IgG beads and eluted through TEV protease cleavage (note the size reduction from XPD-PTP to XPD-P). XPD-P, Med-T5, and U2-40K were detected in input material (Inp), supernatant (SN), and immune precipitate (P). x values denote relative amounts of material analyzed. (E) The complete final eluate (Elu 100%) of a standard Med-T5-PTP purification was separated on a 10 to 20% SDS-polyacrylamide gradient gel and stained with Coomassie blue. Arrows on the right identify Med-T subunits, whereas diamonds on the left indicate TFIIH subunits. For comparison, 0.002% of crude extract (Inp) and 5% of the TEV protease eluate (Elu TEV) were coanalyzed. (F) Sedimentation of Med-T5-PTP-purified material in a 10% to 40% sucrose gradient which was fractionated from top to bottom. In comparison to the sedimentation experiment results shown in panel A, ultracentrifugation was shortened by 3 h to better separate Med-T and Med-T/TFIIH complexes. Proteins from fractions 5 through 19 were separated by SDS-PAGE and stained with Sypro Ruby. Asterisks to the right of lanes 10 and 14 and diamonds to the left of lane 14 mark Med and TFIIH subunits, respectively. As markers with known S values, mouse IgG (6.6S), apoferritin (17S), and thyroglobulin (19S) were sedimented in parallel gradients and analyzed by measuring the optical densities of individual fractions at 280 nm.

A sucrose gradient sedimentation of the Med-T5-PTP-purified proteins showed that all nine newly identified proteins cosedimented with a peak in fraction 10 as an ∼7S particle, strongly indicating that they form a single protein complex (Fig. 1F). Moreover, approximately 10% of this new complex was associated with TFIIH in a combined complex sedimenting as an ∼17.5S particle, mainly in fractions 14 and 15 (Fig. 1F).

Overall, these findings strongly indicated that this TFIIH-associated complex is not the trypanosome cyclin-activating kinase (CAK) complex, because none of its subunits is a component of the well-characterized trypanosome kinome (35) or harbors one of the highly conserved kinase domains (data not shown). Moreover, none of the proteins possesses an N-terminal C3HC4 RING finger domain, which is the hallmark of the CAK subunit MAT1/TFB3 (mammalian/yeast nomenclature).

Med-T5 colocalizes with TFIIH predominantly in two perinucleolar foci. To colocalize Med-T5 and TFIIH by indirect immunofluorescence microscopy, we generated a clonal cell line in which Med-T5 and the p52 orthologous subunit of TFIIH were C-terminally tagged with PTP and HA, respectively. As expected, both proteins were localized in the nucleus, typically with the strongest accumulation in two perinucleolar foci (Fig. 2; see Fig. S11 in the supplemental material) which likely correspond to the SLRNA tandem repeat regions (50). Accordingly, both proteins colocalized in these two foci, whereas there was limited signal overlap in the remainder of the nucleus. Quantification of the fluorescence signals indicated an overlap of 32.65% (standard deviation of 8.95%); this result is in agreement with our biochemical characterizations (Fig. 1A and F), in which only a subfraction of the complexes associated with each other. This result also suggests that TFIIH and the new complex cooperate predominantly at SLRNA genes but have independent functions elsewhere in the nucleus. Another observation was that the TFIIH signal appeared to be present in the nucleolar area, which by DAPI staining is apparent as a spherical structure with reduced staining, whereas the Med-T signal always was excluded from this area. This finding supports a previous study in which the silencing of the TFIIH subunit gene XPD affected nucleolus-based RNA pol I-mediated transcription (26).

• Open in new tab
• Download powerpoint

FIG. 2.

Colocalization of the HA-tagged TFIIH p52 ortholog and of Med-T5-PTP by indirect immunofluorescence.

Med-T is essential for parasite viability and important for SLRNA transcription in vivo.Since TFIIH has an essential role in SLRNA transcription and since TFIIH and the new complex colocalized in the SLRNA foci, we analyzed whether the latter complex is a transcription factor. We introduced a T7 promoter construct for tetracycline-inducible expression of Med-T5 double-stranded RNA (dsRNA) into a genetically engineered procyclic cell line that expresses both the tetracycline repressor and T7 RNA polymerase (52). Three independently derived clonal cell lines were obtained. After induction of Med-T5 dsRNA synthesis by addition of doxycycline to the medium, we observed cessation of culture growth and cell death at 2 days after induction. A representative growth curve for one of the cell lines is shown in Fig. 3A. Semiquantitative RT-PCR analysis of RNA prepared from this cell line showed that Med-T5 mRNA was specifically degraded without an effect on mRNA levels of either the TFIIH subunit XPD or a spliceosomal control protein (Fig. 3B). This finding was corroborated by an immunoblot analysis which showed that Med-T5 abundance was reduced to less than 10% at 3 days after induction whereas levels of the TFIIH subunit TSP2 and of TFIIB remained constant (Fig. 3C). These results strongly indicated that Med-T5 expression was specifically silenced without affecting the abundance of the associated TFIIH complex.

• Open in new tab
• Download powerpoint

FIG. 3.

Med-T5 expression silencing is lethal and affects SL RNA abundance and synthesis. (A) Representative growth curves for one of three independently derived clonal cell lines in the absence and presence of doxycycline, which induces the expression of Med-T5 dsRNA. (B) Semiquantitative RT-PCR analysis of Med-T5, XPD, and U2-40K mRNA levels in total RNA of cells in which Med-T5 dsRNA synthesis was induced for the specified periods. (C) Corresponding immunoblot in which Med-T5, the TFIIH subunit TSP2, and TFIIB were detected with specific polyclonal immune sera. (D) SL RNA (SL)- and U2 snRNA (U2)-specific primer extension analysis of total RNA prepared from noninduced and doxycycline-induced cells. DNA marker bands are indicated on the left and extension products on the right. (E) Nascent RNA, radiolabeled in permeabilized induced cells, was separated by denaturing PAGE and visualized by autoradiography.

When we analyzed SL RNA abundance in total RNA prepared from Med-T5-silenced cells within 3 days of induction, we saw a clear decrease of SL RNA relative to the level of spliceosomal U2 snRNA, which in trypanosomes is synthesized by RNA pol III (Fig. 3D). Radiolabeling of nascent RNA using a cell permeabilization approach allows direct visualization of newly synthesized SL RNA by denaturing PAGE and autoradiography (49). Employing this method, we found that Med-T5 silencing affected SL RNA synthesis (Fig. 3E); in two independent experiments, the SL RNA signal relative to the tRNA signal was reduced on average to 64% and 23% on days 2 and 3, respectively. Taken together, these results strongly indicated that Med-T5 is essential for parasite viability and is involved in SLRNA transcription.

To confirm that the essential Med-T5 function accounts for the whole Med-T complex, we silenced the expression of the trypanosome-specific subunit Med-T4. As observed with Med-T5 silencing, three procyclic clonal cell lines ceased culture growth within 2 days of doxycycline induction of Med-T4 dsRNA synthesis. Figure 4A shows a representative growth curve of one of these cell lines. Importantly, the knockdown of Med-T4 mRNA did not affect the abundance of the Med-T5 mRNA or of the mRNA for the TFIIH subunit XPD, indicating that expression silencing was specific to Med-T4 (Fig. 4B). Finally, as with Med-T5, silencing of Med-T4 resulted in a specific decrease of SL RNA (Fig. 4C), strongly indicating that it is the Med-T complex that serves an essential function in SLRNA transcription.

• Open in new tab
• Download powerpoint

FIG. 4.

Expression silencing of trypanosome-specific Med-T4 is lethal and affects SL RNA abundance. (A) Representative growth curves for one of three independently derived clonal cell lines in the absence and presence of doxycycline, which induces the expression of Med-T4 dsRNA. (B) RT-PCR analysis of Med-T4, Med-T5, XPD, and U2-40K mRNA abundances in total RNA preparations of noninduced cells and of cells which were induced by doxycycline for 1, 2, or 3 days. (C) PCR and quantitative PCR (qPCR) analyses of SL RNA and U2 snRNA cDNAs derived with gene-specific primers from the same RNA preparations. Error bars indicate standard deviations.

Med-T5 is indispensable for SLRNA transcription in vitro.To further demonstrate the role of Med-T in SLRNA transcription, we employed a homologous in vitro transcription system and cotranscribed the SLRNA promoter construct SLins19 with a construct containing the GPEET procyclin promoter, which recruits RNA pol I (25). In a first set of experiments, we prepared extract from noninduced RNAi cells and from cells in which Med-T5 expression was silenced for 3 days. As expected, the silencing of Med-T5 strongly reduced SLRNA promoter transcription, whereas there was no major effect on GPEET promoter transcription (Fig. 5A, top panel, compare lanes 1 and 2). Adding back the final eluate of the Med-T5-PTP purification (Med-T5-P eluate) in amounts that restored the endogenous Med-T5 level (Fig. 5A, lower panel) resulted in specific and full reconstitution of SLRNA transcription. This finding clearly demonstrated that Med-T5 is important for transcription of SLRNA genes and confirmed that it is targeted specifically by the RNAi pathway in induced cells.

• Open in new tab
• Download powerpoint

FIG. 5.

In vitro transcription analysis of Med-T5 depletion and reconstitution. (A) Upper panel, in vitro transcription reactions were carried out with templates SLins19 and GPEET-trm (GPEET) and with extracts prepared from noninduced cells (no RNAi) and from cells in which Med-T5 expression was silenced for 72 h (RNAi). Depleted extract was reconstituted with purified Med-T5-P eluate (depl + Med-T5-P) at two different concentrations. Lower panel, anti-Med-T5 immunoblot analysis of extracts used in the transcription reactions. On the right, endogenous (end.) Med-T5 and TEV protease-cleaved Med-T5-P and a nonspecific band (non-sp) are identified. (B) Top panel, corresponding transcription reactions were carried out with extracts from cells which exclusively expressed Med-T5-PTP. The extract was either mock treated or depleted of Med-T5-PTP (depl) by IgG affinity chromatography. Depleted extract was reconstituted with eluates of PTP-purified TRF4, Med-T5, and XPD. Middle panel, extract depletion of Med-T5-PTP and reconstitution with Med-T5-P was monitored by anti-Med-T5 immunoblotting. Bottom panel, the same blot was probed with a polyclonal antibody directed against the TFIIH subunit TSP2 (α-TSP2).

Since we had generated a cell line which exclusively expressed Med-T5-PTP, we were able to efficiently deplete the tagged protein by IgG affinity chromatography from transcription extracts (Fig. 5B, middle panel, compare lanes 1 and 2). As expected, due to the interaction of Med-T5 with TFIIH, the immunoblot signal of the TFIIH subunit TSP2 was reduced as well, albeit to a much lesser extent (Fig. 5B, lower panel, compare lanes 1 and 2). As shown in the top panel of Fig. 5B, Med-T5 depletion nearly abolished SLRNA transcription, whereas there was no effect on transcription from the GPEET control template (compare lanes 1 and 2). Again, adding back Med-T5-P eluate restored SLRNA transcription in a dose-dependent manner (lanes 4 and 5). Importantly, this positive effect was not attributable to a nonspecifically copurified protein, because the PTP-purified, transcriptionally active TRF4 complex (40) was unable to restore transcription (compare lanes 2 and 3). Moreover, adding back PTP-purified, transcriptionally active TFIIH (27) resulted in only a modest reconstitution of SLRNA transcription (Fig. 4B, top panel, compare lanes 1 and 6), most likely because the Med-T5 level, in contrast to the TSP2 level, was restored only to a marginal extent (Fig. 5B, middle and bottom panels, compare lanes 1 and 6). These results confirmed that Med-T5 is important for SLRNA transcription.

The structure of the TFIIH-associated complex resembles that of the mediator head module.The importance of Med-T for SLRNA transcription in vivo and in vitro, the association of the new complex with TFIIH, and the multisubunit nature of the complex suggested that it might correspond to the trypanosome mediator (Med-T). Because the structure of mediator is essentially conserved from yeast to humans, we reasoned that a structural characterization of the Med-T and Med-T/TFIIH complexes by single-particle electron microscopy (EM) might confirm the identity of Med-T. As shown in Fig. 6A, single particles were well preserved and were homogeneous in size and overall shape. Iterative reference-free image alignment and classification generated three similar Med-T projection maps that revealed slight changes in the overall conformation and architecture of a smaller corner-shaped domain on the bottom-right side of the structure (Fig. 6B). Three-dimensional structures of each Med-T conformation were obtained at a ∼35-Å resolution (see Fig. S12 and S13 in the supplemental material) from images of tilted stained particles (Fig. 6C) using the random conical tilt method (36). The structural variability evident in projection maps was clearly reflected in the 3D structures (Fig. 6D). Med-T exhibited clear structural similarities to the head module of yeast mediator (yHead) (6). The structures of both complexes are reminiscent of a crescent wrench with a handle and jaws (Fig. 7A). Moreover, Med-T exhibited conformational variability similar to that of yHead, namely, changes in the positions of the “handle” and “jaws” regions as depicted in Fig. 7A (right panel).

• Open in new tab
• Download powerpoint

FIG. 6.

Single-particle EM analysis of Med-T. (A) A typical micrograph showing single Med-T particles preserved in uranyl acetate. The scale bar corresponds to 200 Å. (B) Three different conformations of Med-T were identified through reference-free alignment and classification of EM images. Med-T particles were nearly evenly distributed among the three different conformations, which differed mainly in the orientation of the elongated proximal end and the architecture of a smaller corner-shaped domain on the bottom-right side of the structure. Scale bar, 100 Å. (C) Different views of Med-T 3D reconstructions of the three conformations. Scale bar, 100 Å. (D) Superposition of Med-T conformer I (green), conformer II (red), and conformer III (blue) structures shows that the structural flexibility exists mainly in the “handle” and “jaws” domains.

• Open in new tab
• Download powerpoint

FIG. 7.

Trypanosome Med-T is structurally and functionally similar to yHead. (A) Comparison of class averages (left panel) and of 3D reconstructions (right panel) of yHead (6) and Med-T demonstrates the structural similarity between the two complexes. The location of the Med11-Med22-Med17 subunits in yHead is marked by the black dashed ellipse. Black arrows denote the domains showing striking structural flexibility. Scale bars, 100 Å. (B) 2D reconstructions of Med-T/TFIIH, of Med-T, and of TFIIH reveal the TFIIH and Med-T protein densities in the Med-T/TFIIH structure (left panel). In the right panel, a structural model of Med-T/TFIIH is shown, in which the red dashed ellipse marks the TFIIH-interacting Med-T region that corresponds to the Med11-Med22-Med17 subunits of yHead and the arrow points to the TFIIH/Med-T interface.

The combination of conformational flexibility and relatively small size (molecular mass of ∼270 kDa) makes Med-T a challenging target for higher-resolution single-particle EM analysis. Thus, to bolster the assertion that Med-T corresponds to yHead, we analyzed the interaction of Med-T and TFIIH. A stoichiometric and stable Med-T/TFIIH complex sample derived from a sucrose gradient fraction was analyzed by single-particle EM. In the Med-T/TFIIH structure (Fig. 7B, left panel), we could clearly identify Med-T density, as well as additional density along the tip of the jaw region. This non-Med-T density appears as a ring topped by a small protrusion, which is highly reminiscent of the low-resolution EM structure of human and yeast TFIIH (8, 42). To confirm the identity of the non-Med-T density, we determined the structure of free trypanosome TFIIH (Fig. 7B, bottom of middle panel). The structure of free TFIIH closely matches that of the presumed TFIIH portion of the Med-T/TFIIH 2D map, bolstering the identification of the trypanosome mediator and establishing the nature of the Med-T/TFIIH interaction: Med-T interacts with TFIIH along the tip of the Med-T jaws region (Fig. 7B, right panel), which in yHead corresponds to subunits Med11-Med22-Med17 (6). Since a direct interaction between Med11 of yHead and XPD/Rad3 of TFIIH recently was demonstrated (14), our structural analysis strongly indicated that Med-T and trypanosome TFIIH interact in the same manner. In summary, single-particle EM structural analysis demonstrated that Med-T and yHead have similar overall shapes, exhibit the same flexible regions, and use the same region to interact with TFIIH.

Med-T5 silencing disrupts PIC formation at the SLRNA promoter in vivo.Lastly, we wanted to confirm the specific function of Med-T in vivo. Employing anti-Med-T5 chromatin immunoprecipitation (ChIP) experiments, we confirmed that this protein is specifically cross-linked to the SLRNA promoter. While the SLRNA promoter region was readily detectable in precipitated DNA, the amplification signal of the SLRNA region, located 500 bp downstream of the TIS, was strongly reduced, whereas only background signals for the class I GPEET promoter and the coding region of α-tubulin were observed (Fig. 8A).

• Open in new tab
• Download powerpoint

FIG. 8.

Med-T5 binds to the SLRNA promoter and is indispensable for PIC integrity. (A) Anti-Med-T5 ChIP assays using a polyclonal immune serum (IS) or, in negative-control experiments, a nonspecific immune serum (nsp-IS) show that Med-T5 is efficiently cross-linked to the SLRNA promoter (SLRNA prom). Control PCR amplifications comprised an internal SLRNA region from position +407 to +540 downstream of the TIS (SLRNA int), the GPEET promoter (GPEET prom), and a fragment of the α-tubulin-coding region. Standard and quantitative real-time PCR analyses are shown in the upper and lower panels, respectively. Inp, input material. (B) ChIP assays were conducted with chromatin prepared from noninduced cells or from cells in which Med-T5 was silenced (Med-T5 RNAi) for 72 h. Chromatin was precipitated with anti-TSP2 or anti-TFIIB antisera (IS) or with a polyclonal antibody directed against the protein A domains of the PTP tag fused to the SNAP2 protein. The SLRNA promoter region was amplified. Note the different scales in the qPCR panel; the antitag antibody reproducibly led to a better enrichment than the antisera directed against the native proteins. (C) Corresponding ChIP assays were conducted as a control for general death phenotypes with chromatin prepared from noninduced cells or from cells in which LSm2, encoding an essential spliceosomal protein, was silenced for 48 or 72 h (LSm2 RNAi). Error bars indicate standard deviations.

In other systems, mediator was shown to be essential for TFIIH/TFIIE recruitment to the PIC in vivo (14), to assist TFIIB recruitment to the PIC (2), and to play a general role in PIC assembly and/or stabilization (6, 47). To analyze whether Med-T5 has corresponding functions, we conducted ChIP experiments to investigate in vivo binding of the trypanosome TFIIH subunit TSP2 and of TFIIB to the SLRNA promoter upon expression silencing of Med-T5. Anti-TSP2 and anti-TFIIB ChIP experiments with chromatin of noninduced cells confirmed previous results that both TFIIB and TFIIH bind to the SLRNA promoter (Fig. 8B) (39, 40). However, binding of both factors was lost when Med-T5 was silenced for 72 h. Importantly, the loss of TSP2 binding could not be explained by a concomitant loss of TFIIH in response to silencing a subunit of its partner complex Med-T, because we had shown for XPD on the RNA level (Fig. 3B) and for TSP2 on the protein level (Fig. 3C) that Med-T5 silencing did not affect the abundances of these TFIIH subunits. Moreover, to exclude the possibility that the defect in PIC formation at the SLRNA promoter was a general phenotype of dying trypanosomes rather than a specific effect of Med-T5 silencing, we analyzed cells in which the expression of the spliceosomal protein LSm2 was silenced. We previously have shown that interference with LSm2 expression halts trypanosome culture growth about 1 day earlier than does Med-T5 silencing (1). We therefore repeated the ChIP assays with cells in which LSm2 was silenced for 48 and 72 h. At the 48-h time point, which, according to the growth curve, was comparable to the 72-h time point of the Med-T5 silencing, both TSP2 and TFIIB remained clearly bound to the SLRNA promoter (Fig. 8C). Since this binding remained largely intact even after 72 h, when cell numbers already were in decline, the above-described PIC formation defects were specifically due to Med-T5 silencing and not to pathways of cell death.

The only other basal transcription factor known in trypanosomes is the SNAPc/TRF4/TFIIA complex which binds through its SNAP2 subunit directly to the upstream element of the SLRNA promoter (11). To determine whether Med-T5 silencing affected binding of SNAP2, we C-terminally PTP tagged the factor in Med-T5 and LSm2 RNAi cells and subjected it to ChIP experiments with a ChIP-grade polyclonal anti-protein A antibody. These experiments clearly showed that, in contrast to the case for TFIIB and TFIIH, SNAP2 binding to the SLRNA promoter was not affected by Med-T5 silencing (Fig. 8B and C).

Consequently, we concluded that Med-T, the mediator head complex of trypanosomes, functions analogously to human and yeast mediators in basal transcription initiation: it is indispensable for PIC formation or PIC integrity at the SLRNA promoter in vivo and is specifically required for recruitment of TFIIB and TFIIH. Therefore, Med-T is a basal factor for transcription of this crucially important small nuclear RNA gene.