Participation of TFIIIB Subunit Brf1 in Transcription Regulation in the Human Pathogen Leishmania major

In yeast and higher eukaryotes, transcription factor TFIIIB is required for accurate initiation of transcription by RNA Polymerase III (Pol III), which synthesizes transfer RNAs (tRNAs), 5S ribosomal RNA (rRNA), and other essential RNA molecules. TFIIIB is composed of three subunits: B double prime 1 (Bdp1), TATA-binding protein (TBP), and TFIIB-related factor 1 (Brf1). Here, we report the molecular characterization of Brf1 in Leishmania major (LmBrf1), a parasitic protozoan that shows distinctive transcription characteristics, including the apparent absence of Pol III general transcription factors TFIIIA and TFIIIC. Although single-knockout parasites of LmBrf1 were obtained, attempts to generate LmBrf1-null mutants were unsuccessful, which suggests that LmBrf1 is essential in promastigotes of L. major. Notably, Northern blot analyses showed that the half-lives of the messenger RNAs (mRNAs) from LmBrf1 and other components of the Pol III transcription machinery (Bdp1 and Pol III subunit RPC1) are very similar (~40 min). Stabilization of these transcripts was observed in stationary-phase parasites. Chromatin immunoprecipitation (ChIP) experiments showed that LmBrf1 binds to tRNA, small nuclear RNA (snRNA), and 5S rRNA genes. Unexpectedly, the results also indicated that LmBrf1 associates to the promoter region of the 18S rRNA genes and to three Pol II-dependent regions here analyzed. Tandem affinity purification and mass spectrometry analyses allowed the identification of a putative TFIIIC subunit. Moreover, several proteins involved in transcription by all three RNA polymerases co-purified with the tagged version of LmBrf1.


Introduction
More than 20 different species of Leishmania produce the three main forms of leishmaniases (cutaneous, mucocutaneous, and visceral) in tropical and subtropical regions of the world. These protozoan parasites are transmitted to humans through the bite of infected female sandflies of the genera Phlebotomus and Lutzomyia. Leishmania and other trypanosomatids, such as Trypanosoma cruzi and Trypanosoma brucei, are ancient organisms that diverged early from the main eukaryotic lineage [1]. Consequently, their genes are expressed by atypical mechanisms, which include polycistronic transcription and trans-splicing [2][3][4].
Three distinctive classes of RNA polymerase (Pol) participate in transcription of nuclear DNA in eukaryotic cells: Pol I, II, and III [5]. Pol I is responsible for the synthesis of 18S, 5.8S, and 25-28S ribosomal RNAs (rRNAs) [6], whereas Pol II catalyzes the production of messenger RNAs (mRNAs), and most small nuclear RNAs (snRNAs) and microRNAs [7].

Generation of Plasmids
To produce vector p∆Brf1-pac for knockout experiments, the pac gene (1.7 kb) was obtained by digesting the p∆75-pac plasmid [38] with EcoRI and SacI. The 5 -targeting region from LmBrf1 (LmjF. 25.0440) (531 bp) was amplified by PCR with primers BRF1-5-for-XbaI (5 -ATCTAGAACGGCACACATCTACTCGCG) and BRF1-5-rev-EcoRI (5 -AGAATTCGGT GAAGCGTCAACTGCTGG). The 3 -targeting region (508 bp) was amplified with oligonucleotides BRF1-3-for-SacI (5 -AGAGCTCCGAATACGCTGCCAGAGTAG) and BRF1-3rev-XhoI (5 -ACTCGAGAGCTTTCGTCTCTTAGGCCC). To generate vector p∆Brf1-hyg, the hyg gene (1.8 kb) was obtained from the p∆75-hyg plasmid [38] after SpeI and SacI digestion. The 5 -targeting region was amplified with primers BRF1-5-for-XbaI and BRF1-5-rev-SpeI (5 -AACTAGTGGTGAAGCGTCAACTGCTGG), and the 3 -targeting region with oligonucleotides BRF1-3-for-SacI and BRF1-3-rev-XhoI. The PCR products from 5 and 3 targeting regions were digested with the restriction enzymes indicated in the name of each oligonucleotide. For both knockout vectors, the selectable-marker gene and the two flanking regions were cloned into pBluescript II KS (digested with XbaI and XhoI), by a single four-fragment ligation step. The correct order of the fragments was verified by sequencing. The targeting cassettes were obtained from p∆Brf1-pac and p∆Brf1-hyg by digesting with XbaI and XhoI. To generate the LmBrf1-PTP vector, for immunofluores-cence, tandem affinity purifications, and ChIP assays, the RPB6 gene from the pB6-PTP plasmid [29] was substituted with the LmBrf1 gene. For this purpose, pB6-PTP was digested with XmaI and XbaI, and the RBP6 gene was eliminated by agarose gel electrophoresis. The LmBrf1 gene (without the terminal codon) was amplified with primers BRF-XmaI-5 (5 -ATCCCGGGATGTCCAGCTGCACCCATCCCA) and BRF-XbaI-3 (5 -ATTCTAGAGTCCCAGTCCGCCTGTGGCTCA). The PCR product was first cloned into the pGEM-T Easy vector (Promega), digested with XmaI and XbaI, and then ligated into the pB6-PTP backbone. The vector was sequenced to confirm that the LmBrf1 gene and the tag were in frame.

Indirect Immunofluorescence Assays
The subcellular localization of LmBrf1 labeled with a PTP tag was determined by indirect immunofluorescence, as previously described [29,39]. Parasites fixed with paraformaldehyde were stained for double indirect immunofluorescence analysis using a mixture of primary antibodies against Prot C (Delta Biolabs) (to detect LmBrf1-PTP) and LmNop56, followed by anti-rabbit immunoglobulin G (IgG) conjugated with Alexa Fluor 568 (red) and anti-mouse IgG coupled with Alexa Fluor 488 (green) secondary antibodies. DNA was stained with 4',6-diamidino-2-phenylindole (DAPI) (blue). Individual optical sections were obtained with a Nikon A1R+ STORM confocal microscope, analyzed, and prepared for presentation using the ZEN 2012 software (Blue edition).

Chromatin Immunoprecipitation Assays
The ChIP procedures were performed as previously described [41]. Briefly, 1.2 × 10 8 promastigotes were cross-linked with formaldehyde (final concentration of 1%) for 5 min at 37 • C. A Vibra-Cell VCX130 ultrasonic processor (Sonics) was employed to lyse the cells (15 s on/off, 40% amplitude, for 5 min). Nuclei were pelleted and resuspended in sonication buffer (1% SDS, 10 mM EDTA, and 50 mM Tris-HCl, pH 8.0; 1× protease inhibitors). Chromatin was sonicated to an average DNA size of about 200 to 500 bp with a BioRuptor UCD-200 (Diagenode) (30 s on/30 s off, high intensity) for 30 cycles. The sonicated material was pre-cleared by adding protein A/G plus-agarose beads (Santa Cruz Biotechnology) and mixing for 1 h at 4 • C. Chromatin samples were incubated overnight at 4 • C with rabbit anti-Prot A antibody (Sigma) or nonspecific rabbit immune serum (negative control). Protein-DNA complexes were incubated for 2 h with protein A/G plus-agarose beads and 20 µg of sonicated salmon sperm DNA, and then washed as previously described [42]. Cross-links were reversed with 200 mM NaCl at 65 • C overnight. Samples were treated with RNase A and proteinase K. DNA was precipitated with sodium acetate and ethanol and quantified. Each ChIP experiment was performed at least three times.

Northern and Southern Blot Analyses
For determination of the half-life of the different mRNAs, Northern blot experiments were performed with RNA isolated at several time points from cells incubated in the presence of the transcription inhibitor actinomycin D (10 µg/mL). Total RNA was extracted with TRI reagent (Sigma), separated in agarose-formaldehyde gels and transferred to Hybond-N filters (Amersham). Hybridization was performed with probes that were labeled with [α-32 P]dCTP using the High Prime labeling system (Amersham). The Brf1 probe was amplified with primers BRF-XmaI-5 and BRF-XbaI-3 . The Bdp1 probe was amplified with oligonucleotides B"-BamHI-5 (5 -ATGGATCCATGGACGACAACGAGTTCGAA) and B"-XbaI-3 (5 -ATTCTAGACTCAAACGAGAAGTCCGAGTC) and the RPC1 probe with primers C160-1-5 (5 -GATCATCAACGCCAACAAGAGC) and C160-2-3 (5 -TTGAGCAGGATGTTGAA GCTGC). The α-tubulin probe was amplified with primers Alfa-tub-5 (5 -AGAAGTCCAA GCTCGGCTACAC) and Alfa-tub-3 (5 -GTAGTTGATGCCGCACTTGAAG), and the 18S rRNA probe with oligonucleotides Lm-rRNA18S 5 (5 -CGGCCTCTAGGAATGAAGG) and Lm-rRNA18S 3 (5 -CCCCTGAGACTGTAACCTC). Filters were washed to a final stringency of 0.1× SSC and 0.1% SDS at 65 • C. RNA signals from three independent experiments were quantified by densitometry using the MultiGauge software and plotted. Half-lives were determined from the linear regression estimates derived from the linear segment of the mRNA decay curves. For Southern blot analysis, 5 µg of genomic DNA was digested, separated by electrophoresis, and transferred by capillary to Hybond-N membranes. The radioactive probe corresponded to the 5 -targeting region of Brf1. Filters were washed as described above.

Tandem Affinity Purifications and Mass Spectrometry Analysis
Tandem affinity purification experiments were performed, as previously described [43,44], using whole-cell extracts obtained from mid-log phase promastigotes (3 L at 3-4 × 10 7 cells per mL). Cells were harvested, washed with phosphate-buffered saline (PBS), and resuspended in 14 mL of IPP-150 buffer (10 mM Tris-HCl, pH 8.0, 150 mM NaCl, and 0.1% NP-40) that contained two tablets of Complete Mini, EDTA-free protease inhibitor cocktail (Roche Molecular). Then, 2 mL of 10% Triton X-100 was added and the samples were maintained on ice for around 20 min, until cells were totally lysed. To obtain the cleared lysate, samples were centrifuged at 10,000× g for 15 min at 4 • C. Then, the extract was added to a 20 mL disposable column, in which 200 µL of IgG Sepharose 6 Fast Flow beads (GE Healthcare) was equilibrated with IPP-150 buffer and incubated with rotation for 3 h at 4 • C. The column was drained by gravity flow and washed three times with 20 mL of cold IPP-150 buffer. PTP-tagged proteins were eluted by resuspending the beads in 2 mL of TEV cleavage buffer (20 mM Tris-HCl, pH 7.7, 150 mM KCl, 3 mM MgCl 2 , 0.5 mM dithiothreitol (DTT), 0.1% Tween-20, and 0.5 mM EDTA) with 200 U of tobacco etch virus (TEV) protease (Sigma), and incubating overnight at 4 • C. The IgG-Sepharose was drained and the eluted material was mixed with 4 mL of PC-150 buffer (20 mM Tris-HCl, pH 7.7, 150 mM KCl, 3 mM MgCl 2 , 0.5 mM DTT, 0.1% Tween-20, and 2× protease inhibitors). CaCl 2 was added to the eluate to a final concentration of 3 mM before transfer to a 10 mL column containing 200 µL of anti-Prot C affinity matrix (Roche). Samples were incubated at 4 • C for 4 h with rotation, washed with 30 mL of PC-150, and Prot C-tagged proteins were eluted with EGTA elution buffer (5 mM Tris-HCl pH 7.7, 10 mM EGTA, 5 mM EDTA, and 2× protease inhibitors). Eluted proteins were concentrated with Amicon Ultracel 3K columns (Millipore) or by evaporation in a vacuum concentrator and analyzed by SDS-PAGE and SYPRO Ruby (Molecular Probes) staining. Individual lanes from the gels were sliced into two pieces and proteins subjected to in-gel tryptic digestion prior to liquid chromatography-mass spectrometry/mass spectrometry (LC-MS/MS) at the Core Facility for Proteomics and Mass Spectrometry from Upstate Medical University. The collision-induced dissociation (CID) spectra were compared with the L. major protein database from TriTrypDB page.

The Sequence and Predicted Structure of LmBrf1 Are Conserved
LmjF. 25.0440 was previously identified as the L. major ortholog of Brf1 (LmBrf1) [35]. A sequence comparison between LmBrf1 and its ortholog in Saccharomyces cerevisiae (ScBrf1) showed a relatively low sequence identity of 20%. Similarity was slightly higher in the N-terminal half of the protein (30%) than in the C-terminal half (18%). Nevertheless, the in silico analysis demonstrated that LmBrf1 contains the three conserved domains located in the N-terminal half of the protein: a zinc ribbon motif and two imperfect cyclin repeats (also known as TFIIB-related repeats) ( Figure 1A). Moreover, LmBrf1 also contains moderately conserved Homology Blocks I, II, and III, which are characteristic of the C-terminal region of Brf1 from yeast species and human ( Figure 1A) [45]. Notably, the predicted size of the LmBrf1 (77.2 kDa) is more than 10 kDa larger than ScBrf1 (66.9 kDa), due to insertions in the C-terminal region of the former. These insertions are present in other species of Leishmania, but absent in T. cruzi and T. brucei ( Figure S1, Supplementary Materials). moderately conserved Homology Blocks I, II, and III, which are characteristic of the Cterminal region of Brf1 from yeast species and human ( Figure 1A) [45]. Notably, the predicted size of the LmBrf1 (77.2 kDa) is more than 10 kDa larger than ScBrf1 (66.9 kDa), due to insertions in the C-terminal region of the former. These insertions are present in other species of Leishmania, but absent in T. cruzi and T. brucei ( Figure S1 25.0440) and Saccharomyces cerevisiae (Sc, CAA68968.1). Please note that the last two lines of the alignment (Homology Block II and Homology Block III) are not contiguous with each other and with the rest of the alignment. Conservation is denoted by black shading, conserved substitutions are indicated by dark-gray shading, and semiconserved substitutions are denoted by light-gray shading, according to the ClustalΩ program. The zinc ribbon, both cyclin repeats, and the Brf1 homology blocks I, II, and III are indicated. Predicted secondary structure elements are shown for L. major (above the sequence) and for S. cerevisiae (below the sequence). The β-strands are denoted by yellow arrows, and α-helices by rounded rectangles. The five α-helices of the first and the second cyclin repeats are shown in the same colors. (B) Predicted three-dimensional structure of the N-terminal region of LmBrf1. Homology modeling was performed for LmBrf1 using the crystal structure from ScBrf1 as a template. A merge figure is also shown. All the structures are shown in the same colors presented in panel A. The quality of the models was estimated with Mod Eval server, showing a score of 0.70 (0.89 if we compare only the first and second cyclin repeats).
The zinc ribbon fold in ScBrf1 consists of three antiparallel β strands [46]. Similarly, the zinc-binding domain in LmBrf1 is predicted to fold into three β strands, while the rest of the protein is mainly composed of α-helices ( Figure 1A). Each cyclin repeat in LmBrf1 is predicted to fold into five α-helices ( Figure 1A), as reported in other organisms [18,47]. The structure of LmBrf1 was further analyzed by generating the hypothetical three-dimensional structure of the N-terminal region (zinc ribbon, cyclin repeats, and homology block I) by homology modeling, using the crystal structure of ScBrf1 as a template. As shown in Figure 1B, the zinc ribbon in LmBrf1 exhibits the typical structure of the three antiparallel β strands, and the cyclin repeats present the characteristic globular organization of the five α-helices. Moreover, the hypothetical architecture of homology block I, folded into one helix, is conserved in LmBrf1 ( Figure 1B). Thus, the predicted structure of LmBrf1 is similar to that of ScBrf1.

LmBRF1 Is a Nuclear Protein
To examine the subcellular localization of LmBrf1, a cell line where this protein is labeled with a C-terminal PTP tag was obtained to carry out indirect immunofluorescence assays. This tag is composed of Protein A (Prot A) and Protein C (Prot C) epitopes separated by a cleavage site of the TEV protease [43]. The expression of the LmBrf1-PTP protein was confirmed by Western blot analysis with an anti-Prot C antibody, showing a protein of~97 kDa (expected from the fusion of the 20 kDa PTP tag to the~77 kDa LmBrf1) ( Figure 2A). The same antibody was used for the immunofluorescence experiments on fixed and permeabilized parasites. As shown in Figure 2B, LmBrf1 is localized in the nucleus of the parasites, as anticipated for a transcription factor. The nuclear distribution of LmBrf1 is very similar to that of LmBdp1 [33]. Notably, a small fraction of LmBrf1 seems to colocalize with the nucleolar protein Nop56 in the periphery of the nucleolus (in yellow in the Merge panels of Figure 2B).

The Half-Lives of the mRNAs from LmBrf1 and Other Components of the Pol III Transcription Machinery Are Very Similar
In order to determine the decay rate of the LmBrf1 mRNA, Northern blot experiments were performed with total RNA from mid-log parasites that were grown in the presence of actinomycin D (10 mg/mL) for different times (between 0 and 16 h). The half-life of the LmBrf1 mRNA was around 39 min ( Figure 3). Notably, the half-lives of the mRNAs from LmBdp1 (other TFIIIB subunit) and LmRPC1 (the largest Pol III subunit) were~36 min and 40 min, respectively ( Figure 3). Thus, the decay rate of these three proteins involved in Pol III transcription is very similar. In contrast, a much longer half-life of~4.3 h was observed for the α-tubulin mRNA ( Figure 3). Similar experiments performed with stationary phase parasites showed that the half-lives of LmBrf1, LmBdp1, and LmRPC1 were~60 min, 72 min, and~62 min, respectively ( Figure 3, right panels). Thus, an increase in the halflives of all three mRNAs was observed in stationary-phase promastigotes. This is different from the α-tubulin mRNA, whose half-life was reduced to~2.25 h in nonproliferative cells ( Figure 3). ) and wild-type cells (WT). Membranes were incubated with an antibody against Prot C, which recognizes the PTP tag. As a loading control, a human α/β-tubulin antibody was used. (B) PTP-tagged LmBrf1 was detected by immunofluorescence with a rabbit anti-Prot C polyclonal antibody and an anti-rabbit immunoglobulin G (IgG) coupled with Alexa Fluor 568 (red). Nop56 was detected with a mouse anti-Nop56 antibody and an anti-mouse IgG secondary antibody conjugated with Alexa Fluor 488 (green). Nuclei (N) and kinetoplast (K) DNA were stained with 4',6-diamidino-2-phenylindole (DAPI; blue). Scale bar indicates 2 m.

The Half-Lives of the mRNAs from LmBrf1 and Other Components of the Pol III Transcription Machinery Are Very Similar
In order to determine the decay rate of the LmBrf1 mRNA, Northern blot experiments were performed with total RNA from mid-log parasites that were grown in the presence of actinomycin D (10 mg/mL) for different times (between 0 and 16 h). The halflife of the LmBrf1 mRNA was around 39 min ( Figure 3). Notably, the half-lives of the mRNAs from LmBdp1 (other TFIIIB subunit) and LmRPC1 (the largest Pol III subunit) were ~36 min and ~40 min, respectively ( Figure 3). Thus, the decay rate of these three proteins involved in Pol III transcription is very similar. In contrast, a much longer halflife of ~4.3 h was observed for the α-tubulin mRNA ( Figure 3). Similar experiments performed with stationary phase parasites showed that the half-lives of LmBrf1, LmBdp1, and LmRPC1 were ~60 min, ~72 min, and ~62 min, respectively ( Figure 3, right panels). Thus, an increase in the half-lives of all three mRNAs was observed in stationary-phase promastigotes. This is different from the α-tubulin mRNA, whose half-life was reduced to ⁓2.25 h in nonproliferative cells (Figure 3).

Promastigotes Need at Least One Copy of the LmBrf1 Gene to Survive
To study the role of LmBrf1 on cell growth and transcription in L. major, we planned to produce LmBrf1-null mutant parasites by replacing the endogenous genes with selectable marker genes. Since L. major Friedlin is a diploid organism [48], two consecutive rounds of targeted gene disruption were intended to generate double-knockout parasites. To fulfill this purpose, we produced plasmids p∆LmBrf1-pac and p∆LmBrf1-hyg, where the puromycin (pac) and hygromycin (hyg) resistance genes were bounded by 5 and 3 flanking regions of LmBrf1.
The single-knockout cell line for LmBrf1 was generated by transfecting wild-type promastigotes with the targeting cassette from p∆LmBrf1-pac, and clones were selected with puromycin. The correct substitution of one LmBrf1 allele with the pac gene was demonstrated in one of the clones by Southern blot analysis using the 5 -targeting region as a probe, as the expected bands of~1158 and~5251 bp were observed when digesting with SalI and NotI, respectively ( Figure 4). The bands of the remaining endogenous LmBrf1 gene were also observed:~889 bp with SalI and 4134 bp with NotI ( Figure 4). Thus, these results demonstrate that one copy of LmBrf1 was replaced with the pac gene. It is worth noting that the single-knockout cell line did not show any growth defect in relation to wild-type cells.
To generate the double-knockout cell line for LmBrf1, the single-knockout clone was transfected with the targeting cassette from vector p∆LmBrf1-hyg, and the culture was selected with both puromycin and hygromycin. However, we were unable to generate null mutants of LmBrf1, as all the transfected cultures died during the selection process.
To verify that the hyg cassette functions properly, it was used to transfect wild-type promastigotes, and clones were successfully obtained in media with hygromycin. The correct substitution of one LmBrf1 allele with the hyg gene was shown by PCR analyses. Therefore, single-knockout parasites could be generated with both cassettes. This culture was electroporated with the pac cassette, but we did not obtain transfected cells, again. Thus, these results suggest that promastigotes of L. major require at least one copy of the LmBrf1 gene in order to survive.

Promastigotes Need at Least One Copy of the LmBrf1 Gene to Survive
To study the role of LmBrf1 on cell growth and transcription in L. major, we planned promastigotes with the targeting cassette from pΔLmBrf1-pac, and clones were selected with puromycin. The correct substitution of one LmBrf1 allele with the pac gene was demonstrated in one of the clones by Southern blot analysis using the 5'-targeting region as a probe, as the expected bands of ~1158 and ~5251 bp were observed when digesting with SalI and NotI, respectively (Figure 4). The bands of the remaining endogenous LmBrf1 gene were also observed: ~889 bp with SalI and 4134 bp with NotI ( Figure 4). Thus, these results demonstrate that one copy of LmBrf1 was replaced with the pac gene. It is worth noting that the single-knockout cell line did not show any growth defect in relation to wild-type cells.
. To generate the double-knockout cell line for LmBrf1, the single-knockout clone was transfected with the targeting cassette from vector pΔLmBrf1-hyg, and the culture was selected with both puromycin and hygromycin. However, we were unable to generate null mutants of LmBrf1, as all the transfected cultures died during the selection process. To verify that the hyg cassette functions properly, it was used to transfect wild-type promastigotes, and clones were successfully obtained in media with hygromycin. The correct substitution of one LmBrf1 allele with the hyg gene was shown by PCR analyses. Therefore, single-knockout parasites could be generated with both cassettes. This culture was electroporated with the pac cassette, but we did not obtain transfected cells, again. Thus, these results suggest that promastigotes of L. major require at least one copy of the LmBrf1 gene in order to survive.  As an alternative procedure, the cell line that expresses the LmBrf1-PTP recombinant protein was transfected with the targeting cassette from p∆LmBrf1-hyg, and LmBrf1 singleknockout parasites were obtained after selecting with hygromycin. This cell line was then transfected with the pac cassette to try to generate the double-knockout parasites. However, we could not delete the second endogenous copy of LmBrf1, perhaps due to technical reasons that could have prevented the efficient transfection and/or drug selection of the electroporated parasites. Alternatively, it is possible that the expression of the recombinant LmBrf1-PTP protein was not high enough to allow the removal of the second allelic copy of LmBrf1, or that the presence of the PTP-tag might have reduced the activity of LmBrf1. Subsequent complementation experiments with new vectors will help to clarify this issue.

LmBrf1 Binds to Pol III Promoters and Other Genomic Regions
The in vivo association of LmBrf1 to Pol III promoters was analyzed by ChIP experiments, using the L. major cell line that expresses LmBrf1 with a C-terminal PTP fusion. Chromatin was precipitated with a ChIP-grade anti-Prot A antibody that recognizes the two Prot A domains present in the PTP tag [49], and with a nonspecific mouse immune serum as negative control. The binding of LmBrf1-PTP to the purified DNA was evaluated by qPCR assays.
Results obtained from three independent experiments demonstrated that LmBrf1 associates with the tRNA-like and, to a lesser extent, to the tRNA-Ala, which are required for the expression of the U2 snRNA gene ( Figure 5). Enrichment was also detected within the U2 snRNA gene, whose first nucleotides are needed for Pol III transcription of the gene [27]. Similar results were observed in the U4 snRNA locus, as high enrichment was detected in the upstream tRNA-like region, and lower association was found in the U4 snRNA gene ( Figure 5). We also analyzed a genomic region from chromosome 11 that contains a 5S rRNA gene and two tRNA genes (Met and Arg) embedded within a Pol II transcription unit. High association of LmBrf1 was observed in the 5S rRNA and the tRNA-Met gene.
Interestingly, occupancy of LmBrf1 was found in the LmjF.11.0930 gene (which encodes a hypothetical protein and is transcribed by Pol II) and in the intergenic region between this gene and the 5S rRNA gene (Inter 11.0930) ( Figure 5). Unexpectedly, association of LmBrf1 was found in the SSR from chromosome 1 (which controls transcription of all the protein-coding genes located on this chromosome) [50] and the spliced-leader (SL) RNA loci (transcribed by Pol II), as well as in the promoter region of the rRNA unit (transcribed by Pol I) ( Figure 5).

Proteins That Participate in Pol I, Pol II, and Pol III Transcription Were Co-Purified with LmBrf1-PTP
To identify proteins that interact with LmBrf1, tandem affinity purification experiments were performed with the cell line that expresses the PTP-tagged version of LmBrf1. The correct expression of LmBrf1-PTP was confirmed by Western blotting and immunofluorescence ( Figure 2). LmBrf1-PTP and associated proteins were purified, in successive steps, by IgG affinity chromatography, TEV protease elution, and anti-Prot C affinity chromatography. The purified proteins were eluted from the anti-Prot C column with a buffer that contains EGTA, and they were concentrated and analyzed by SDS-PAGE ( Figure 6). Several different bands were observed, including one whose size corresponds to the tagged LmBrf1 (~81 kDa after TEV digestion, labeled with an asterisk in Figure 6). In the control experiment (mock purification using wild-type extract), some bands were also detected, which were identified by mass spectrometry as human keratins, bovine serum albumin, and multiple L. major ribosomal proteins, heat-shock proteins, translation elongation factors, actin, and αand β-tubulins. These proteins are common contaminants in tandem affinity purifications [51]. steps, by IgG affinity chromatography, TEV protease elution, and anti-Prot C affinity chromatography. The purified proteins were eluted from the anti-Prot C column with a buffer that contains EGTA, and they were concentrated and analyzed by SDS-PAGE ( Figure 6). Several different bands were observed, including one whose size corresponds to the tagged LmBrf1 (⁓81 kDa after TEV digestion, labeled with an asterisk in Figure 6). In the control experiment (mock purification using wild-type extract), some bands were also detected, which were identified by mass spectrometry as human keratins, bovine serum albumin, and multiple L. major ribosomal proteins, heat-shock proteins, translation elongation factors, actin, and α-and β-tubulins. These proteins are common contaminants in tandem affinity purifications [51]. In addition to LmBrf1, mass spectrometry analysis of the LmBrf1-PTP sample allowed the identification of multiple proteins that were divided in seven categories: transcription factors, Pol subunits, regulators of transcription and/or chromatin remodelers, DNA-or RNA-binding proteins, DNA replication, kinases or phosphatases, and other functions ( Table 1). As anticipated, TBP was identified as one of the putative LmBrf1 partners. Nevertheless, Bdp1 (the third TFIIIB subunit) did not co-purify with LmBrf1-PTP. Notably, a putative ortholog of the TFIIIC subunit Tau131 was identified (LmjF.12.0560) ( Table 1). This is a tetratricopeptide repeat protein annotated as hypothetical in the L. major genome database; however, sequence analysis with the HHpred server identified it as a probable ortholog of yeast Tau131 (E-value of 1.1 × 10 −41 ). Pol III subunits RPC1, RPAC1 (shared with Pol I), and RPB5 (shared with Pol II) were also identified. Moreover, La protein, which binds to Pol III transcripts, co-purified with LmBrf1-PTP.  In addition to LmBrf1, mass spectrometry analysis of the LmBrf1-PTP sample allowed the identification of multiple proteins that were divided in seven categories: transcription factors, Pol subunits, regulators of transcription and/or chromatin remodelers, DNA-or RNA-binding proteins, DNA replication, kinases or phosphatases, and other functions ( Table 1). As anticipated, TBP was identified as one of the putative LmBrf1 partners. Nevertheless, Bdp1 (the third TFIIIB subunit) did not co-purify with LmBrf1-PTP. Notably, a putative ortholog of the TFIIIC subunit Tau131 was identified (LmjF.12.0560) ( Table 1). This is a tetratricopeptide repeat protein annotated as hypothetical in the L. major genome database; however, sequence analysis with the HHpred server identified it as a probable ortholog of yeast Tau131 (E-value of 1.1 × 10 −41 ). Pol III subunits RPC1, RPAC1 (shared with Pol I), and RPB5 (shared with Pol II) were also identified. Moreover, La protein, which binds to Pol III transcripts, co-purified with LmBrf1-PTP. Interestingly, several proteins related to Pol II transcription were also purified. These include Pol II subunits RPB1 and RPB2, transcription elongation factors TFIIS and SPT6, some subunits from the PAF1 and the CCR4-NOT complexes, and FACT complex subunit SPT16 (Table 1). Moreover, several proteins involved in Pol I transcription were also identified, including Pol I subunits RPA1 and RPB5z, and high mobility group protein TDP1.

Discussion
In this work, we characterized TFIIIB subunit Brf1 in the early-branching eukaryote L. major. Bioinformatic analyses demonstrated that the N-terminal half of LmBrf1 contains the distinctive zinc ribbon, cyclin repeats, and homology block I (Figure 1). Three-dimensional structure predictions show that the zinc ribbon folds into three antiparallel β strands, and each cyclin repeat folds into five α-helices with a globular organization. Homology block I is predicted to fold into a helix that is conserved in ScBrf1 ( Figure 1). Notably, relatively conserved homology blocks II and III were identified in LmBrf1 (Figure 1). These two conserved sequences are characteristic of yeast species, and they are involved in the interactions of Brf1 with TBP and Bdp1 [23,24,45]. Homology blocks II and III were not identified previously in Brf1 from T. brucei [35]; however, sequence alignments indicated that they are actually present in this parasite, as well as in all the other trypanosomatids that were here analyzed ( Figure S1, Supplementary Materials). These two homology blocks represent the most conserved sequences in the C-terminal half of Brf1 in trypanosomatids ( Figure S1, Supplementary Materials).
Immunofluorescence assays showed that LmBrf1 is a nuclear protein (Figure 2). A similar punctate nuclear signal was reported for other Pol III transcription factors in L. major and T. brucei, including Bdp1 [33] and Maf1 [41], as well as for the L. major 5S rRNA genes [29]. Even though the vast majority of LmBrf1 signal is nucleoplasmic, a small part of the immunostaining colocalized with Nop56 in the nucleolar periphery, which could support the association of LmBrf1 with the rRNA transcription unit that was found in ChIP experiments ( Figure 5) and the possible interaction with proteins involved in Pol I transcription (Table 1), as discussed below.
The fact that we were unable to obtain LmBrf1-null mutants suggests that this protein is essential for the viability of L. major promastigotes. In support of this possibility, inducible knockdown experiments in T. brucei procyclic forms demonstrated that Brf1 is indispensable for cell survival [35]. In other organisms, including yeast, Brf1 is an essen-tial protein [52]. Similarly, LmBdp1 is an essential molecule for the viability of L. major promastigotes [33]. Other strategies, such as CRISPR/Cas9 genome editing, should be performed to demonstrate the essentiality of LmBrf1 for L. major growth.
Cellular RNA levels are determined through the interplay among RNA synthesis, processing, and degradation [53]. As Pol II transcription is constitutive in trypanosomatids, mRNA abundance is mainly regulated through processing by trans-splicing and polyadenylation, and by degradation [54]. It has been shown in mammals and T. brucei that the mRNAs encoding proteins with similar functions share comparable decay rates [53,54]. In line with these findings, our Northern blot analyses showed that the mRNAs from Lm-Brf1, Bdp1, and RPC1 (all involved in Pol III transcription) have a half-life of~40 min, while the transcript from α-tubulin presents a half-life of~4.3 h (Figure 3). These decay rates are relatively short in relation to human mRNAs, which show half-lives of 7 to 10 h [55]. However, even shorter decay rates have been reported in T. brucei, where the median mRNA half-life is about 20 min [54].
Compared to logarithmic-phase cells, in stationary-phase parasites, we observed an increase in the half-lives of the mRNAs encoding LmBrf1 (from 39 to 60 min), Bdp1 (from 36 to 72 min), and RPC1 (from 40 to 62 min) ( Figure 3). The stabilization of mRNAs during the stationary phase was observed in T. cruzi, where it was suggested the presence of a regulation mechanism that protects mRNAs from degradation in stationary epimastigotes, as a strategy to perpetuate through this quiescent stage [56]. A similar mechanism might be present in nonproliferative promastigotes of Leishmania. Interestingly, unlike the other transcripts that we analyzed, the half-life of the α-tubulin mRNA was reduced in stationary phase cells from 4.30 to 2.25 h (Figure 3). Destabilization of α-tubulin transcripts in nonproliferative promastigotes has been reported in Leishmania [57,58]. Consequently, the stability of α-tubulin mRNA must be regulated by a different mechanism.
ChIP experiments demonstrated that LmBrf1 binds to tRNA, snRNA, and 5S rRNA genes ( Figure 5). Similarly, TBP and Bdp1, the other subunits of TFIIIB, have been shown to bind to these genes in L. major [32,33]. ChIP results also indicated the unexpected binding of LmBrf1 to the promoter region of the rRNA genes (transcribed by Pol I), to the SSR from chromosome 1, the LmjF.11.0930 gene, and to the SL RNA locus (transcribed by Pol II). The co-purification of proteins involved in Pol II and Pol I transcription with LmBrf1-PTP supports this finding (Table 1) (see below). Future genome-wide studies would reveal if LmBrf1 binds to other Pol II-dependent regions.
Tandem affinity purification experiments allowed the identification of several proteins that co-purified with LmBrf1-PTP, including TBP (Table 1). It was not surprising that we did not identify Bdp1, the third TFIIIB subunit, taking into consideration reports in other species that show that, while Brf1 interacts tightly with TBP and co-purifies, Bdp1 is weakly associated within the TFIIIB complex [59]. Nevertheless, it is also possible that the presence of the PTP-tag in LmBrf1 affected the interaction with Bdp1. Thus, other approaches should be employed to demonstrate the association between Bdp1 and Brf1 in Leishmania. Notably, a putative ortholog of TFIIIC subunit Tau131 (LmjF.12.0560) co-purified with LmBrf1-PTP. It is worth noting that our group identified the T. brucei ortholog (Tb927.1.3860) as an interacting partner of Pol III subunit RPC3 by tandem affinity purifications (manuscript in preparation). Thus, contrary to what was presumed [60], at least one subunit of TFIIIC is present in trypanosomatid parasites. Additionally, three Pol III subunits, the La protein, and a tRNA-binding domain-containing protein were co-purified with LmBrf1-PTP ( Table 1).
The multiple proteins identified by the mass spectrometry analyses (Table 1), including the proteins related to Pol II and Pol I transcription, represent potential interacting partners of LmBrf1-PTP. Future studies will help to determine if they actually associate, directly or indirectly, with LmBrf1. Among the Pol II-related proteins, we identified the transcription elongation factors TFIIS and SPT6, as well as two subunits of CCR4-NOT (Not1 and Not5), a complex that regulates gene expression at all steps, from transcription of mRNAs in the nucleus to their degradation in the cytoplasm [61,62]. We also identified four components of the PAF1 complex, which binds to Pol II to regulate transcriptional elongation and termination, as well as mRNA processing and export [63]. Interestingly, recent findings indicate that PAF1 also participates in Pol I and Pol III transcription [64]. Among the PAF1 subunits, a putative ortholog of RTF1 (HHpred E-value of 0.0011) was identified. Another protein that co-purified with LmBrf1-PTP is a RuvB-like DNA helicase, component of the INO80 chromatin remodeling complex that plays crucial roles in transcription by interacting with Pol II and TBP [65]. We also identified a putative Staphylococcal nuclease homolog/Tudor domain-containing protein (Table 1). These are evolutionarily conserved proteins involved in virtually all pathways of gene expression, from transcription to RNA silencing [66]. Regarding Pol I transcription, we identified Pol I subunits RPA1 and RPB5z, and the high-mobility group protein TDP1, which replaces histones on active Pol I transcription units in T. brucei to maintain an open chromatin structure and facilitate their transcription [67].
Several DNA-and RNA-binding proteins that could regulate gene expression were co-purified with LmBrf1-PTP, including a KD domain containing protein, putative ATPdependent RNA helicase SUB2, and ALBA-domain protein 3 (Table 1). Protein phosphatase 2C and several protein kinases, whose role in transcription regulation has been extensively studied in other organisms [68][69][70], were also identified. Some factors that participate in DNA replication were also found (Table 1), which might suggest that DNA replication is coupled to transcription in L. major, as has been reported in human cells [71]. A reduced number of peptides was identified for some of the putative interacting partners (Table 1), which may indicate a weak and/or transient interaction with LmBrf1.
At the present time, we do not know if the association of LmBrf1 with some genes transcribed by Pol I and Pol II ( Figure 5) and with proteins involved in Pol I and Pol II transcription (Table 1) has a functional significance. Several reports have demonstrated that, rather than functioning independently, the different types of RNA polymerases and their transcription factors can frequently work with one another to globally regulate and coordinate gene expression [72][73][74][75][76]. Therefore, it is possible that LmBrf1 helps to regulate transcription by all three RNA polymerases in L. major. Nevertheless, the interaction of LmBrf1 with Pol I-and Pol II-dependent regions and their transcription regulators might be fortuitous. A genome-wide study showed that TBP is highly enriched at most SSRs, the SL RNA genes, and the rRNA transcription units [32]. Thus, LmBrf1 could be attracted to these loci by TBP without playing a functional role. Future work will allow clarification of this subject.

Conclusions
In this work, we showed that, despite sequence divergence, LmBrf1 possesses the six conserved domains of Brf1 orthologs. Moreover, the predicted three-dimensional structure of the N-terminal region of LmBrf1 is conserved. Our results also suggest that LmBrf1 is an essential nuclear protein. Northern blot assays showed that the mRNAs encoding LmBrf1, Bdp1, and RPC1 (involved in Pol III transcription) share similar half-lives. All three mRNAs were stabilized in stationary-phase parasites. ChIP experiments demonstrated that LmBrf1 associates to 5S rRNA, tRNA, and snRNA genes. However, unlike other species, LmBrf1 associates to the Pol I-dependent 18S rRNA gene promoter and to three regions transcribed by Pol II. Moreover, our results suggest the interaction of LmBrf1 with transcriptional complexes from all three RNA polymerases. Subunit Tau131 of TFIIIC was identified as one of the putative interacting partners of LmBrf1. Data Availability Statement: All data generated or analyzed in this study are available from the corresponding author on reasonable request.