Maf1 Is a Negative Regulator of Transcription in Trypanosoma Brucei

Molecular Microbiology (2017) 103(3), 452–468 ᭿ doi:10.1111/mmi.13568 First published online 25 November 2016 Maf1 is a negative regulator of transcription in Trypanosoma brucei

Gabriela Romero-Meza,1,2 Interestingly, 5S rRNA levels do not change after Daniel E. Velez-Ram ırez,1 TbMaf1 ablation or overexpression. Notably, our data Luis E. Florencio-Martınez,1 also revealed that TbMaf1 regulates Pol I transcrip- Fiordaliso C. Roman-Carraro, 1 tion of procyclin gene and Pol II transcription of SL Rebeca Manning-Cela,2 RNA genes. Rosaura Hernandez-Rivas 2 and Santiago Martınez-Calvillo1* 1Unidad de Biomedicina, Facultad de Estudios Superiores Iztacala, Universidad Nacional Autonoma de Mexico, Av. De los Barrios 1, Col. Los Reyes Introduction Iztacala, Tlalnepantla, Edo. de Mexico, CP, 54090, Trypanosoma brucei is a flagellated protozoan parasite Mexico. that causes human African trypanosomiasis, commonly 2 Departamento de Biomedicina Molecular, Centro de known as sleeping sickness. If left untreated, this dis- Investigacion y de Estudios Avanzados del IPN, ease produces severe neurological and psychiatric dis- Apartado Postal 14-740, Mexico, DF, 07360, Mexico. orders leading to death. The parasite is transmitted primarily by bites from infected tsetse flies (Glossina spp). T. brucei also infects cattle, where it produces the Summary disease nagana, causing major economic impact in sub- Saharan Africa (Kennedy, 2013). The parasite pos- RNA polymerase III (Pol III) produces small RNA mol- sesses a complex life cycle that involves several devel- ecules that play essential roles in mRNA processing opmental forms that alternate between the mammalian and translation. Maf1, originally described as a nega- and the invertebrate hosts. In order to differentiate and tive regulator of Pol III transcription, has been stud- adapt to the diverse environments, T. brucei requires ied from yeast to human. Here we characterized Maf1 strict control of gene expression (Kramer, 2012). in the parasitic protozoa Trypanosoma brucei T. brucei and other trypanosomatids, such as Trypa- (TbMaf1), representing the first report to analyse nosoma cruzi and Leishmania spp., have unusual mech- Maf1 in an early-diverged eukaryote. While Maf1 is anisms of gene expression. For instance, RNA generally encoded by a single-copy gene, the T. bru- polymerase II (Pol II) transcription of protein-coding cei genome contains two almost identical TbMaf1 genes is polycistronic and apparently constitutive, and genes. The TbMaf1 protein has the three conserved mature mRNAs are produced from polycistronic tran- sequences and is predicted to fold into a globular scripts by trans-splicing and polyadenylation (Gunzl structure. Unlike in yeast, TbMaf1 localizes to the et al., 2007; Martinez-Calvillo et al., 2010). Trans- nucleus in procyclic forms of T. brucei under normal splicing involves the addition of a capped 39-nucleotide growth conditions. Cell lines that either downregu- spliced leader (SL) RNA to the 50 end of each mRNA late or overexpress TbMaf1 were generated, and (Michaeli, 2011). In a cluster of protein-coding genes, growth curve analysis with them suggested that Pol II transcription initiates mainly upstream of the first TbMaf1 participates in the regulation of cell growth of gene (Martinez-Calvillo et al., 2003; Kolev et al., 2010); T. brucei. Nuclear run-on and chromatin immunopre- although these regions do not seem to contain typical cipitation analyses demonstrated that TbMaf1 Pol II promoter elements, they are enriched in histone represses Pol III transcription of tRNA and U2 snRNA modifications and histone variants associated with tran- genes by associating with their promoters. scription initiation (Siegel et al., 2009; Thomas et al., 2009; Rudenko, 2010). Accepted 26 October, 2016. *For correspondence. E-mail scalv@ campus.iztacala.unam.mx; Tel. 1(52) (55) 5623-1333 ext. 39782; The SL RNA genes are independently transcribed by Fax 1(52) (55) 5623-1138. Pol II in a monocistronic fashion, and its expression is

VC 2016 John Wiley & Sons Ltd Characterization of Maf1 in T. brucei 453 controlled by defined promoter sequences (Gunzl et al., C160 subunit of Pol III, and with the BRF1 subunit of 1997; Gilinger and Bellofatto, 2001). Several transcrip- TFIIIB (Desai et al., 2005; Reina et al., 2006). tion factors have been shown to bind the SL RNA pro- Interestingly, a study from Johnson et al. (2007) moter region in T. brucei: TATA binding protein (TBP) showed that human Maf1 is a negative regulator of the (Ruan et al., 2004), SNAPc (Das et al., 2005), TFIIA expression of TBP, which is a transcription factor (Schimanski et al., 2005a), TFIIB (Palenchar et al., required for all three RNA polymerases. Downregulated 2006; Schimanski et al., 2006), TFIIH (Lecordier et al., expression of TBP by Maf1 repressed Pol I transcription 2007; Lee et al., 2007) and the mediator complex (Lee of rRNA genes (Johnson et al., 2007). Also, a recent et al., 2010). Interestingly, the sequence of most of report showed that Maf1 controls the Pol II transcription them is extremely divergent comparing to their ortho- of several genes related to cell proliferation (Lee et al., logues in yeast and vertebrates. 2015). By binding to nearby Pol III promoters, Maf1 In T. brucei Pol I not only synthesizes rRNAs, but also seems to inhibit the recruitment of Pol II machinery and has the extraordinary ability to produce the mRNAs of other factors that mediate active histone marks to Pol II the variant surface glycoproteins (VSG) and the EP/ promoters (Lee et al., 2015). Thus, human Maf1 is a GPEET procyclins, which are two of the most abundant repressor of transcription by all three RNA polymerases proteins in the parasite (Gunzl et al., 2003). VSGs par- (Palian et al., 2014). Notably, it was found in yeast that ticipate in the process of antigenic variation, a survival Pol II transcription of two gluconeogenetic genes is strategy that allows bloodstream forms of T. brucei to downregulated by Maf1 (Morawiec et al., 2013). Conse- escape host cell immunological attacks (Borst, 2002; quently, it is possible that, as in human cells, Maf1 regu- Pays et al., 2004). Only one Pol I transcription factor, lates the transcription of all RNA polymerases in yeast. named CITFA, has been identified and characterized in In this study we characterized the Maf1 orthologue in T. brucei (Brandenburg et al., 2007). It is needed for the early-branched eukaryote T. brucei (TbMaf1). transcription of rRNAs, VSG and procyclin genes. Knock-down and overexpression analyses indicate that Pol III is involved in the production of small essential TbMaf1 is involved in the control of cell growth in procy- RNAs, such as tRNAs, 5S rRNA and U6 snRNA (Dieci clic T. brucei forms, and that it inhibits transcription of et al., 2007; Moir and Willis, 2013). In trypanosomatids, Pol III-dependent genes. Interestingly, the data indicates Pol III transcription is also unusual, as it transcribes all that TbMaf1 also inhibits transcription of procyclin and snRNAs genes (not only U6) (Fantoni et al., 1994). SL RNA genes. Moreover, snRNA genes in T. brucei have a divergently oriented tRNA gene in their 50-flanking region and, interestingly, internal sequences from the neighbouring tRNA Results genes are required for expression of the snRNAs (Nakaar et al., 1994). In yeast and higher eukaryotes, Maf1 is encoded by two genes in T. brucei Pol III requires three general transcription factors to initi- Maf1 is a negative regulator of transcription that is ate transcription: TFIIIA, TFIIIB and TFIIIC (Geiduschek highly conserved across eukaryotes. In order to deter- and Kassavetis, 2001). An orthologue of subunit BRF1 mine if a Maf1 orthologue is present in the protozoan of TFIIIB has been characterized in T. brucei (Velez- Ramirez et al., 2015). However, neither TFIIIA nor parasite T. brucei, BLAST analysis was performed. Two TFIIIC have been identified in this parasite. Maf1genes were identified in the T. brucei genome, one Maf1 is a negative regulator of transcription con- located on chromosome 4 (Tb927.4.5140) that we served from yeast to human. It was originally discovered named TbMaf1-4, and the other one on chromosome 8 as a repressor of Pol III transcription in Saccharomyces (Tb927.8.7250) that we called TbMaf1-8. Both genes cerevisiae, where Maf1 mutants showed elevated levels encode 265-amino acid proteins, with a predicted molec- of tRNAs (Pluta et al., 2001). Also, Maf1 was found to ular mass of approximately 30 kDa, that show 95.8% mediate several signal pathways by repressing tRNA identity. A sequence alignment with human Maf1 transcription in response to diverse stresses (Boguta, revealed that, although the overall homology with 2013). Similarly, it was reported that Maf1 reduces the TbMaf1-4 and TbMaf1-8 is low (15% identity), these levels of tRNAs, 5S rRNA and U6 snRNA in human cells proteins share the three conserved regions (domains A, (Reina et al., 2006; Goodfellow et al., 2008). Although B and C) (26%–30% identity) (Fig. 1A), which are pres- the overall homology among Maf1 from different species ent in all Maf1 orthologues. The length of the sequence is low, all Maf1 orthologues contain three conserved located between the A and B domains, known as linker sequence regions, named A, B and C, which show no region, is highly variable among species; for instance, homology with protein domains of known function. To these domains are fused in Encephalitozoon cuniculi, inhibit transcription, Maf1 physically interacts with the but they are separated by 182 amino acids in

VC 2016 John Wiley & Sons Ltd, Molecular Microbiology, 103, 452–468 454 G. Romero-Meza et al. ᭿

Fig. 1. TbMaf1 sequence and structure analyses. A. Sequence alignment of Maf1 proteins from S. cerevisiae (ScMaf1), H. sapiens (HsMaf1) and T. brucei (TbMaf1-4 and TbMaf1-8). Conserved and semi-conserved residues are shaded in black and grey respectively. Maf1 conserved domains are boxed in red (domain A), blue (domain B) and green (domain C). For ScMaf1, only the sequences that correspond to domains A–C are shown, because the sequences of the linker and carboxy-terminal regions are larger and divergent in relation to the other Maf1 orthologues. The two Maf1 signature sequences are shown above domains B and C, with conserved residues shaded in black. The putative nuclear localization signal is underlined. Amino acids that are different between TbMaf1-4 and TbMaf1-8 are indicated with arrowheads. Arrows show amino acids that are required for Maf1 function in yeast or mammals that are conserved in T. brucei. Circles show amino acids that are phosphorylated in human (Michels, 2011) (not conserved in T. brucei). Diamonds show important amino acids for Maf1 function in yeast (Roberts et al., 2006; Gajda et al., 2010) (not conserved in T. brucei). Predicted secondary structure elements are shown below the alignment. The b-strands are represented with arrows and a-helices with rounded rectangles, and they are coloured in accordance to the conserved domain where they are located. B. Predicted three-dimensional structure of TbMaf1. Homology modelling was performed for TbMaf1 (left image) using the crystal structure of human Maf1 as a template. As control, the structure of human Maf1 was predicted (centre image). A merge between TbMaf1 and HsMaf1 is shown in the right image. The linker region and the C-terminal region of Maf1 were not included in the modelling and are represented with dashed white lines. The colours of the domains correspond to the colours shown in panel A. The putative nuclear localization signal is coloured in purple. The quality of the homology models was evaluated with the Mod Eval program, showing a score of 0.97.

VC 2016 John Wiley & Sons Ltd, Molecular Microbiology, 103, 452–468 Characterization of Maf1 in T. brucei 455

S. cerevisiae (Gajda et al., 2010). In T. brucei, the linker localization signal (NKRA/TR), which overlaps with part region is composed of 36 amino acids, whereas in of the signature sequence located in domain C, was human it is 33 amino acids long (Fig. 1A). identified in TbMaf1 (Fig. 1A). Homology modelling Partially-conserved Maf1 signature sequences indicates that this signal (shown in purple in Fig. 1B) is (PDYDFS and LWSxxYFxYNKRxKR) were found within surface accessible, which suggests that TbMaf1 is a domains B and C in TbMaf1-4 and TbMaf1-8 (Fig. 1A). nuclear protein. Thus, to investigate the TbMaf1 cellular Several amino acids that are required for Maf1 function localization, a polyclonal antibody was generated by in repression have been characterized in yeast and immunizing mice with recombinant GST-TbMaf1 protein mammals (Moir et al., 2006; Roberts et al., 2006; Gajda (Supporting Information Fig. S2). Immunoblot analysis et al., 2010; Michels, 2011). Interestingly, four of these performed with purified nuclear and cytoplasmic protein amino acids (S-76, D-117, S-119 and K-298) are con- fractions from procyclic parasites showed that the anti- served in TbMaf1 (Fig. 1A). body reacts with an approximately 37-kDa protein only A single Maf1 gene was identified in T. cruzi and in the nuclear fraction (Fig. 2A). The same membrane Leishmania species, with the exception of L. braziliensis was stripped and incubated with anti-aldolase and anti- that contains two identical genes (Supporting Informa- H3K18ac antibodies, which were employed as cytoplas- tion Fig. S1). Relatively low conservation in the pre- mic and nuclear controls respectively (Fig. 2A). To verify dicted amino acid sequences was found between the nuclear localization of TbMaf1, indirect immunofluo- T. brucei and T. cruzi (58.9% identity) and between rescence assays were carried out with the anti-TbMaf1 T. brucei and Leishmania (47% identity) (Supporting antibody. However, no signal was detected in these Information Fig. S1). By contrast, the sequence of the experiments (data not shown). Maf1 proteins is highly conserved among Leishmania As an alternative approach, we produced a cell line species (90%–96% identities) (Supporting Information where TbMaf1 was tagged with a carboxy-terminal PTP Fig. S1). epitope to perform indirect immunofluorescence experi- The secondary structure of TbMaf1 was predicted ments. The PTP tag is composed of Protein A (ProtA) using the PSIPRED program, which indicated that and Protein C (ProtC) epitopes separated by a tobacco domain A folds into two b-sheets and two a-helices, etch virus (TEV) protease cleavage site (Schimanski whereas domain B is composed of two a-helices et al., 2005b). To generate the TbMaf1-PTP cell line, (Fig. 1A). Domain C is predicted to form an a-helix and plasmid pC-TbMaf1-PTP was obtained by cloning the 30 three b-sheets. By contrast, no secondary structure ele- terminal coding sequence of Tb-Maf1-8 into the ments are predicted for the linker region and the genome-integration vector pC-PTP-BLA (Schimanski carboxy-terminal region (Fig. 1A). A similar secondary et al., 2005b). Because the nucleotide sequence of structure was reported for human Maf1 (Vannini et al., TbMaf1-4 and TbMaf1-8 is almost identical, plasmid pC- 2010). The predicted three-dimensional structure of PTP-BLA could integrate into either gene. Southern blot TbMaf1 was generated by homology modelling, using analyses using as probes the gene-specific 30-UTRs as a template the crystal structure of human Maf1, (Supporting Information Fig. S3A) showed that the vec- which was obtained from a Maf1 variant that lacked the tor actually integrated into the TbMaf1-4 locus (Support- last a-helix from domain A, the linker region and the ing Information Fig. S3B and C). After the expression of carboxy-terminal region (Vannini et al., 2010); conse- the recombinant TbMaf1-PTP protein was confirmed quently, these regions were not included in the TbMaf1 by immunoblot analysis with an anti-ProtC antibody model either (Fig. 1B). The obtained three-dimensional (Fig. 2B), indirect immunofluorescence experiments structure of TbMaf1 is very similar to the structure were performed using the same antibody. Analysis of reported for human Maf1, as they both show a globular the images revealed a dotted pattern inside the nucleus architecture with five centrally located b-sheets that are of the transfected parasites (Fig. 2C). Therefore, our flanked by four a-helices (Fig. 1B) (Vannini et al., 2010). results showed that TbMaf1 localizes to the nucleus in Altogether, our sequence and structure analyses procyclic forms of T. brucei. strongly suggest that TbMaf1-4 and TbMaf1-8 are the Maf1 orthologues in T. brucei. TbMaf1 plays a role in controlling cellular growth

Studies in Drosophila melanogaster have shown that TbMaf1 is a nuclear protein Maf1 regulates body size and developmental timing by In S. cerevisiae, Maf1 is normally dispersed throughout controlling tRNA synthesis (Rideout et al., 2012). More- the cell, but stress conditions lead to its delocalisation to over, in several organisms Maf1 has been shown to the nucleus (Roberts et al., 2006). A putative nuclear inhibit cell proliferation and lipogenesis (Khanna et al.,

VC 2016 John Wiley & Sons Ltd, Molecular Microbiology, 103, 452–468 456 G. Romero-Meza et al. ᭿

Fig. 2. TbMaf1 is a nuclear protein. A. Nuclear and cytoplasmic fractions from procyclic T. brucei parasites were analysed by immunoblot using the TbMaf1 antiserum (top image). For fraction purity control, membranes were stripped and incubated with an aldolase antibody (middle image) or a histone H3K18ac antibody (bottom image). In all cases, 50 mg of protein were analysed. B. Immunoblot analysis with total protein from wild type cells and parasites that express the TbMaf1-PTP recombinant protein using an anti-ProtC monoclonal antibody. C. PTP-tagged TbMaf1 was located by indirect immunofluorescence experiments with anti-ProtC monoclonal antibody and an Alexa-Fluor 488 conjugated secondary antibody. Nucleus and kinetoplast are stained with DAPI. Three representative images of three independent experiments are shown.

2014; Palian et al., 2014). Thus, in order to determine if Northern blot analysis showed that the levels of TbMaf1 is involved in cell growth in T. brucei,we TbMaf1 mRNA were reduced by around 80% after 48 h knocked-down TbMaf1 in vivo by RNA interference of RNAi induction with doxycycline (Fig. 3A). Immunoblot (RNAi). For that purpose, a 561 bp fragment from the analysis with the TbMaf1 polyclonal antiserum demon- TbMaf1-8 gene was cloned into pZJM, a vector that strated that the amount of TbMaf1 protein was decreased contains two opposite T7 RNA polymerase promoters by approximately 60% after 48 h of induction and by that are inducible with doxycycline to produce double- approximately 70% after 96 h of induction (Fig. 3B). stranded RNA (Wang et al., 2000). Because the nucleo- Thus, a strong reduction in the levels of TbMaf1 mRNA tide sequence of TbMaf1-4 and TbMaf1-8 is almost and protein was observed in the knock-down cell line identical, it is expected that parasites transfected with after RNAi induction. In order to evaluate the effect of the this vector would decrease the levels of mRNA from TbMaf1 knock-down on the growth of T. brucei,cultures both TbMaf1 genes. The resultant vector was trans- of induced (Dox1) and non-induced (Dox2)parasites fected into procyclic T. brucei cell line 29–13, which were counted and diluted daily for seven days. As pre- expresses the T7 RNA polymerase and the tetracycline sented in Fig. 3C, which shows a representative growth repressor (Wirtz et al., 1999). The transfected popula- curve of five independent experiments, slightly higher cell tion was cloned by limiting dilution, and a clonal cell line densities were reached in the induced cultures. Similar was selected for further analysis. effects have been reported in other organisms, including

VC 2016 John Wiley & Sons Ltd, Molecular Microbiology, 103, 452–468 Characterization of Maf1 in T. brucei 457 yeast (Pluta et al., 2001), fruit fly (Rideout et al., 2012) experiments demonstrated that the levels of TbMaf1 and worms (Cai and Wei, 2016). mRNA were clearly increased after 48 h induction (Fig. 4A). Similarly, immunoblot analysis showed that the levels of the TbMaf1 protein started to rise 48 h TbMaf1 is a negative regulator of transcription post-induction, reaching a 2.4-fold increase after 96 h The role for Maf1 in repressing Pol III transcription is induction (Fig. 4B). As expected, growth curve analy- conserved from yeast to mammals (Khanna et al., sis showed the opposite effect observed in the knock- 2015). In order to investigate if TbMaf1 is indeed down cell line (Fig. 3C), since Maf1-overexpressing involved in the regulation of Pol III transcription in T. brucei cells (Dox1) reach lower cell densities than T. brucei, nuclear run-on experiments were performed. the non-induced (Dox2) cells (Fig. 4C). Thus, these In these assays, radiolabelled nascent transcripts were results support the participation of TbMaf1 in the regu- obtained from TbMaf1 knock-down cultures that were lation of cell growth in procyclic forms of T. brucei. induced for 96 h (Dox1) or non-induced (Dox2) Overexpression studies in human cells have demon- (Fig. 3D). The Pol III-transcribed genes that we ana- strated a similar growth effect (Johnson et al., 2007; lysed were 5S rRNA, 7SL RNA, U2 snRNA, tRNA-Arg, Palian et al., 2014). tRNA-Phe and tRNA-Tyr. Also, a-tubulin, Elp3b and SL The effect of the increased expression of TbMaf1 on RNA, which are transcribed by Pol II, were included in transcription was studied by nuclear run-on analysis, the analysis. Two Pol I-transcribed genes, the 18S rRNA using cells that were induced for 48 h (Dox1) or non- and procyclin, were also analysed. As expected, the induced (Dox2). As expected, the overexpression of transcription signal of the Pol III-transcribed genes was TbMaf1 reduced Pol III transcription (Fig. 4D), as signal increased in the Dox1 culture (Fig. 3D). Quantification of 7SL RNA, U2 snRNA and tRNAs was reduced to of the hybridization signals observed in Fig. 3D and two approximately 45%–55% of the control value (Fig. 4E). more independent experiments (Fig. 3E) revealed that Thus, this data confirms that TbMaf1 is a negative the increase in signal varies from 80% (tRNA-Tyr) to regulator of Pol III transcription in T. brucei. Interestingly, 120% (snRNA U2). Interestingly, the increase observed signal of the 5S rRNA was reduced to a lesser extent with 5S rRNA (20%) was not statistically significant. (80%), showing non-significant differences with the Nevertheless, these results confirm that TbMaf1 inhibits control. Regarding Pol II transcription, while signal of a- Pol III transcription in procyclic forms of T. brucei. Inter- tubulin and Elp3b was reduced to 60% and 50%, estingly, a reproducible increase in transcription signal respectively, transcription of SL RNA was not signifi- was also observed for the three genes transcribed by cantly affected (Fig. 4D and E). Pol I transcription of Pol II that were analysed: a-tubulin (40%), Elp3b (30%) procyclin was reduced to 60%, whereas transcription and SL RNA (70%) (Fig. 3D and E). Moreover, tran- of 18S rRNA was not affected by the overexpression of scription signal for procyclin, transcribed by Pol I, TbMaf1 (Fig. 4D and E). Altogether, our results suggest showed a 60% increase after TbMaf1 RNAi induction that TbMaf1 could participate in the expression of some (Fig. 3D and E). Notably, transcription signal of 18S RNAs transcribed by Pol II and Pol I in T. brucei. rRNA, also transcribed by Pol I, remain unaffected. Therefore, these results suggest that TbMaf1 might con- TbMaf1 associates with promoter regions trol the transcription of some genes transcribed by Pol II and Pol I in T. brucei. The involvement of Maf1 in the To investigate the in vivo binding of TbMaf1 to T. brucei regulation of Pol I and Pol II transcription has been dem- promoter regions, chromatin immunoprecipitation (ChIP) onstrated in other organisms (Johnson et al., 2007; experiments were conducted. For these assays, we Palian et al., 2014; Lee et al., 2015). used the cell line that expresses the PTP-tagged version of TbMaf1, and immunoprecipitations were performed with a ChIP-grade anti-ProtA antibody, which recognizes Overexpression analysis of TbMaf1 the two ProtA domains present in the PTP tag (Lee To further explore the possible role of TbMaf1 in regu- et al., 2010), and with a nonspecific rabbit immune lating cell growth and transcription in T. brucei,acell serum as negative control. Recovered DNA was ana- line that overexpresses TbMaf1 was generated. To lysed by qPCR to explore the binding of TbMaf1-PTP to this end, the TbMaf1-8 coding region was cloned into different regions of the T. brucei genome (Fig. 5A). Four pLew100 (Wirtz et al., 1999), and the resultant vector Pol III promoter regions were analysed: the internal pro- was transfected into the procyclic T. brucei cell line moter found in tRNA-Arg, tRNA-Ala and 5S rRNA 29-13. Synthesis of exogenous TbMaf1 mRNA was genes, and the extragenic promoter from the U2 snRNA induced by the addition of doxycycline. RT-PCR gene. As negative controls, we analysed fragments

VC 2016 John Wiley & Sons Ltd, Molecular Microbiology, 103, 452–468 458 G. Romero-Meza et al. ᭿

located upstream of the two tRNA genes (tRNA inter) TbMaf1 was found in the 5S rRNA genes (Fig. 5B). As and the U2 snRNA gene (which do not contain an inter- expected, TbMaf1 did not bind the tRNA intergenic nal promoter region) (Fig. 5A). The results from three regions and the U2 snRNA gene. Thus, these data con- independent experiments showed that, as anticipated, firm the involvement of TbMaf1 in Pol III transcription in TbMaf1-PTP is highly enriched in the tRNA genes and T. brucei. Several fragments from Pol I- and Pol II- the U2 snRNA promoter (Fig. 5B). Lower enrichment of dependent genes were also analysed. Regarding Pol II

VC 2016 John Wiley & Sons Ltd, Molecular Microbiology, 103, 452–468 Characterization of Maf1 in T. brucei 459

Fig. 3. RNAi knock-down analysis of TbMaf1. A. Northern blot analysis of TbMaf1 mRNA in non-induced parasites (Dox2) and parasites induced for 24 or 48 h. To normalize the TbMaf1 mRNA levels, the filter was stripped and re-hybridized with an a-tubulin probe. Densitometric analysis from this hybridization and two independent experiments is shown below the blot. B. Immunoblot analysis of TbMaf1 protein in non-induced parasites (Dox2) and parasites induced for 24, 48, 72 and 96 h using the TbMaf1 antiserum. The bands shown here and from two independent experiments were quantified and plotted. TbMaf1 protein levels were normalized to the level of the b-tubulin protein (loading control). C. Growth curve of a clonal cell line under non-induced (Dox2) and doxycycline-induced (Dox1) conditions. A representative growth curve of five independent experiments is shown. The average doubling times were 12.6 and 10.7 h for the Dox2 and Dox1 cultures respectively. Statistical significance was determined by two-sample t-test (P < 0.0001). D. Run-on assay conducted with isolated nuclei from a clonal cell line that was either induced with doxycycline for 96 h (Dox1) or non- induced (Dox2). Labelled nascent RNA was used to hybridize dot blots of double-stranded DNAs (2 lg) cloned into the pGEM-T Easy vector. The genes transcribed by Pol III analysed were: 5S rRNA (5S), 7SL RNA (7SL), U2 snRNA (U2), tRNA-Arg (Arg), tRNA-Phe (Phe) and tRNA- Tyr (Tyr). The Pol I-transcribed 18S rRNA (18S) and procyclin (Procyc) were also analysed. Also, three genes transcribed by Pol II were studied: a-tubulin (a-tub), Elp3b (Elp3) and SL RNA. An empty vector (pGEMT) was analysed as control. E. The signals obtained for each gene in panel D, and from two more independent experiments, were quantified and plotted, considering as 100% the signal obtained in the Dox2 experiment. Values represent means of the three experiments. Standard deviation bars are shown. RNA levels were normalized to the level of the 18S rRNA. Statistical significance was determined by two-sample t-test. *P < 0.05; **P < 0.001; ***P < 0.0001. transcription, TbMaf1 was enriched in the SL RNA pro- Because sequences within 30-UTRs play a major role in moter, but not in a fragment from the intergenic region of post-transcriptional regulation in T. brucei (Clayton, the SL RNA tandem array, the a-tubulin gene and a frag- 2013), it is tempting to speculate that the expression of ment from the strand switch region (SSR) upstream of the TbMaf1-4 and TbMaf1-8 is differentially regulated and transcription unit that contains the a-tubulin genes (Fig. that they might participate in some gene-specific func- 5B). Interestingly, TbMaf1-PTP was enriched in the Pol I tions. Attempts to specifically downregulate TbMaf1-4 or 0 promoter region of the procyclin gene, but absent from the TbMaf1-8 by RNAi using the 3 -UTRs were unsuccess- Pol I promoter of the rRNA transcription unit (Fig. 5). As ful, since the abundance of the mRNAs was not reduced expected, we did not observe association of TbMaf1 with after the induced production of double-stranded RNA (Supporting Information Fig. S4). Interestingly, while the procyclin and 18S rRNA genes. Consequently, these TbMaf1-8 mRNA was observed as a sharp and discrete results confirm that TbMaf1 has a role in regulating Pol I band in Northern blot experiments, transcription of transcription of the procyclin gene and Pol II transcription TbMaf1-4 produced a smeared band that seems to con- of the SL RNA gene in procyclic forms of T. br ucei. tain a population of mRNAs of different sizes (Support- ing Information Fig. S4). This observation is supported Discussion by high-throughput RNA sequencing analysis that revealed that the length of the 30-UTR from TbMaf1-8 is In this work we identified and characterized Maf1 in the constant (it ranges from 410 to 424 bases), whereas the protozoan parasite T. brucei. Our results show that length of the 30-UTR from TbMaf1-4 is heterogeneous TbMaf1 is a negative regulator of transcription, a function (it varies from 400 to 1300 bases) (Kolev et al., 2010). that is conserved in multiple organisms including yeast, The Northern blot analysis also reveals that TbMaf1-4 is insects, worms, vertebrates and plants (Soprano et al., highly upregulated compared with TbMaf1-8 (Supporting 2013; Khanna et al., 2015). Thus, the presence of Information Fig. S4). The higher expression levels of Maf1 in T. brucei, an early diverging eukaryote, suggests TbMaf1-4 may explain why the PTP vector only inte- that Maf1 was most likely present in the last eukaryotic grated into this locus, as differential levels of transcrip- common ancestor, as has been suggested for other pro- tion would lead to selective drug sensitivity of teins with wide phylogenetic distribution (Rathore et al., transfectants between the two loci. 2016). This is different from other regulators of Pol III and A single Maf1 gene is present in all the species of Pol I transcription, such as Myc, p53 and Rb, which have Leishmania, except for L. brazileinsis that possesses been studied in higher eukaryotes (White, 2008), but do two identical Maf1 genes. Unlike T. brucei, Maf1 is a not seem to be present in trypanosomatids. single-copy gene in T. cruzi. Analysis of the TriTrypDB It is worth noting that, while all species studied con- databases of other Trypanosoma species showed that tain a single Maf1 gene, T. brucei possesses two non- T. congolense, T. evansi and T. vivax contain two Maf1 identical Maf1 genes, namely TbMaf1-4 and TbMaf1-8. genes. Thus, the number of Maf1 genes seems to be Interestingly, 30-UTR sequences differ considerably variable within trypanosomatids. Interestingly, no single between TbMaf1-4 and TbMaf1-8, as they show only Maf1 gene was identified in the T. rangeli genome; 32% identity (Supporting Information Fig. S3A). moreover, there are no Maf1 degenerate sequences

VC 2016 John Wiley & Sons Ltd, Molecular Microbiology, 103, 452–468 460 G. Romero-Meza et al. ᭿

Fig. 4. Overexpression analysis of TbMaf1. A. The levels of TbMaf1 mRNA were assayed by RT-PCR in cells that were induced with doxycycline for 48 h (Dox1) to overexpress Maf1, or non-induced (Dox2). Densitometric analysis from three independent experiments is shown below the gel. As a control, a fragment of a-tubulin was amplified too. B. Immunoblot analysis of TbMaf1 protein in non-induced parasites (Dox2) and parasites induced for 24, 48, 72 and 96 h using the TbMaf1 antiserum. Densitometric analysis of three independent experiments is shown below. TbMaf1 protein levels were normalized to the level of the b-tubulin protein (loading control). C. Growth curve of a clonal cell line under non-induced (Dox2) and doxycycline-induced (Dox1) conditions. A representative growth curve of three independent experiments is shown. The average doubling times were 13.6 and 17.5 h for the Dox2 and Dox1 cultures respectively. Statistical significance was determined by two-sample t-test (P < 0.002). D. Run-on assay performed with isolated nuclei from a clonal cell line that was either induced with doxycycline for 48 h (Dox1) or non-induced (Dox2). Labelled nascent RNA was used to hybridize dot blots of double-stranded DNAs (2 lg) cloned into the pGEM-T Easy vector, as indicated in Fig. 3D. E. The signals obtained for each gene in panel D, and from two more independent experiments, were quantified and plotted, considering as 100% the signal obtained in the Dox2 experiment. Values represent means of the three experiments. Standard deviation bars are shown. RNA levels were normalized to the level of the 18S rRNA. Statistical significance was determined by two-sample t-tests. *P < 0.05; **P < 0.001; ***P < 0.0001.

VC 2016 John Wiley & Sons Ltd, Molecular Microbiology, 103, 452–468 Characterization of Maf1 in T. brucei 461

Fig. 5. ChIP analysis of TbMaf1. A. Schematic drawn to scale of the genes and promoters analysed by ChIP assays. Dark grey areas represent promoter regions. Scale bar represents 100 bp for all the loci, except for the a-tubulin locus, where it represents 500 bp. B. Chromatin from a clonal cell line that expresses the recombinant protein TbMaf1- PTP was precipitated with a ChIP-grade anti-ProtA antibody. Precipitated DNA was analysed by qPCR. The results from three independent ChIP experiments, each analysed by three qPCR reactions, are shown. Error bars indicate standard deviations. Results are presented as fold enrichment over negative control precipitations.

(pseudogenes) in the corresponding regions of synteny. obtained by homology modelling showed that TbMaf1 Further analysis is required to demonstrate the absence forms a globular structure with five centrally located b- of Maf1 in T. rangeli and its possible biological sheets that are flanked by one a-helix on one side and implications. three a-helices on the other side (Fig. 1B). This struc- Our results showed that the TbMaf1-4 and TbMaf1-8 ture is very similar to the one reported for human Maf1 proteins possess all the sequence and structural fea- (Vannini et al., 2010). For the most part, domains A, B tures present in Maf1 orthologues from other organisms and C seem to correspond to independently folded (Fig. 1). For example, sequence comparisons revealed structural modules. The putative nuclear localization sig- that TbMaf1-4 and TbMaf1-8 contain domains A, B and nal, which overlaps with part of the signature sequence C, which are conserved in Maf1 proteins across eukar- located in domain C, is exposed at the surface of the yotes. The two Maf1 signature sequences located on globular protein structure (Fig. 1B). The apparent molec- domains B and C are present in TbMaf1-4 and TbMaf1- ular mass of TbMaf1 (37 kDa) (Fig. 2A) is a little 8 as well. The predicted tridimensional structure higher than its predicted mass (30 kDa). This fact has

VC 2016 John Wiley & Sons Ltd, Molecular Microbiology, 103, 452–468 462 G. Romero-Meza et al. ᭿

been reported for Maf1 in several species, and it is procyclins and rRNA genes are similar, but not identical, apparently caused by post-translational modifications and it is expected that they recruit different transcription (Oler and Cairns, 2012; Rohira et al., 2013). It remains factors (Sherman et al., 1991; Laufer et al., 1999). to be determined if the function of TbMaf1 is regulated Therefore, it is possible that TbMaf1 associates with the by post-translational modifications. procyclin promoter by interacting with a procyclin- Immunoblot (Fig. 2A) and indirect immunofluorescence specific transcription factor. In mammalian cells, Maf1 (Fig. 2C) analyses showed that TbMaf1 localizes to the regulates 18S rRNA transcription by directly controlling nucleus in exponentially-growing procyclic forms of the expression of TBP (Johnson et al., 2007), which is a T. brucei. In contrast, in S. cerevisiae Maf1 is spread subunit of several transcription factors, including the Pol throughout the cell under normal growth conditions, and I transcription factor SL1. In T. brucei, TBP is required unfavourable conditions lead to its enrichment to the for transcription of procyclin genes, but it is not appa- nucleus (Roberts et al., 2006). In two different types of rently needed for transcription of the rRNA unit (Ruan human cancer cells, Maf1 has been either localized to et al., 2004; Thomas et al., 2006). Interestingly, RT-PCR the nucleus (Rollins et al., 2007), or distributed evenly and qPCR analyses indicate that TBP expression is not throughout the cell (Shor et al., 2010). In the latter case, affected after TbMaf1 ablation or overexpression in mTOR inhibitors caused nuclear accumulation of Maf1. T. brucei (Supporting Information Fig. S5). Thus, unlike It has been shown that Maf1 and BRF2 (a subunit of other organisms, TBP does not seem to play an impor- transcription factor TFIIIB in vertebrates) co-localize in tant role in Maf1 regulation of transcription in T. brucei. human cells (Rollins et al., 2007). In agreement with this Depletion of TbMaf1 increased 30%–40% the Pol II result, the punctate nuclear distribution of TbMaf1 transcription signal of a-tubulin and Elp3b (Fig. 3E), (Fig. 2C) strongly resembles the nuclear pattern reported whereas, in agreement, overexpression of TbMaf1 for BRF1 in T. brucei (Velez-Ramirez et al., 2015). reduced to 50%–60% the transcription of these genes Nuclear run-on analyses with cell lines where TbMaf1 (Fig. 4E). The nuclear run-on results obtained for the SL was either depleted (Fig. 3D) or overexpressed RNA gene were somehow contradictory, since transcrip- (Fig. 4D) showed that TbMaf1 is a negative regulator of tion signal was increased 70% after TbMaf1 ablation Pol III transcription in procyclic forms of T. brucei. More- (Fig. 3E), and not significantly affected after TbMaf1 over, ChIP experiments demonstrated that TbMaf1 overexpression (Fig. 4E). Since SL RNA is encoded by associates with promoter regions from tRNA-Arg, tRNA- multiple genes, and each of them contains its own pro- Ala, U2 snRNA and 5S rRNA genes (Fig. 5). Similarly, moter region, it is possible that in the overexpression ChIP on chip analyses in yeast showed the binding of experiments the levels of Maf1 were not high enough to Maf1 to Pol III-transcribed genes in conditions of active reduce the transcription from every SL RNA gene. growth, with an increased association of Maf1 under However, ChIP analysis showed the association of unfavourable growth conditions (Oficjalska-Pham et al., TbMaf1 with the SL RNA gene promoter (Fig. 5). Thus, 2006; Roberts et al., 2006). High levels of Maf1 were these results show the participation of TbMaf1 in the found to reduce 5S rRNA transcription in mammalian transcriptional regulation of SL RNA. In human cells, cells (Reina et al., 2006; Goodfellow et al., 2008). Also, Maf1 regulates Pol II transcription of genes related to increased transcription of 5S rRNA genes was observed cell proliferation, including CDKN1A (cyclin-dependent in Maf1-deficient yeast extracts (Pluta et al., 2001). kinase inhibitor 1A), that contain tRNA-derived SINEs Notably, although our data indicates that TbMaf1 binds (short interspersed elements) in their promoter regions 5S rRNA genes (Fig. 5), we did not observe significant (Lee et al., 2015). Under normal conditions, Pol III binds changes in the transcription of 5S rRNA after TbMaf1 the SINE within the CDKN1A promoter, which in turn depletion or overexpression (Figs. 3E and 4E). Thus, induces CDKN1A transcription by recruiting Pol II, gen- the role of TbMaf1 in the regulation of 5S rRNA tran- eral transcription factors and other proteins that gener- scription in T. brucei requires further investigation. ate histone modifications associated with gene Our results also indicate that TbMaf1 is involved in activation. Adverse growth conditions lead to the binding the regulation of some genes transcribed by Pol I and of Maf1 to the SINE, which prevents the recruitment of Pol II. Regarding Pol I-dependent genes, procyclin tran- Pol III and Pol II machinery to inhibit CDKN1A transcrip- scription was reproducibly increased after TbMaf1 tion (Lee et al., 2015). depletion (Fig. 3E) and decreased after TbMaf1 overex- In human and Caenorhabditis elegans, Maf1 nega- pression (Fig. 4E), whereas 18S rRNA transcription was tively regulates genes required for de novo lipogenesis not affected. Accordingly, our ChIP assays revealed that by binding to their promoter regions (Palian et al., 2014; TbMaf1 associates with the procyclin promoter region, Khanna et al., 2014). In yeast, Maf1 represses Pol II but not with the promoter region from the rRNA unit transcription of two gluconeogenetic genes by an (Fig. 5). It is worth noting that promoter regions of unknown mechanism (Morawiec et al., 2013). Microarray

VC 2016 John Wiley & Sons Ltd, Molecular Microbiology, 103, 452–468 Characterization of Maf1 in T. brucei 463 analysis in C. elegans showed that Maf1 controls the sequence of Maf1 for RNAi experiments was chosen using expression of numerous Pol II-transcribed genes the TrypanoFAN page (http://trypanofan.path.cam.ac.uk/try- involved in essential functions, such as protein synthe- panofan/main/). To predict the protein secondary structure, the PSIPRED Protein Sequence Analysis Workbench was sis, growth and reproduction (Khanna et al., 2014). used (http://bioinf.cs.ucl.ac.uk/psipred/). Homology modelling Interestingly, a whole-body knock-out of Maf1 in mice was performed with SWISS-MODEL (http://swissmodel. showed no significant changes in the levels of Pol II expasy.org/interactive) and Swiss-PDBViewer program transcripts (Bonhoure et al., 2015). (http://www.expasy.org/spdbv/), using the crystallographic If, indeed, TbMaf1 inhibits Pol II transcription of structure of human Maf1 (PDB id: 3nr5) as a template. protein-coding genes in T. brucei, the mechanism must be different from those present in other organisms, con- Cell culture and transfection sidering that Pol II transcription in this parasite is polycistronic and that mRNA abundance is mainly regulated The procyclic form of T. brucei strain 29-13 (Wirtz et al., at a post-transcriptional level (Clayton, 2013). Notably, 1999) was cultured at 288C in SDM-79 medium supple- 21 we did not observe the association of TbMaf1 with a mented with 10% fetal bovine serum (GIBCO), 15 mgml G418 (Sigma) and 50 mgml21 hygromycin B (Sigma). 250-bp fragment from the SSR located upstream of the Transfections were performed as previously described a-tubulin locus (Fig. 5). However, we cannot rule out the (Velez-Ramirez et al., 2015). Transfectants were selected possibility that TbMaf1 binds to other parts of the SSR, with phleomycin (2.5 mgml21) (for RNA interference) or which is 3.6 kb long. In contrast, our data show the with blasticidin (10 mgml21) (for PTP-tagging and Maf1 binding of TbMaf1 to the upstream element of the SL overexpression). Clonal cell lines were obtained by serial RNA promoter region (Fig. 5). Even though domain IV dilution in 96-well plates. Silencing and overexpression of TbMaf1 was induced by adding doxycycline (2 mgml21)to of the rRNA gene promoter and the upstream element the medium (Dox1 or induced culture). As a control, the of the SL RNA promoter share sequence similarity same cell line was grown without doxycycline (Dox2 or (Schimanski et al., 2004), TbMaf1 does not associate to non-induced culture) in parallel with the induced culture. the former (Fig. 5). This suggests that TbMaf1 might Parasites were counted daily and diluted to 2 3 106 cells recognize Pol II transcription factors associated with the ml21. For growth curves, cumulative cell density was plot- SL RNA promoter, which are not present in the rRNA ted. Doubling times of the different cell lines were calcu- gene promoter. lated as previously reported (Shi et al., 2004). In conclusion, our results demonstrate that, unlike most organisms, TbMaf1 is encoded by two genes in Plasmids the protozoan parasite T. brucei. Both TbMaf1 proteins contain the three typical Maf1 domains and are pre- To obtain plasmid pZ-TbMaf1 for RNAi experiments, a 561- bp fragment from the TbMaf1-8 gene (Tb927.8.7250) was dicted to fold into a conserved globular structure. Our amplified with primers RNAi-Fw (50-AGGTCCAGCTTG- data also show that TbMaf1 localizes to the nucleus in GAGGCTTTCCATGTCGTT) and RNAi-Rv (50-CTTATCTTT T. brucei procyclic forms. Growth analysis of cell lines GCGGAGGGAAACTCGAGT) and cloned into the pZJM that either downregulate or overexpress TbMaf1 sug- vector (Wang et al., 2000) using the XhoI and BamHI gests that TbMaf1 has a role in controlling cell growth of restriction sites. Before transfection, pZ-TbMaf1 was linear- 0 T. brucei. As expected for a Maf1 orthologue, our results ized with NotI. The 3 -UTR of TbMaf1-4 (Tb927.4.5140) was amplified with oligonucleotides Maf4iRNAFw (50-GGAT show that TbMaf1 inhibits transcription of Pol III- CCAAGCTTTGTGATTGATAGTGGAGCGG) and Maf4iR- dependent genes. Moreover, our data indicates that NARv (50-TTCTTTCCGTACTTCCACCGCTCGAG); while TbMaf1 also regulates Pol I transcription of the procyclin the 30-UTR of TbMaf1-8 was amplified with primers Maf8iR- gene and Pol II transcription of the SL RNA genes. NAFw (50-GGATCCAAGCTTTACTGGAGCTAGGAAGAGGA) Therefore, the function of Maf1 is highly conserved and Maf8iRNARv (50-TATTCCATCGTGTGTGCTCCCTCGA through evolution. G). Both regions were also cloned into pZJM. To generate vector pLew-TbMaf1 for overexpression of TbMaf1, gene Tb927.8.7250 was amplified with primers HindIII-Fw (50-AA GCTTTTCTAATGGAGTTCGTTC) and BamHI-Rv (50-TATTA Experimental procedures TTATCGCCCGTGAGGATCC) and cloned into pLew100v5- In silico analyses BSD vector (Wirtz et al., 1999), using the restriction sites indicated in the primer’s name. For transfection, pLew- Sequences were obtained from the TriTrypDB database TbMaf1 was linearized with NotI. To obtain vector (release 27) (http://tritrypdb. org/tritrypdb/) and from the pC-TbMaf1-PTP for PTP tagging, a 489-bp fragment from National Center for Biotechnology Information (NCBI) data- the C-terminal TbMaf1 coding region was amplified with pri- base (http://www. ncbi.nlm.nih.gov/). Multiple sequence align- mers PTP-Fw (50-GGGCCCCTAAACAGCATTTATGGTGAG) ments were produced with the Clustal Omega program and PTP-Rv (50-GTACATTATTATTATCGCCCGGCGCGGCC (http://www.ebi.ac.uk/Tools/msa/clustalo/). The most suitable GC) and cloned into the ApaIandNotI restriction sites of the

VC 2016 John Wiley & Sons Ltd, Molecular Microbiology, 103, 452–468 464 G. Romero-Meza et al. ᭿

pC-PTP-Bla plasmid (Nguyen et al., 2007). Prior transfec- system (Amersham). This fragment was amplified by PCR tion, the plasmid was linearized with NcoI. To obtain vector with primers TbMaf1-RNAiF (50-AGGTCCAGCTTGGAGGC pX-TbMaf1 for the expression of recombinant GST-TbMaf1 TTTCCATGTCGTT) and TbMaf1-RNAiR (50-CTTATCTTT protein, the entire TbMaf1 gene was amplified with oligonu- GCGGAGGGAAACTCGAGT). Filters were washed to a cleotides BamHi-Fw (50-CGGGATCCATGGAGTTCGTTC final stringency of 0.13 SSC and 0.1% SDS at 658C. Other CCCTCGAG) and NotI-Rv (50-GTACATTATTATTATCGCCC experiments were carried out with approximately 300-bp GGCGGCCGC) and cloned into the pGEX-4T3 vector (GE probes corresponding to the 30-UTRs of TbMaf1-4 or T healthcare) using the restriction sites indicated in the pri- bMaf1-8, which were amplified as described in the Plas- mer’s name. All constructs were verified by sequencing. mids section. For Southern blot analysis, 5 mg of genomic DNA, isolated as reported elsewhere (Carrington, 1993), were digested with AgeI, BamHI, MfeI and SphI (New Eng- Indirect immunofluorescence land Biolabs). DNA was separated by electrophoresis and transferred by capillary to Hybond-N membranes. The radi- These assays were performed as previously described oactive probes corresponded to the 30-UTRs of TbMaf1-4 (Velez-Ramirez et al., 2015), using 1.5 3 107 mid-log para- or TbMaf1-8. Filters were washed as described above. sites that expressed the recombinant TbMaf1-PTP protein. Cells were incubated with a rabbit anti-ProtC polyclonal antibody (Delta Biolabs) and a secondary goat anti-rabbit Generation of TbMaf1 polyclonal antibody antibody conjugated with Alexa-Fluor 488 (Life Technolo- gies). Images were obtained with a Zeiss Axiovert A1 Plasmid pX-TbMaf1 was transformed into Escherichia coli microscope and analysed with the LEN2 software. JM109 competent cells, and expression of the GST-TbMaf1 recombinant protein was induced with 0.5 mM IPTG. Pro- tein was purified by glutathione-affinity chromatography (GE Protein extraction and immunoblot analysis Healthcare) according to the manufacturer’s specifications. Purified recombinant protein (30 mg) was used to inoculate Whole-cell extracts were obtained by standard procedures 4-week-old female BALB-C mice intraperitoneally in com- (Ji, 2010). Nuclear and cytoplasmic fractions were obtained plete Freund’s adjuvant (Sigma) at a 1:1 ratio. Animals from 1.8 3 106 parasites, which were harvested and were boosted after 3 and 6 weeks with a similar dose of washed once with lysis buffer (10 mM Tris pH 7, 3 mM the antigen in incomplete Freund’s adjuvant (Sigma). CaCl , 2 mM MgCl ). The pellet was resuspended in lysis 2 2 Serum was collected 15 days after the last immunization. buffer with 0.5% NP-40 and protease inhibitor cocktail Pre-immune normal mouse serum was collected before (Sigma). Cells were transferred to a Dounce homogenizer immunization. and broken with 20–30 strokes. The extract was centrifuged at 1500g for 10 min and the supernatant was recovered and kept at 2208C as the cytoplasmic fraction. The pellet Nuclear run-on analysis was resuspended in extraction buffer (20 mM HEPES, 420 mM NaCl, 1 mM EDTA, 1 mM EGTA, 0.5 mM DTT, These assays were conducted as described previously protease inhibitor cocktail) and incubated for 40 min in a (Padilla-Mejia et al., 2015; Velez-Ramirez et al., 2015), tube rotator. The extract was centrifuged and the superna- using isolated nuclei from 2 3 108 mid-log cells incubated tant was collected and kept at 2208C as the nuclear frac- in the presence of doxycycline for 96 h (for knock-down tion. All steps were performed at 48C. For immunoblot studies) or 48 h (for overexpression analysis). Labelled nas- analysis, 50–80 mg of protein were fractionated on SDS- cent RNA was hybridized to Hybond-N1 filters (Amersham) 10% polyacrylamide gels and blotted to PVDF membranes containing dots of 2 mg of plasmid DNA. These plasmids (Bio-Rad). Membranes were incubated with monoclonal contain fragments of genes transcribed by Pol I (18S rRNA anti-ProtC antibody HPC4 (1:250 dilution, Roche), polyclo- and procyclin), Pol II (a-tubulin, Elp3b and SL RNA) and nal anti-TbMaf1 antiserum (1:3000 dilution), polyclonal anti Pol III (5S rRNA, 7SL RNA, U2 snRNA, tRNA-Arg, tRNA- human b-tubulin (1:500, Invitrogen), polyclonal anti-human Phe and tRNA-Tyr), as published recently (Velez-Ramirez acetyl-histone H3 (Lys 18) (1:25 000, Millipore), or anti- et al., 2015). After hybridization, filters were washed to a Plasmodium aldolase antibody (1:25 000, Abcam). Proteins final stringency of 0.13 SSC and 0.1% SDS at 658C. RNA were detected with a horseradish peroxidase-conjugated signals were quantified by densitometry using the Multi- secondary antibody, and developed using an ECL kit (GE Gauge software. Healthcare).

RT-PCR assays Northern and Southern blot analyses To analyse the TbMaf1 mRNA levels in the TbMaf1 over- For Northern blot experiments, total RNA was extracted expression studies, RT-PCR assays were performed with with TRI reagent (Sigma), separated in agarose- the SuperScript One-Step RT-PCR with Platinum Taq formaldehyde gels and transferred to Hybond-N filters System (Invitrogen), following the manufacturer’s instruc- (Amersham). Hybridization was performed with an approxi- tions. TbMaf1 (Tb927.8.7250) was amplified with primers mately 500-pb probe from the TbMaf1 gene that was HindIII-Fw (50-AAGCTTTTCTAATGGAGTTCGTTC) and labelled with [a-32P]dCTP using the High Prime labelling BamHI-Rv (50-TATTATTATCGCCCGTGAGGATCC). The

VC 2016 John Wiley & Sons Ltd, Molecular Microbiology, 103, 452–468 Characterization of Maf1 in T. brucei 465 a-tubulin (Tb927.1.2340) transcript was amplified with the gene (Tb927.6.510) was amplified with primers Procyclin-5 primers Alfa-tub 50 (50-AGAAGTCCAAGCTCGGCTACAC) (50-ATGGCACCTCGTTCCCTTTA) and ProcqRv (50-CTTTG and Alfa-tub 30 (50-CTTCAAGTGCGGCATCAACTAC). For CCTCCCTTCACGATAAC); and the procyclin promoter region TBP mRNA quantification after TbMaf1 knock-down, the with primers GPEET-promoter-F (50-TGATAGGTATCTCTTAT- TBP transcript (Tb927.10.1550) was amplified with TAGTATAG) and GPEET-promoter-R (50-GTACTCACCCG oligonucleotides TBP 50 (50-CCTCTGGTCTTCTTGGCATC) ATAACCCC), as reported previously (Lee et al., 2010). The and TBP 30 (50-CAACTTCTCCAGCGTCCAGT); and strand switch region (SSR) fragment was amplified with oligo- TbMaf1 was amplified with the oligonucleotides Maf1qFw nucleotides SSRatub50 (50-GGTGGGCGTGTATTCTTTGT) (50-ATTCTTTGCTTCAGGGTGTG) and Maf1qRv (50- and SSRatub30 (50-TATGTGTTTTCGGTCTGGGG); and the GAAGCCGTATGACGTCCGCTT). Amplification products a-tubulin gene (Tb927.1.2340) with TubqFw (50-GGGCTTCC were resolved by electrophoresis in agarose gels stained TCGTGTATCA) and TubqRv (50-GCTTGGACTTCTTGCC with ethidium bromide and signals were quantified by den- ATAG). The SL RNA promoter region was amplified with oli- sitometry using the MultiGauge software. gonucleotides SL-promoter-F (50-CTACCGACACATTTCTGG C) and SL-promoter-R (50-GCTGCTACTGGGAGCTTCTC ATACC); and SL RNA intergenic region with primers Chromatin immunoprecipitation analysis SL-inter-F (50-ATGGCTTATACGTGCTCGTTTCTCC) and SL-inter-R (50-AGCAGACTTTAAAGCGCCTATATGTG) (Lee ChIP experiments were carried out as previously described et al., 2010). The rRNA 5S gene (Tb927.8.1381)wasampli- (Park et al., 2011; Vizuet-de-Rueda et al., 2016), with some fied with primers 5SrRNA-F1 (50-GAGTACGACCACACTTG 8 modifications. For each assay, 1 3 10 parasites were AGTG) and 5SrRNA-R1 (50-GAGTACGGCACTCAGGGTT). cross-linked with formaldehyde (final concentration of 1%) ThetRNA-Alagene(Tb927.7.6821) was amplified with for 5 min at 378C. Cells were broken with a Vibra-Cell oligonucleotides AlaqFw (50-GGGGATGTAGCTCAGATGG) VCX130 ultrasonic processor (Sonics) (15 s on/off, 40% and AlaqRv (50-TGGAGAAGTTGGGTATCGATC); and the amplitude) for 5 min. Nuclei were pelleted and resuspended tRNA-Arg gene (Tb927.8.2859) with oligonucleotides ArgqFw in sonication buffer (1% SDS, 10 mM EDTA, 50 mM Tris- (50-GGTCTCGTGGCGCAATG) and ArgqRv (50-CGATCCC HCl pH 8.1). Chromatin was sonicated to an average DNA GGCAGGACTC), as reported before (Velez-Ramirez et al., size of about 200–500 bp using a BioRuptor UCD-200 (Dia- 2015). The intergenic region upstream of the tRNA-Ala gene genode) (30 s on/30 s off, high intensity) for 40 cycles. The was amplified with oligonucleotides InterAla50 (50-CACTC sonicated material was pre-cleared by adding protein A/G TCCCGAGAATCGAAG) and InterAla30 (50-AAAAGTCGAC plus-agarose beads (Santa Cruz Biotechnology) and rotat- TCCACACCCA); and the intergenic region upstream of the ing for 1 h at 48C. Chromatin samples were incubated over- tRNA-Arg gene with primers InterArg50 (50-GGCTGAAAATAG night at 48C with rabbit anti-ProtA antibody (Sigma) or CGGAAGTG) and InterArg30 (50-GACTAACGACACGGGCT nonspecific rabbit immune serum as negative control. The AGC). The U2 promoter region was amplified with primers protein-DNA complexes were incubated for 1 h with protein U2Prom50 (50-CACAACCTGTAGTGGCGGTA) and U2Prom30 A/G plus-agarose beads and 200 ng of sonicated salmon (50-TTACGACACCTGGTCCATTG); and U2 snRNA gene sperm DNA, and washed as previously described (Vizuet- (Tb927.2.5680) with primers U2qFW (50-CTCGGCTATTT de-Rueda et al., 2016). Cross-links were reversed with AGCTAAGATCAAGT) and U2qRV (50-AAACTGTTGGCTGT 200 mM NaCl at 658C overnight. Following RNase A and CCCG). To analyse the TBP mRNA levels after TbMaf1 over- proteinase K treatment, DNA was precipitated with sodium expression, 1 lg of total RNA from the 48 h induced and acetate and ethanol and quantified. Each ChIP experiment non-induced parasites was used as template for the first was performed at least three times. strand cDNA synthesis using SuperScriptTM III Reverse Tran- scriptase (Invitrogen) and 50 ng of random hexamers (Invitro- gen). The cDNA was analysed by quantitative real-time PCR Quantitative real-time PCR (qPCR) assays using the Platinum SYBR Green qPCR SuperMix-UDG kit, as explained above, and the obtained Around 2 ng of immunoprecipitated DNA was analysed by data was analysed by the 22DDCq method. The a-tubulin quantitative real-time PCR (qPCR) using the Platinum SYBR gene was amplified with primers TubqFw (50-GGGCTTCCT Green qPCR SuperMix-UDG kit (Invitrogen). Data were CGTGTATCA) and TubqRv (50-GCTTGGACTTCTTGCCAT analysed by the 22DDCq method, as reported previously AG); and the TBP gene was amplified with the primers (Velez-Ramirez et al., 2015; Vizuet-de-Rueda et al., 2016). TBP 50 (50-CCTCTGGTCTTCTTGGCATC) and TBP 30 All qPCR reactions were performed in triplicate, using primers (50-CAACTTCTCCAGCGTCCAGT). TbMaf1 was amplified and conditions that were optimized to produce a single ampli- with oligonucleotides Maf1qFw (50-ATTCTTTGCTTCAGGGT con of the correct size. Results are presented as fold enrich- GTG) and Maf1qRv (50-GAAGCCGTATGACGTCCGCTT). ment over negative control precipitations. Domain IV of the rRNA promoter (18S-IV) was amplified with oligonucleotides 18S4Fw (50-CACCCTCAAGACCGTAGCTC) and 18S4Rv (50- Acknowledgements TGTGTTGATAAGGGACGGGT); and domains I/II (18S-I/II) with primers 18S1-2Fw (50-CTGTGGGGAACACACAAACA) We thank Arthur Gunzl for providing a detailed ChIP protocol and 18S1-2Rv (50-TGGTGTTTCCCTCTACAGGG). The 18S and Magdalena Aguirre for providing the b-tubulin antibody. rRNA gene (Tb927.2.1452) was amplified with oligonucleo- We also thank Juan C. Vizuet, Norma E. Padilla and Lourdes tides 18SrRNA-50 (50-CGGCTTCCAGGAATGAAGG) and Mejıa for technical help. This work was supported by grant 18SrRNA-30 (50-CCCCTGAGACTGTAACCTC). Procyclin 251831 from CONACyT, and grants IN210712 and IN214715

VC 2016 John Wiley & Sons Ltd, Molecular Microbiology, 103, 452–468 466 G. Romero-Meza et al. ᭿

from PAPIIT (UNAM) to S. Martınez-Calvillo; and by grants small nuclear RNA gene is controlled by both intragenic 132312 and 139898 from CONACYT to R. Manning-Cela. and extragenic regulatory elements. Mol Cell Biol 14: Gabriela Romero-Meza was the recipient of a doctoral fellow- 2021–2028. ship from CONACyT (Fellowship 344986, CVU 376635). The Gajda, A., Towpik, J., Steuerwald, U., Muller, C.W., authors declare that they have no conflict of interest. Lefebvre, O., and Boguta, M. (2010) Full repression of RNA polymerase III transcription requires interaction between two domains of its negative regulator Maf1. Author contributions J Biol Chem 285: 35719–35727. Geiduschek, E.P., and Kassavetis, G.A. (2001) The RNA GRM performed most of the experiments, analysed data polymerase III transcription apparatus. J Mol Biol 310: and drafted the manuscript. DEVR performed some in 1–26. silico analyses and helped generate some vectors. Gilinger, G., and Bellofatto, V. (2001) Trypanosome spliced leader RNA genes contain the first identified RNA poly- LEFM helped perform transfections, Northern blot and merase II gene promoter in these organisms. Nucleic indirect immunofluorescence analyses. FCRC helped Acids Res 29: 1556–1564. perform Southern blot and qPCR experiments. RMC Goodfellow, S.J., Graham, E.L., Kantidakis, T., Marshall, L., analysed data and participated in the writing and editing Coppins, B.A., Oficjalska-Pham, D., et al. (2008) Regula- of the manuscript. RHR analysed data and participated tion of RNA polymerase III transcription by Maf1 in mam- in the writing and editing of the manuscript. SMC con- malian cells. J Mol Biol 378: 481–491. ceived the study, analysed data and wrote the manu- Gunzl, A., Ullu, E., Dorner, M., Fragoso, S.P., Hoffmann, K.F., Milner, J.D., et al. (1997) Transcription of the Trypa- script. All authors read and approved the manuscript. nosoma brucei spliced leader RNA gene is dependent only on the presence of upstream regulatory elements. Mol Biochem Parasitol 85: 67–76. References Gunzl, A., Bruderer, T., Laufer, G., Schimanski, B., Tu, L.C., Chung, H.M., et al. (2003) RNA polymerase I transcribes Boguta, M. (2013) Maf1, a general negative regulator of procyclin genes and variant surface glycoprotein gene RNA polymerase III in yeast. Biochim Biophys Acta 1829: expression sites in Trypanosoma brucei. Eukaryot Cell 2: 376–384. 542–551. Bonhoure, N., Byrnes, A., Moir, R.D., Hodroj, W., Preitner, Gunzl, A., Vanhamme, L., and Myler, P.J. (2007) Transcrip- F., Praz, V., et al. (2015) Loss of the RNA polymerase III tion in trypanosomes: a different means to the end. In repressor MAF1 confers obesity resistance. Genes Dev Trypanosomes: After the Genome. Barry, J.D., McCulloch, 29: 934–947. R., Mottram, J.C., and Acosta-Serrano, A. (eds). Wymon- Borst, P. (2002) Antigenic variation and allelic exclusion. ham: Horizon Bioscience, pp. 177–208. Cell 109: 5–8. Ji, H. (2010) Lysis of cultured cells for immunoprecipitation. Brandenburg, J., Schimanski, B., Nogoceke, E., Nguyen, Cold Spring Harb Protoc 2010: pdb.prot5466. T.N., Padovan, J.C., Chait, B.T., et al. (2007) Multifunc- Johnson, S.S., Zhang, C., Fromm, J., Willis, I.M., and tional class I transcription in Trypanosoma brucei depends Johnson, D.L. (2007) Mammalian Maf1 is a negative reg- on a novel protein complex. Embo J 26: 4856–4866. ulator of transcription by all three nuclear RNA polymer- Cai, Y., and Wei, Y.H. (2016) Stress resistance and lifespan ases. Mol Cell 26: 367–379. are increased in C. elegans but decreased in S. cerevisiae Kennedy, P.G. (2013) Clinical features, diagnosis, and treat- by mafr-1/maf1 deletion. Oncotarget 7: 10812–10826. ment of human African trypanosomiasis (sleeping sick- Carrington, M. (1993) Preparation of DNA and RNA from ness). Lancet Neurol 12: 186–194. Trypanosoma brucei. Methods Mol Biol 21: 101–111. Khanna, A., Johnson, D.L., and Curran, S.P. (2014) Physio- Clayton, C. (2013) The regulation of trypanosome gene logical roles for mafr-1 in reproduction and lipid homeo- expression by RNA-binding proteins. PLoS Pathog 9: stasis. Cell Rep 9: 2180–2191. e1003680. Khanna, A., Pradhan, A., and Curran, S.P. (2015) Emerging Das, A., Zhang, Q., Palenchar, J.B., Chatterjee, B., Cross, roles for maf1 beyond the regulation of RNA polymerase G.A., and Bellofatto, V. (2005) Trypanosomal TBP func- III activity. J Mol Biol 427: 2577–2585. tions with the multisubunit transcription factor tSNAP to Kolev, N.G., Franklin, J.B., Carmi, S., Shi, H., Michaeli, S., direct spliced-leader RNA gene expression. Mol Cell Biol and Tschudi, C. (2010) The transcriptome of the human 25: 7314–7322. pathogen Trypanosoma brucei at single-nucleotide reso- Desai, N., Lee, J., Upadhya, R., Chu, Y., Moir, R.D., and lution. PLoS Pathog 6: e1001090. Willis, I.M. (2005) Two steps in Maf1-dependent repres- Kramer, S. (2012) Developmental regulation of gene expression of transcription by RNA polymerase III. J Biol Chem sion in the absence of transcriptional control: the case of 280: 6455–6462. kinetoplastids. Mol Biochem Parasitol 181:61–72. Dieci, G., Fiorino, G., Castelnuovo, M., Teichmann, M., and Laufer, G., Schaaf, G., Bollgonn, S., and Gunzl, A. (1999) Pagano, A. (2007) The expanding RNA polymerase III In vitro analysis of alpha-amanitin-resistant transcription transcriptome. Trends Genet 23: 614–622. from the rRNA, procyclic acidic repetitive protein, and Fantoni, A., Dare, A.O., and Tschudi, C. (1994) RNA poly- variant surface glycoprotein gene promoters in Trypano- merase III-mediated transcription of the trypanosome U2 soma brucei. Mol Cell Biol 19: 5466–5473.

VC 2016 John Wiley & Sons Ltd, Molecular Microbiology, 103, 452–468 Characterization of Maf1 in T. brucei 467

Lecordier, L., Devaux, S., Uzureau, P., Dierick, J.F., Hernandez-Rivas, R., et al. (2015) The Selenocysteine Walgraffe, D., Poelvoorde, P., et al. (2007) Characteriza- tRNA Gene in Leishmania major Is Transcribed by both tion of a TFIIH homologue from Trypanosoma brucei. Mol RNA Polymerase II and RNA Polymerase III. Eukaryot Microbiol 64: 1164–1181. Cell 14: 216–227. Lee, J.H., Nguyen, T.N., Schimanski, B., and Gunzl, A. Palenchar, J.B., Liu, W., Palenchar, P.M., and Bellofatto, V. (2007) Spliced leader RNA gene transcription in Trypano- (2006) A divergent transcription factor TFIIB in trypanosoma brucei requires transcription factor TFIIH. Eukaryot somes is required for RNA polymerase II-dependent Cell 6: 641–649. spliced leader RNA transcription and cell viability. Eukar- Lee, J.H., Cai, G., Panigrahi, A.K., Dunham-Ems, S., yot Cell 5: 293–300. Nguyen, T.N., Radolf, J.D., et al. (2010) A TFIIH- Palian, B.M., Rohira, A.D., Johnson, S.A., He, L., Zheng, associated mediator head is a basal factor of small N., Dubeau, L., et al. (2014) Maf1 is a novel target of nuclear spliced leader RNA gene transcription in early- PTEN and PI3K signaling that negatively regulates onco- diverged trypanosomes. Mol Cell Biol 30: 5502–5513. genesis and lipid metabolism. PLoS Genet 10: Lee, Y.L., Li, Y.C., Su, C.H., Chiao, C.H., Lin, I.H., and Hsu, e1004789. M.T. (2015) MAF1 represses CDKN1A through a Pol III- Park, S.H., Nguyen, T.N., Kirkham, J.K., Lee, J.H., and dependent mechanism. Elife 4: e06283. Gunzl, A. (2011) Transcription by the multifunctional RNA Martinez-Calvillo, S., Yan, S., Nguyen, D., Fox, M., Stuart, polymerase I in Trypanosoma brucei functions independ- K., and Myler, P.J. (2003) Transcription of Leishmania ently of RPB7. Mol Biochem Parasitol 180: 35–42. major Friedlin chromosome 1 initiates in both directions Pays, E., Vanhamme, L., and Perez-Morga, D. (2004) Anti- within a single region. Mol Cell 11: 1291–1299. genic variation in Trypanosoma brucei: facts, challenges Martinez-Calvillo, S., Vizuet-de-Rueda, J.C., Florencio- and mysteries. Curr Opin Microbiol 7: 369–374. Martinez, L.E., Manning-Cela, R.G., and Figueroa- Pluta, K., Lefebvre, O., Martin, N.C., Smagowicz, W.J., Angulo, E.E. (2010) Gene expression in trypanosomatid Stanford, D.R., Ellis, S.R., et al. (2001) Maf1p, a negative parasites. J Biomed Biotechnol 2010: 525241. effector of RNA polymerase III in Saccharomyces cerevi- Michaeli, S. (2011) Trans-splicing in trypanosomes: machin- siae. Mol Cell Biol 21: 5031–5040. ery and its impact on the parasite transcriptome. Fut Rathore, O.S., Faustino, A., Prudencio, P., Van, D.P., Cox, Microbiol 6: 459–474. C.J., and Martinho, R.G. (2016) Absence of N-terminal Michels, A.A. (2011) MAF1: a new target of mTORC1. Bio- acetyltransferase diversification during evolution of chem Soc Trans 39: 487–491. eukaryotic organisms. Sci Rep 6: 21304. Moir, R.D., and Willis, I.M. (2013) Regulation of pol III tran- Reina, J.H., Azzouz, T.N., and Hernandez, N. (2006) Maf1, scription by nutrient and stress signaling pathways. Bio- a new player in the regulation of human RNA polymerase chim Biophys Acta 1829: 361–375. III transcription. PLoS ONE 1: e134. Moir, R.D., Lee, J., Haeusler, R.A., Desai, N., Engelke, Rideout, E.J., Marshall, L., and Grewal, S.S. (2012) Dro- D.R., and Willis, I.M. (2006) Protein kinase A regulates sophila RNA polymerase III repressor Maf1 controls body RNA polymerase III transcription through the nuclear size and developmental timing by modulating tRNAiMet localization of Maf1. Proc Natl Acad Sci U S A 103: synthesis and systemic insulin signaling. Proc Natl Acad 15044–15049. Sci U S A 109: 1139–1144. Morawiec, E., Wichtowska, D., Graczyk, D., Conesa, C., Roberts, D.N., Wilson, B., Huff, J.T., Stewart, A.J., and Lefebvre, O., and Boguta, M. (2013) Maf1, repressor of Cairns, B.R. (2006) Dephosphorylation and genome-wide tRNA transcription, is involved in the control of gluconeo- association of Maf1 with Pol III-transcribed genes during genetic genes in Saccharomyces cerevisiae. Gene 526: repression. Mol Cell 22: 633–644. 16–22. Rohira, A.D., Chen, C.Y., Allen, J.R., and Johnson, D.L. Nakaar, V., Dare, A.O., Hong, D., Ullu, E., and Tschudi, C. (2013) Covalent small ubiquitin-like modifier (SUMO) (1994) Upstream tRNA genes are essential for expres- modification of Maf1 protein controls RNA polymerase III- sion of small nuclear and cytoplasmic RNA genes in try- dependent transcription repression. J Biol Chem 288: panosomes. Mol Cell Biol 14: 6736–6742. 19288–19295. Nguyen, T.N., Schimanski, B., and Gunzl, A. (2007) Active Rollins, J., Veras, I., Cabarcas, S., Willis, I., and Schramm, RNA polymerase I of Trypanosoma brucei harbors a L. (2007) Human Maf1 negatively regulates RNA poly- novel subunit essential for transcription. Mol Cell Biol 27: merase III transcription via the TFIIB family members 6254–6263. Brf1 and Brf2. Int J Biol Sci 3: 292–302. Oficjalska-Pham, D., Harismendy, O., Smagowicz, W.J., Ruan, J.P., Arhin, G.K., Ullu, E., and Tschudi, C. (2004) Gonzalez de, P.A., Boguta, M., Sentenac, A., and Functional characterization of a Trypanosoma brucei Lefebvre, O. (2006) General repression of RNA polymer- TATA-binding protein-related factor points to a universal ase III transcription is triggered by protein phosphatase regulator of transcription in trypanosomes. Mol Cell Biol type 2A-mediated dephosphorylation of Maf1. Mol Cell 24: 9610–9618. 22: 623–632. Rudenko, G. (2010) Epigenetics and transcriptional control Oler, A.J., and Cairns, B.R. (2012) PP4 dephosphorylates in African trypanosomes. Essays Biochem 48: 201–219. Maf1 to couple multiple stress conditions to RNA poly- Schimanski, B., Laufer, G., Gontcharova, L., and Gunzl, A. merase III repression. Embo J 31: 1440–1452. (2004) The Trypanosoma brucei spliced leader RNA and Padilla-Mejia, N.E., Florencio-Martinez, L.E., Moreno- rRNA gene promoters have interchangeable TbSNAP50- Campos, R., Vizuet-de-Rueda, J.C., Cevallos, A.M., binding elements. Nucleic Acids Res 32: 700–709.

VC 2016 John Wiley & Sons Ltd, Molecular Microbiology, 103, 452–468 468 G. Romero-Meza et al. ᭿

Schimanski, B., Nguyen, T.N., and Gunzl, A. (2005a) nia tarentolae TATA box-binding protein LtTBP associates Characterization of a multisubunit transcription factor with a subset of promoters. Int J Parasitol 36: 1217–1226. complex essential for spliced-leader RNA gene tran- Thomas, S., Green, A., Sturm, N.R., Campbell, D.A., and scription in Trypanosoma brucei. MolCellBiol25: 7303– Myler, P.J. (2009) Histone acetylations mark origins of 7313. polycistronic transcription in Leishmania major. BMC Schimanski, B., Nguyen, T.N., and Gunzl, A. (2005b) Highly Genom 10: 152. efficient tandem affinity purification of trypanosome pro- Vannini, A., Ringel, R., Kusser, A.G., Berninghausen, O., tein complexes based on a novel epitope combination. Kassavetis, G.A., and Cramer, P. (2010) Molecular basis Eukaryot Cell 4: 1942–1950. of RNA polymerase III transcription repression by Maf1. Schimanski, B., Brandenburg, J., Nguyen, T.N., Caimano, Cell 143: 59–70. M.J., and Gunzl, A. (2006) A TFIIB-like protein is indis- Velez-Ramirez, D.E., Florencio-Martinez, L.E., Romero-Meza, pensable for spliced leader RNA gene transcription in G., Rojas-Sanchez, S., Moreno-Campos, R., Arroyo, R., Trypanosoma brucei. Nucleic Acids Res 34: 1676–1684. et al. (2015) BRF1, a subunit of RNA polymerase III tran- Sherman, D.R., Janz, L., Hug, M., and Clayton, C. (1991) scription factor TFIIIB, is essential for cell growth of Tr ypa- Anatomy of the parp gene promoter of Trypanosoma bru- nosoma brucei. Parasitology 142: 1563–1573. cei. Embo J 10: 3379–3386. Vizuet-de-Rueda, J.C., Florencio-Martinez, L.E., Padilla- Shi, H., Chamond, N., Tschudi, C., and Ullu, E. (2004) Mejia, N.E., Manning-Cela, R., Hernandez-Rivas, R., and Selection and characterization of RNA interference- Martinez-Calvillo, S. (2016) Ribosomal RNA genes in the deficient trypanosomes impaired in target mRNA degra- protozoan parasite Leishmania major possess a nucleo- dation. Eukaryot Cell 3: 1445–1453. somal structure. Protist 167: 121–135. Shor, B., Wu, J., Shakey, Q., Toral-Barza, L., Shi, C., Wang, Z., Morris, J.C., Drew, M.E., and Englund, P.T. Follettie, M., and Yu, K. (2010) Requirement of the (2000) Inhibition of Trypanosoma brucei gene expression mTOR kinase for the regulation of Maf1 phosphorylation by RNA interference using an integratable vector with and control of RNA polymerase III-dependent transcrip- opposing T7 promoters. J Biol Chem 275: 40174–40179. tion in cancer cells. J Biol Chem 285: 15380–15392. White, R.J. (2008) RNA polymerases I and III, non-coding Siegel, T.N., Hekstra, D.R., Kemp, L.E., Figueiredo, L.M., RNAs and cancer. Trends Genet 24: 622–629. Lowell, J.E., Fenyo, D., et al. (2009) Four histone variants Wirtz, E., Leal, S., Ochatt, C., and Cross, G.A. (1999) A mark the boundaries of polycistronic transcription units in tightly regulated inducible expression system for condi- Trypanosoma brucei. Genes Dev 23: 1063–1076. tional gene knock-outs and dominant-negative genetics in Soprano, A.S., Abe, V.Y., Smetana, J.H., and Benedetti, Trypanosoma brucei. Mol Biochem Parasitol 99: 89–101. C.E. (2013) Citrus MAF1, a repressor of RNA polymerase III, binds the Xanthomonas citri canker elicitor PthA4 and suppresses citrus canker development. Plant Physiol Supporting information 163: 232–242. Thomas, S., Yu, M.C., Sturm, N.R., and Campbell, D.A. Additional supporting information may be found in the (2006) A non-universal transcription factor? The Leishma- online version of this article at the publisher’s web-site.

VC 2016 John Wiley & Sons Ltd, Molecular Microbiology, 103, 452–468