|Home | About | Journals | Submit | Contact Us | Français|
Trypanosomatid parasites share a gene expression mode which differs greatly from that of their human and insect hosts. In these unicellular eukaryotes, protein-coding genes are transcribed polycistronically and individual mRNAs are processed from precursors by spliced leader (SL) trans splicing and polyadenylation. In trans splicing, the SL RNA is consumed through a transfer of its 5′-terminal part to the 5′ end of mRNAs. Since all mRNAs are trans spliced, the parasites depend on strong and continuous SL RNA synthesis mediated by RNA polymerase II. As essential factors for SL RNA gene transcription in Trypanosoma brucei, the general transcription factor (GTF) IIB and a complex, consisting of the TATA-binding protein-related protein 4, the small nuclear RNA-activating protein complex, and TFIIA, were recently identified. Although T. brucei TFIIA and TFIIB are extremely divergent to their counterparts in other eukaryotes, their characterization suggested that trypanosomatids do form a class II transcription preinitiation complex at the SL RNA gene promoter and harbor orthologues of other known GTFs. TFIIH is a GTF which functions in transcription initiation, DNA repair, and cell cycle control. Here, we investigated whether a T. brucei TFIIH is important for SL RNA gene transcription and found that silencing the expression of the highly conserved TFIIH subunit XPD in T. brucei affected SL RNA gene synthesis in vivo, and depletion of this protein from extract abolished SL RNA gene transcription in vitro. Since we also identified orthologues of the TFIIH subunits XPB, p52/TFB2, and p44/SSL1 copurifying with TbXPD, we concluded that the parasite harbors a TFIIH which is indispensable for SL RNA gene transcription.
Trypanosoma brucei, Trypanosoma cruzi, and Leishmania spp. are members of the eukaryotic family Trypanosomatidae, they cause devastating, often lethal diseases in humans, and they share a peculiar mode of gene expression which does not occur in their mammalian and insect hosts. In the genomes of these parasites, protein-coding genes are arranged in large tandem arrays which are transcribed polycistronically. Individual mRNAs are then processed from precursor RNA by spliced leader (SL) trans splicing and polyadenylation. A key molecule in this process is the SL RNA, because its capped 5′-terminal sequence, the SL or mini-exon, is trans spliced onto the 5′ end of each mRNA (reviewed in reference 20). Since unspliced mRNA is rapidly degraded, the parasites need to continuously synthesize large amounts of the SL RNA for the production of functional mRNA.
To accommodate the strong need for SL RNA, trypanosomatid cells harbor up to ~200 SL RNA gene copies, which are transcribed by RNA polymerase II (Pol II) in a monocistronic fashion (3, 7). The detailed characterization of the SL RNA gene promoter in three trypanosomatid species revealed a bipartite upstream sequence element (USE) (8, 14, 38) and an initiator (21). A focus of trypanosomatid SL RNA biology that has recently emerged is the characterization of the transcription machinery required for SL RNA synthesis. The initial discovery was the purification and partial characterization of three proteins which specifically associated with the active SL RNA gene template by binding to the USE (4). One of these proteins was identified as the orthologue of the human small nuclear RNA (snRNA)-activating protein 50 (SNAP50), which suggested that a SNAP complex (SNAPc)-like factor was important for SL RNA gene transcription (4). Furthermore, trypanosomatid genomes encode a single homologue of the TATA-binding protein (TBP) termed TBP-related factor 4 (TRF4). Silencing of the corresponding gene by conditional RNA interference (RNAi) in T. brucei clearly affected SL RNA gene expression in vivo, and chromatin immunoprecipitation demonstrated specific interaction of TRF4 with the SL RNA gene promoter (28). Subsequently, a combination of tandem affinity purification and in vitro transcription analysis established that SL RNA gene transcription essentially depends on a factor of six subunits that specifically binds to the SL RNA gene USE and comprises three SNAPc subunits, TRF4, the orthologue of the small subunit of the general transcription factor (GTF) TFIIA, and a highly divergent protein which might be the orthologue of the large TFIIA subunit (5, 31). Moreover, an extremely divergent member of the TFIIB family of proteins was recently discovered (25, 29; C. Tschudi, Yale University, unpublished data) and shown to be indispensable for SL RNA gene transcription. Although the amino acid sequence of this protein does not unambiguously identify it as TFIIB, its functional properties are in accordance with such an assignment: like TFIIB of yeast and humans (reviewed in reference 13), the T. brucei protein localized to the nucleus, specifically bound to a class II promoter (SL RNA gene), and interacted with both TRF4 and RNA Pol II (25, 29).
TFIIA, TFIIB, and the other class II GTFs TBP/TFIID, TFIIE, TFIIF, and TFIIH form the transcription preinitiation complex (PIC), recruit RNA Pol II to the correct transcription initiation site, separate the DNA strands, and mediate the polymerase's escape from the promoter (13). Although PIC formation is of fundamental importance to eukaryotes, identification of TFIIA and TFIIB in T. brucei was surprising because, except for the multifunctional TRF4 (28) and three putative TFIIH subunits, which in other organisms are known to function beyond PIC formation in DNA repair and cell cycle control, GTFs were not identified by the initial annotation of trypanosomatid genomes (15). The apparent lack of both GTFs and class II promoters for protein-coding gene transcription (22) as well as observations of seemingly nonspecific transcription initiation by RNA Pol II on chromosomal and extrachromosomal DNA (6, 37) fit well with the notion that RNA Pol II-mediated transcription of trypanosomatid protein-coding genes is not promoter driven (22, 34). Nevertheless, the finding that highly divergent TFIIA and TFIIB proteins are involved in SL RNA gene transcription suggested that trypanosomatids do form a PIC at the SL RNA gene promoter. If so, SL RNA gene transcription may depend on additional GTFs, such as TFIIH.
TFIIH consists of at least nine subunits and is recruited into the PIC after RNA Pol II. The factor is involved in DNA strand separation at the transcription initiation site and in phosphorylation of the carboxy-terminal domain of the largest RNA Pol II subunit, RPB1 (reviewed in reference 39). Accordingly, eukaryotic TFIIH consists of two subcomplexes: the core complex, which consists of four to five subunits that include the DNA helicase Xeroderma pigmentosum B (XPB) and p52, and the cyclin-activating kinase complex, which contains cyclin-dependent kinase 7. Both subcomplexes are linked by the DNA-RNA helicase XPD and its regulator, p44. Trypanosomatid genomes encode clear orthologues of XPB and XPD, whereas the claimed identification of cyclin-dependent kinase 7 by sequence homology (15) is ambiguous (12, 23).
Here, we show that silencing of XPD gene expression in T. brucei affects SL RNA synthesis in vivo and in vitro, that XPD in extract binds specifically to the SL RNA gene promoter, and that XPD is essential for SL RNA gene transcription in a crude cell extract. Moreover, the finding that the orthologues of the TFIIH subunits XPB, p44, and p52 copurified with XPD and that p52 localized to the nucleus confirmed that T. brucei harbors a TFIIH which is required for SL RNA gene transcription.
pTbXPD-stl is a derivative of the T. brucei stem-loop RNAi vector (1, 33) containing the TbXPD region from position −5 to position +483 relative to the translation initiation codon as inverted repeats around a stuffer sequence. pXPD-PTP-NEO was derived from pC-PTP-NEO (32) by fusing the C-terminal 532 bp of the TbXPD coding sequence to the PTP sequence via restriction sites ApaI and NotI. For C-terminal hemagglutinin (HA) tagging of the p52 orthologue, the vector p52-HA-BLA was generated. The first cassette of this vector consists of 925 bp of the C-terminal coding region fused to the HA sequence and the 3′ gene flank of TbRPA1. In the downstream selectable marker cassette, the blasticidin coding region is flanked 5′ and 3′ by the intergenic regions of heat shock protein 70 genes 2 and 3 and of β- and α-tubulin, respectively. Template DNAs SLins19 and GPEET-trm have been described previously (8, 19), but since both plasmids contained T7 promoters which were incompatible with extracts of T7 RNA Pol-expressing 29-13 cells, the template DNAs were recloned into vector pUC19 using restriction sites EcoRI and HindIII for SLins19-V.2 and KpnI and EcoRI for GPEET-trm-V.2.
Procyclic cell culture, targeted integration of linear DNAs into cells by electroporation, and the generation of stable cell lines by selection and limiting dilution were done as previously described (9, 10). The cell line TbX1 was generated by replacing one TbXPD allele with a PCR product of the hygromycin resistance gene fused to 150 bp of 5′ and 3′ TbXPD gene flank on either side and by inserting the SalI-linearized construct pXPD-PTP-NEO into the second TbXPD allele. For HA tagging of p52, the construct p52-HA-BLA was linearized with StuI and transfected into TbX1 cells. The selection conditions were 20 μg/ml of hygromycin, 40 μg/ml of G418, and 10 μg/ml of blasticidin. Correct integration of DNAs was confirmed by PCR with one primer positioned outside the cloned region.
For TbXPD silencing, T. brucei 29.13.6 cells (36) were transfected with 10 μg of the SacII-linearized construct pTbXPD-stl and cloned by limited dilution in the presence of 50 μg/ml of hygromycin, 15 μg/ml of G418, and 2.5 μg/ml of phleomycin. Silencing of TbXPD expression was induced by adding doxycycline to the culture medium at a final concentration of 10 μg/ml. Cells were counted daily and diluted to 3 × 106 cells/ml.
XPD and, as a control, U2-40K, expression in RNAi experiments was assayed by reverse transcription-PCR of total RNA prepared from induced and noninduced cells in the linear range of the amplification reaction. Total RNA was prepared from 108 cells using the RNeasy Mini kit (QIAGEN, Valencia, CA) according to the manufacturer's protocol. Primer extension reactions for the determination of the relative abundances of SL RNA and U2 snRNA were carried out with the same RNA preparations, the 5′-32P-end-labeled oligonucleotides SLf and U2f (11), and SuperScript II reverse transcriptase (Invitrogen, Carlsbad, CA) according to the manufacturer's protocol. For nascent RNA labeling, cells were permeabilized with the detergent lysolecithin as detailed elsewhere (35). Cells were then mixed into a transcription cocktail containing 20 mM potassium l-glutamate, 20 mM HEPES-KOH pH 7.7, 3 mM MgCl2, 1 mM dithiothreitol, 25 mM creatine phosphate, 0.6 mg/ml creatine kinase, 10 μg/ml leupeptin, 10 μg/ml aprotinin, 2 mM ATP, 1 mM GTP, 1 mM CTP, 5 μM UTP, and 100 μCi of [α-32P]UTP (3,000 Ci/mmol) and incubated for 8 min at 28°C. A small aliquot of total RNA prepared from these cells was standardized according to the total number of counts, separated on 6% polyacrylamide-50% urea gels, and visualized by autoradiography.
Indirect immunofluorescence was carried out as described previously (29) using a rat monoclonal anti-HA antibody at a 1:50 dilution and a secondary anti-rat immunoglobulin G (IgG) antibody conjugated to fluorescein isothiocyanate at a 1:250 dilution. For immunoblot analyses, PTP-, ProtC- or HA-tagged proteins were separated on sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE) gels, electroblotted onto a polyvinylidene difluoride membrane, and detected by the PAP reagent (Invitrogen), with the anti-protein C antibody HPC4 (Roche, Indianapolis, IN) or a peroxidase-labeled rat anti-HA antibody (Roche) in combination with the BM chemiluminescence blotting substrate (Roche) according to the manufacturer's protocol. Immunoprecipitations were essentially carried out as described previously (24). In brief, 100 μl of crude extract was incubated with either 40 μl settled volume of IgG beads (Amersham, Piscataway, NJ) or 25 μl settled volume of anti-HA antibody (Roche) immobilized on protein G-Sepharose (Amersham) and incubated on ice for 60 min. After washing the beads six times with 700 μl of TET100 buffer (100 mM NaCl, 20 mM Tris-HCl, pH 8.0, 3 mM MgCl2, 0.05% Tween 20), PTP-tagged protein was eluted by AcTEV protease (Invitrogen), whereas HA-tagged protein was directly eluted into SDS loading buffer at 85°C for 10 min. Immunoprecipitates were analyzed by immunoblotting as described above.
TbXPD-PTP was tandem affinity purified exactly as described previously (32). Purified proteins were separated on a 10 to 20% SDS-polyacrylamide gradient gel and stained with Gelcode Coomassie stain (Pierce, Rockford, IL). Protein bands were excised and analyzed by liquid chromatography-tandem mass spectrometry.
Promoter pull-downs were carried out as described previously (30) except that the salt concentration in the binding reaction mixture and in the washing buffers was increased to 80 mM potassium chloride. The in vitro transcription system, including the preparation of the crude transcription extract, has been described in detail elsewhere (18, 19). Briefly, 40-μl reaction mixtures were incubated for 1 h at 27°C; they contained 8 μl of extract, 20 mM potassium l-glutamate, 20 mM KCl, 3 mM MgCl2, 20 mM HEPES-KOH, pH 7.7, 0.5 mM of each nucleoside triphosphate, 20 mM creatine phosphate, 0.48 mg/ml of creatine kinase, 2.5% polyethylene glycol, 0.2 mM EDTA, 0.5 mM EGTA, 4 mM dithiothreitol, 10 mg/ml leupeptin, 10 mg/ml aprotinin, 12.5 μg/ml vector DNA, 20 μg/ml GPEET-trm template, and 7.5 μg/ml SLins19 template. For transcription reactions with 29-13 extract, the templates GPEET-trm-V.2 and SLins19-V.2 were used. GPEET-trm and SLins19 transcripts were specifically detected by extension of 32P-end-labeled primers Tag_PE (19) and SLtag (8) as described above. XPD-PTP-depleted transcription extract was prepared from the flowthrough of the IgG chromatography, the first step of XPD-PTP purification. Transcription extract from noninduced and induced 29-13 cells was prepared from 1 liter of cell culture according to a standard protocol (19).
The aim of this study was to assess whether T. brucei harbors TFIIH and whether this factor is involved in SL RNA gene transcription. Of the 10 known TFIIH subunits, homologous sequences of only two subunits, the helicases XPD and XPB, have been unambiguously identified (2). XPD is encoded by a single gene (GeneDB accession number Tb927.8.5980), whereas there are two genes which encode putative homologues of XPB: the predicted proteins of genes Tb927.3.5100 and Tb11.01.7950 have calculated masses of 105 kDa and 89 kDa, respectively, and are highly divergent to each other and to XPBs of other organisms (data not shown), the significance of which is not understood. We therefore chose to begin our analysis with TbXPD. At first, we employed the conditional RNAi system to silence the expression of TbXPD in procyclic trypanosomes. In this system, expression of double-stranded TbXPD RNA was induced by adding doxycycline to the medium. While induced cells grew nearly normal up to 72 h, they subsequently ceased growth and died rapidly (Fig. (Fig.1A).1A). Since we observed this effect with three independently generated cell lines and since the induction of double-stranded RNA (dsRNA) synthesis clearly affected the abundance of XPD RNA (Fig. (Fig.1B),1B), we concluded that TbXPD is encoded by an essential gene.
Next, we used a primer extension assay to compare the amounts of SL RNA and U2 snRNA in total RNA of cells in which XPD expression was silenced. In T. brucei, U2 snRNA synthesis is mediated by RNA Pol III and, therefore, should not depend on XPD/TFIIH. Accordingly, XPD silencing for 48 and 72 hours decreased the abundance of SL RNA to 17% and 13%, respectively, whereas U2 snRNA amounts were not significantly affected (Fig. (Fig.1C).1C). Primer extension of SL RNA typically results in two extension products because, in contrast to newly synthesized SL RNA, mature SL RNA has a unique cap4 structure with unusual methylations of the first four nucleotides which cause premature termination of reverse transcription. Interestingly, the larger extension product detecting unmethylated SL RNA was more strongly reduced than the product of the mature form, suggesting that XPD is involved in SL RNA synthesis rather than in maturation. To more directly analyze the effect of XPD silencing on SL RNA gene transcription, we employed a permeabilized cell system to radiolabel nascent RNA (35). Since the SL RNA is synthesized at a very high rate, it can be directly visualized by autoradiography, between high-molecular-mass pre-rRNA/pre-mRNA and tRNA, when total RNA is separated by denaturing PAGE (Fig. (Fig.1D)1D) (8, 35). Labeling of the SL RNA was reproducibly decreased by approximately 60% in cells in which XPD RNAi was induced for 48 or 72 h. While this effect seems moderate, it was comparable to the reduction caused by RNAi of the essential factors TRF4 (28) and TFIIB (29), raising the possibility that reducing SL RNA synthesis to this extent is sufficient to inhibit in vitro growth of procyclic trypanosomes. Our finding that steady-state levels of SL RNA were more strongly affected (−87%) than the synthesis rate of nascent SL RNA (−60%) is in accordance with this hypothesis and suggests that SL RNA turnover by trans splicing is greater than the rate of SL RNA synthesis in cells in which XPD expression was silenced for a minimum of 48 hours. Together, these experiments indicated that XPD plays a role in SL RNA gene transcription.
As a tool for TFIIH characterization and functional analysis of XPD in vitro, we generated the clonal cell line TbX1, which exclusively expressed XPD C-terminally fused to the PTP tag (Fig. (Fig.2A).2A). The latter is a combination of protein A and protein C (ProtC) epitopes separated by a tobacco etch virus (TEV) protease cleavage site and designed for tandem affinity purification of a tagged protein (32). Since we had shown that TbXPD is essential for trypanosome growth and since TbX1 cell growth was not significantly altered when compared to wild-type cells (data not shown), we inferred that the PTP tag did not critically interfere with the function of TbXPD. We therefore prepared a crude cell extract from TbX1 cells which consisted of a mix of cytoplasmic and extracted nuclear components and PTP-purified XPD sequentially by IgG affinity chromatography, TEV protease elution, and anti-ProtC immunoaffinity chromatography. The protein was finally eluted with either EGTA or the ProtC peptide. Monitoring of the purification by immunoblotting showed that while IgG bound nearly all of the XPD-PTP in the extract, binding of the tagged protein to the anti-ProtC matrix was less efficient (Fig. (Fig.2B,2B, compare lanes 1 and 3 with lanes 2 and 4). Due to the latter step, the overall purification efficiency of XPD-PTP was lower than that of previously reported PTP-tagged proteins (24, 26, 31, 32). Nevertheless, SDS-PAGE and Coomassie staining of the final eluate resulted in the detection of 10 distinct protein bands (Fig. (Fig.2C).2C). Liquid chromatography-tandem mass spectrometry analysis revealed four clear orthologues of TFIIH subunits in the final eluate thus far. The top band was identified as the large XPB protein encoded by gene Tb927.3.5100 and the strong band below as XPD-P.
In addition, analysis of the band with an apparent size of 36 kDa identified a protein which is encoded by gene Tb927.8.6540. The predicted size of this protein is 38 kDa, and it has orthologues in the genomes of Trypanosoma cruzi and Leishmania major (Fig. (Fig.3A).3A). A sequence alignment unambiguously identified this protein as the orthologue of the human TFIIH subunit p44 and the yeast counterpart SSL1p (Fig. (Fig.3A).3A). The C-terminal domain of this subunit is important for transcription initiation and, as a key feature, harbors a zinc-binding C4C4 RING structure (17). Accordingly, all eight C residues of this structure are conserved in trypanosomatids. The central domain of p44 is involved in the formation of the TFIIH core complex harboring an additional zinc-binding motif which is also perfectly conserved in the trypanosomatid sequences (Fig. (Fig.3A3A).
Furthermore, we identified a protein in the doublet band of 53/55 kDa which is encoded by gene Tb10.70.1900. While this protein was annotated as conserved hypothetical, it exhibited homology to the TFIIH subunit p52, which is termed TFB2 in yeast. p52 binds directly to the helicase XPB, and a mutational analysis showed that its C-terminal domain is involved in both transcription and DNA repair anchoring XPB within the TFIIH core complex (16). Accordingly, a sequence comparison between p52/TFB2 of model organisms and its putative trypanosomatid orthologues revealed the highest degree of conservation in the C terminus (Fig. (Fig.3B3B and data not shown).
Overall, the putative p52 orthologue was less conserved than p44, and it did not contain a clear sequence motif. To unequivocally show that the correct protein was identified, we C-terminally HA-tagged T. brucei p52 in TbX1 cells, prepared extract from the new cell line, and conducted reciprocal coimmunoprecipitation assays. Precipitation of XPD-PTP by IgG-coated beads specifically coprecipitated p52-HA and, vice versa, the use of an anti-HA antibody precipitated both p52-HA and XPD-PTP (Fig. (Fig.4).4). This interaction was specific, because the U2 snRNP protein U2-40K, which is predominantly localized in the nucleus (27), did not coprecipitate with either protein. Since we can also exclude a nonspecific interaction between HA and the PTP tag (24), these results clearly confirmed the interaction of XPD and p52.
Finally, we have achieved preliminary identification of five additional XPD copurified proteins (data not shown). All of them have been annotated as conserved hypothetical proteins and, by primary structure, appear to be parasite-specific proteins. We are in the process of confirming these identifications and evaluating the functional significance of each of them.
In summary, PTP purification of TbXPD revealed a minimum of nine copurified proteins, three of which were unambiguously identified as orthologues of TFIIH subunits. This strongly indicated that trypanosomes possess a transcriptionally relevant TFIIH.
The HA fusion to p52 enabled us to localize this protein in the cell. Thus far, we have been unable to utilize the protein A and C epitopes of the PTP tag to localize tagged proteins by indirect immunofluorescence. In contrast, the HA epitope has been successfully used for localization studies in trypanosomes. Accordingly and as expected, the use of the monoclonal rat anti-HA antibody specifically stained the nucleus of TbX1 cells which expressed p52-HA (Fig. (Fig.5),5), whereas there was no detectable signal with TbX1 cells not expressing an HA tag (data not shown). This control also demonstrated that the protein A domains of XPD-PTP, expressed in TbX1 cells, did not interact with the antibody at a detectable level. Hence, the nuclear localization of p52 further supported the correct identification of this protein as a subunit of T. brucei TFIIH.
Since silencing of TbXPD expression affected SL RNA synthesis in vivo, we wanted to confirm a function of TbXPD in SL RNA gene transcription first by analyzing its interaction with the SL RNA gene promoter. We have previously employed a promoter pull-down assay to demonstrate the specific interaction of all subunits of TRF4/SNAPc/TFIIA and of TFIIB with the SL RNA gene promoter (29, 31). In the same assay, XPD-PTP clearly bound to the SL RNA gene promoter but not to nonspecific DNA or the class I GPEET procyclin gene promoter (Fig. (Fig.6,6, compare lanes 3, 4, and 6). We also included in our analysis an SL RNA gene promoter DNA in which the first motif of the USE was mutated. This mutation abolished binding of the TRF4/SNAPc/TFIIA complex to the SL RNA gene promoter (Fig. (Fig.6)6) (31). In contrast, XPD-PTP bound more efficiently to the mutant than to the wild-type promoter, indicating that TFIIH binding is independent of TRF4/SNAPc/TFIIA (Fig. (Fig.6,6, compare lanes 4 and 5). Furthermore, since the PIC dissociates after transcription initiation (13) and our binding reactions may support transcription initiation, the inability of the mutant SL RNA gene promoter to direct transcription initiation (8) may have stabilized the interaction with XPD-PTP. Together, these results established that, in extract, XPD-PTP is specifically recruited to the SL RNA gene promoter.
In a final step, we analyzed the function of TbXPD in an in vitro transcription system. This system is based on a crude transcription extract which consists of cytoplasmic and extracted nuclear components and which is active for both RNA Pol I-mediated and class II SL RNA gene transcription (19). T. brucei utilizes RNA Pol I not only for transcription of the large rRNA genes as other eukaryotes, but also for those gene units which encode the major cell surface antigens variant surface glycoprotein and procyclin (10). Although all class I promoters function in the in vitro system, the GPEET procyclin gene promoter is most effective (19). Hence, we cotranscribed the templates GPEET-trm and SLins19, which harbor the GPEET gene/promoter and SL RNA gene promoter, respectively. Both templates carry unrelated tag sequences downstream of the transcription initiation site, enabling the specific detection of the GPEET-trm and SLins19 RNAs by primer extension assays. To assess the function of XPD for SL RNA gene transcription, we first prepared transcription extract from TbX1 cells which exclusively express XPD-PTP. Subsequently, we depleted XPD from this extract by IgG chromatography (Fig. (Fig.7A).7A). In comparison to mock-treated extract, XPD depletion virtually abolished transcription of template SLins19, whereas it did not affect GPEET-trm transcription (Fig. (Fig.7B).7B). The transcriptional activity of SLins19 could be partially restored in a dose-dependent manner when the final peptide eluate of the XPD-PTP tandem affinity purification was added back (Fig. (Fig.7B,7B, +Elu). This was not a nonspecific effect of the tag, because we have previously shown that PTP-tagged TbSNAP2 and TbRPA1 proteins are unable to stimulate SL RNA gene transcription in TFIIB-depleted extract (29). Hence, these results demonstrated that the observed effects on SL RNA gene transcription were caused by the XPD-PTP purified proteins. They did not, however, unambiguously identify XPD as an essential protein for this activity. To show this, we prepared transcription extract from an XPD-RNAi competent cell line before and 72 h after inducing synthesis of XPD dsRNA. In comparison to the previous experiments, the relative activities of GPEET-trm and SLins19 transcription were different (Fig. 7B and C). This variability was most likely a result of large- and small-scale extract preparations for the depletion and RNAi experiments, respectively. Nevertheless, the signal strengths within each experimental set were consistent. Hence, the results clearly showed that SLins19 transcription was strongly affected by XPD silencing, whereas GPEET-trm transcription was not (Fig. (Fig.7C).7C). Since the final peptide XPD-PTP eluate fully reconstituted the transcriptional activity, again in a dose-dependent manner, a nonspecific RNAi effect could be excluded. Thus, we concluded that XPD is essential for SL RNA gene transcription in T. brucei.
In this study, we have partially characterized TFIIH of T. brucei and shown that the TFIIH subunit XPD is essential for transcription of the SL RNA gene. After the identification of TFIIA and TFIIB, this is the third trypanosome GTF playing a crucial role in SL RNA gene transcription. Together, these findings strongly argue that trypanosomes form a conventional class II PIC at the SL RNA gene promoter. If so, it can be expected that the parasites also harbor orthologues of the missing GTFs, TFIIE and TFIIF. The surprising aspect of these findings is the extreme divergence in primary structure between mammalian and trypanosomatid GTFs which, in the case of TFIIB, exceeds the divergence between the mammalian and archaeal orthologues (29). The observed divergence of trypanosomatid sequences may be a result of relaxed evolutionary constraints due to relatively few PIC formation sites in the genome. The extraordinary divergence level offers the promising perspective that trypanosomatids have evolved unique and essential structural and/or biochemical features in a fundamentally important process, namely, the recruitment of RNA Pol II to DNA.
We are grateful to Mary Ann Gawinowicz (Columbia University, Protein Core Facility) for excellent mass spectrometry, Albrecht Bindereif for providing us with the anti-U2-40K antiserum, and Jens Brandenburg for his help in the indirect immunofluorescence experiments.
This work was supported by National Institutes of Health grant AI059377 to A.G. J.H.L was supported by a grant of the Korea Science and Engineering Foundation (no. C00093).
Published ahead of print on 26 January 2007.