- Email: [email protected]
Revisiting intrinsic transcription termination in mycobacteria: U-tract downstream of secondary structure is dispensable for termination
Biochemical and Biophysical Research Communications xxx (xxxx) xxx
Contents lists available at ScienceDirect
Biochemical and Biophysical Research Communications journal homepage: www.elsevier.com/locate/ybbrc
Revisiting intrinsic transcription termination in mycobacteria: U-tract downstream of secondary structure is dispensable for termination Ezaz Ahmad a, Shubhada R. Hegde b, Valakunja Nagaraja a, c, * a
Department of Microbiology and Cell Biology, Indian Institute of Science, Bengaluru, 560012, India Institute of Bioinformatics and Applied Biotechnology, Bengaluru, 560100, India c Jawaharlal Nehru Centre for Advanced Scientific Research, Bengaluru, 560100, India b
a r t i c l e i n f o
a b s t r a c t
Article history: Received 6 November 2019 Accepted 9 November 2019 Available online xxx
In Escherichia coli, G/C-rich hairpin structure followed by a U-tract in the 30 region of the nascent RNA are crucial determinants for intrinsic or factor independent transcription termination. In mycobacteria, there is a scarcity of such intrinsic terminators. However, secondary structures having G/C-rich stem devoid of any U’s or with suboptimal U-tracts were identified earlier as terminators and found to be functional both in vitro and in vivo. Two different observations - that a mycobacterial RNA polymerase (RNAP) does not function at intrinsic terminators devoid of U-tracts and the identification of an altogether new motif for termination in mycobacteria necessitated re-examining a number of putative terminators for their function as terminators. When these in silico identified non-canonical terminators were subjected to experimental validation, they were found to dissociate RNA from the elongating RNAP. Termination is observed when the U-tracts were reduced, or totally absent both in vitro and in vivo. Our results, thus indicate that the presence of U-tract following the G/C-rich stem in an intrinsic terminator may not be an essential determinant for transcription termination in mycobacteria. © 2019 Elsevier Inc. All rights reserved.
Keywords: Transcription Intrinsic termination Mycobacteria Gene expression RNA polymerase
1. Introduction Transcription termination is a necessary step to end RNA synthesis. Of the two-transcription termination mechanisms in eubacteria, the intrinsic or factor independent termination has been studied in detail. The canonical intrinsic termination signal in DNA is composed of a G/C-rich palindromic element, immediately followed by a T-stretch. A weaker RNA/DNA hybrid generated after transcribing the T-stretch in the template causes transcriptional pause, and stable stem-loop formation in the nascent transcript to facilitate the release of RNA polymerase (RNAP) [1e3]. Intrinsic termination process thus, is considered economical for the cell as no additional protein factors or energy-involving processes are required to terminate transcription. However, the sequencing of a number of genomes indicated that many of them lack canonical
Abbreviations: RNAP, RNA Polymerase; TEC, Transcription elongation complex; EC, Elongation complex; GeSTer, Genome Scanner for Intrinsic Terminators; TRIT, Tuberculosis Rho-independent terminator; MsRNAP, Mycobacterium smegmatis RNA polymerase; SigA, Sigma A. * Corresponding author. Department of Microbiology and Cell Biology, Indian Institute of Science, Bengaluru, 560012, India. E-mail address: [email protected] (V. Nagaraja).
intrinsic terminators towards the 3’ end of the genes [4e7] hinting at the presence of alternative features as yardsticks for termination. To address, GeSTer algorithm (Genome Scanner for Intrinsic Terminators) was developed for in silico analysis, which could identify non-canonical terminators in addition to canonical terminators [7,8]. Unlike a canonical intrinsic terminator, a non-canonical variant has hairpin stem followed by a suboptimal U-tract (<3) or no U residues at all in the nascently transcribed RNA. A few of these predicted terminators were shown to be functional in mycobacteria both in vivo and in vitro [9]. Nearly 90% of in silico predicted intrinsic terminators in a number of mycobacterial species lack U-tracts entirely or possess a mixed A/U trail [10]. Using a probabilistic approach to predict intrinsic terminators, yet another algorithm termed RNIE has been developed [5]. In this in silico prediction model, Gardner et al. claimed that Tuberculosis Rho-independent terminator (TRIT) motif accounts for about 80% of significantly stable terminator structures in Mycobacterium tuberculosis. A cursory examination of some of these sequences indicated that they do not have features of either a typical intrinsic terminator or a non-canonical intrinsic terminator. Notably, none of these predicted terminators were experimentally validated either in vitro or in vivo. In yet another study, but with experimental
https://doi.org/10.1016/j.bbrc.2019.11.062 0006-291X/© 2019 Elsevier Inc. All rights reserved.
Please cite this article as: E. Ahmad et al., Revisiting intrinsic transcription termination in mycobacteria: U-tract downstream of secondary structure is dispensable for termination, Biochemical and Biophysical Research Communications, https://doi.org/10.1016/j.bbrc.2019.11.062
2
E. Ahmad et al. / Biochemical and Biophysical Research Communications xxx (xxxx) xxx
approach using Mycobacterium bovis RNAP, Czyz et al. showed that only canonical structures function as intrinsic terminators in mycobacteria. Their data indicated that non-canonical intrinsic terminators that lack U-tract in RNA do not function as terminators, but, transcription factor NusG promotes novel terminationstimulating activity on intrinsic terminators with suboptimal Utracts [11]. Given these diverse reports with seemingly disparate observations and predictions, it was imperative to reexamine the potential of predicted non-canonical intrinsic terminators to effect termination by carrying out transcription termination assays. Here, several non-canonical structures, especially those with 3 U’s immediately downstream of G/C-rich hairpin stem, predicted by the algorithm GeSTer, were tested experimentally for their ability to terminate transcription and release the transcript from transcription elongation complexes (TEC). 2. Materials and Methods 2.1. Bacterial strains, plasmids, construction of clones, chemicals and enzyme E. coli strains DH10B and BL21 lDE3 were used for cloning and protein overexpression, respectively. Selected putative intrinsic terminators were cloned in pUC18 plasmid containing T7A1 promoter. Oligonucleotides were synthesized by Sigma (Table S1). a32P UTP was purchased from Board of Radiation and Isotope Technology (BRIT), Mumbai. DNA modifying enzymes were purchased either from NEB or Roche. Streptavidin coated agarose beads and Ni2þ-nitrilotriacetic acid (Ni-NTA) beads were from GE healthcare and G-Biosciences respectively. Sigma A (SigA) enriched His-tagged RNAP was purified from Mycobacterium smegmatis SM07 strain as described before [12] and was used for in vitro transcription assays. The enzyme (MsRNAP) was purified by polyethyleneimine precipitation, heparin sepharose chromatography and Ni-NTA affinity chromatography. 2.2. In silico analysis Putative non-canonical intrinsic terminators from M. tuberculosis were identified using GeSTer [7,8]. RNIE algorithm was employed to derive the prediction of these in vitro validated intrinsic terminators in mycobacteria [5]. The Infernal package version infernal-1.0.2 was used to execute the RNIE algorithm. The genome sequences and the gene annotations were downloaded from the NCBI database (https://www.ncbi.nlm.nih.gov/) for both of the analyses. 2.3. In vitro transcription Sequences containing non-canonical intrinsic terminators were cloned in pUC18 between XbaI and HindIII sites. DNA templates for transcription reactions were prepared by PCR. Each DNA template contained pT7A1 promoter fused with a putative non-canonical intrinsic terminator with imperfect or no U-tracts. Putative terminators of mutT1, xerC, Rv3183, Lsr2, hpt, ung, kdpC, metE, mkl, fadD16, cyp144, bfrB, sdaA, rpsD, glnA1, and rrf genes were selected using GeSTer. In vitro transcription was carried out by incubating 10 nM DNA template with 200 nM RNAP in T-Buffer (Tris HCl pH8 50 mM, MgOAc 3 mM, K-Glutamate 100 mM, DTT 0.1 mM, EDTA 0.1 mM, BSA 0.1 mg/ml, Glycerol 5%), first on ice followed by 10 min at 37 C. Transcription was initiated with the addition of 200 mM rNTPs (200 mM mix of rATP, rCTP, rGTP, and 20 mM rUTP) and 10 mCi a-32PUTP, incubated at 37 C for 30 min and terminated by the addition of phenol-chloroform. Following centrifugation, an equal volume of
gel-loading dye (95% deionized Formamide, 0.05% Bromophenol Blue and 0.05% Xylene cyanol) was added to the aqueous phase, heated at 90 C for 2 min, and resolved on 8% 8 M urea-PAGE. The gel was scanned and quantified using Typhoon 9500(GE) Phosphorimager and Multi Gauge V2.3 software respectively. 2.4. Termination assays using immobilized RNAP or DNA template Assays were carried out using immobilized RNAP as described previously [11]. Reactions were set in duplicate. Briefly, 25 nM of DNA template was incubated with 200 nM RNAP in transcription buffer (Tris HCl pH8 20 mM, NaCl 20 mM, MgCl2 10 mM, b-mercaptoethanol 1 mM, BSA 0.1 mg/ml, and Glycerol 2%), first on ice followed by at 37 C for 10 min. The template contains only A and G from þ1 to þ19 positions downstream of the T7A1 promoter. Elongation complexes (ECs) with stalled RNAP were generated by the addition of 40 mM rATP and 40 mM rGTP and incubation at 37 C for 10 min. These complexes were incubated in transcription buffer preequilibrated Ni-NTA beads at room temperature for 30 min with gentle tumbling. Transcription was restarted by the addition of 200 mM rNTPs with supplement of tRNA (final concentration 50 mg/ ml) and 10 mM MgCl2 at room temperature for 30 min. First set of reactions were directly stopped by adding an equal volume of 2 STOP buffer (125 mM Tris HCl pH8, 15 mM EDTA, 333 mM NaCl, 1.25% SDS). The other set was separated into supernatant and beads and processed as described [11]. For in vitro transcription assays using immobilized DNA, templates were generated by PCR amplification using biotinylated pUC18 forward primer. 10 nM biotinylated templates were incubated with streptavidin coated agarose beads preequilibrated with T-buffer as described [11] at 37 C for 15 min. In vitro transcriptions were carried out at 37 C for 30 min. Reactions were separated into supernatant and beads and processed as described above. 2.5. RNA isolation and qRT-PCR Two primer pairs were designed for four of the in vitro tested genes, bfrB, lsr2, mkl, and Rv3183. These primer pairs were positioned to analyse the presence of transcripts upstream and downstream of the predicted terminators. RNA was isolated by using RNAZol-RT reagent (Sigma) from exponentially growing M. tuberculosis H37Ra cells, incubated at 37 C till A600 ¼ 0.5. RNA concentration and purity were checked spectrophotometrically followed by digestion with DNase I (Fermentas) to remove traces of chromosomal DNA. Samples were tested for DNA contaminations using PCR. cDNA was synthesized using Applied Bioscience cDNA synthesis kit following manufacturer’s instructions using gene specific reverse primer. Quantitative PCR (qPCR) was carried out using Bio Rad CFX96 Touch Real Time PCR Machine. RNA copy number of terminated and read-through transcripts were interpolated from a standard curve generated against known copy number of in vitro transcribed RNA for each gene. 3. Results 3.1. Transcription termination at non-canonical intrinsic terminators Our WebGeSTer analysis showed underrepresentation of intrinsic terminators in different species of mycobacteria [10]. Among the predicted structures, a high percentage were of noncanonical nature. For example, in M. tuberculosis genome, 89% were of non-canonical intrinsic terminators [10]. Typically, these comprise of G/C-rich hairpin followed by imperfect U-tracts or no ‘Us’ at all. These structures were localized downstream of the stop
Please cite this article as: E. Ahmad et al., Revisiting intrinsic transcription termination in mycobacteria: U-tract downstream of secondary structure is dispensable for termination, Biochemical and Biophysical Research Communications, https://doi.org/10.1016/j.bbrc.2019.11.062
E. Ahmad et al. / Biochemical and Biophysical Research Communications xxx (xxxx) xxx
codon anywhere between 9 nts to 200 nts. Sequences from the 30 UTR of mutT1, Rv3183, Lsr2, bfrB, cyp144, metE, mkl, kdpC, rrf, glnA1, xerC, hpt, ung, fadD16, sdaA, and rpsD were chosen for in vitro transcription analysis (Fig. 1A and S1). Linear DNA templates harboring these putative intrinsic terminators downstream of T7A1 promoters were generated (Fig. 1B) and transcription assays were carried out with Ni-NTA affinity purified MsRNAP (Fig. 1C and S1). The transcripts in the Figs. 1C and S1 corresponds to the GeSTer predicted termination products for mutT1, Rv3183, Lsr2, bfrB, cyp144, metE, mkl, kdpC, rrf, and glnA1, indicating termination event. 3.2. Release of RNA from MsRNAP at non-canonical terminator templates In the above assays, the truncated RNA products seen could be paused intermediates, other RNA truncations, in addition to actual termination products. Termination assays that rely on the release of RNA from the TEC can differentiate between true termination
3
stimulated by destabilization of the EC, and the stalled RNAP or transcriptional pause. The terminators used in the above assays were tested for their ability to dissociate ECs by detecting the RNA release. For this, in vitro transcriptions were carried out with NiNTA bead bound His-tag MsRNAP stalled complexes. In these assays RNA release was seen for mutT1, Rv3183, Lsr2 (Fig. S2), indicating that they are indeed true terminators. When plasmids harboring putative terminators fused with T7A1 promoter were also used for in vitro transcription assay, the same pattern of termination was observed irrespective of the topological state of the DNA (Fig. S2B). To further validate that the released transcripts are indeed terminated products, RNA release assays were carried out differently. Instead of immobilizing RNAP, the templates were immobilized on streptavidin beads as described in Materials and Methods. In vitro transcription with immobilized DNA templates resulted in a similar pattern to what was observed with transcription assays with immobilized RNAP (Fig. 2). In these reactions also, the terminated RNAs were recovered from the supernatant
Fig. 1. Transcription termination at non-canonical intrinsic terminators. (A) Putative terminator sequences of mutT1, Rv3183 and lsr2 obtained by GeSTer analysis are shown as RNA with a horizontal line and half-ovals to indicate RNA stem-loop structures. (B) Schematic depiction of linear DNA template used for in vitro transcription assay harboring the above sequences fused with the T7A1 promoter. (C) Transcription was carried out using 10 nM of linear DNA templates and MsRNAP (200 nM) at 37 C, and the products were resolved on 6% 8 M urea-PAGE. Terminated products and run-off products are marked with open circles (B) and closed circle (C) respectively with size marker to determine the length of the transcripts. Termination efficiencies were calculated (Termination efficiency: (Intrinsic termination)/(Intrinsic termination þ run-off) with standard deviations from 3 experimental replicates (shown below the gel lanes).
Please cite this article as: E. Ahmad et al., Revisiting intrinsic transcription termination in mycobacteria: U-tract downstream of secondary structure is dispensable for termination, Biochemical and Biophysical Research Communications, https://doi.org/10.1016/j.bbrc.2019.11.062
4
E. Ahmad et al. / Biochemical and Biophysical Research Communications xxx (xxxx) xxx
Fig. 2. Detection of released RNA products in transcription with immobilized DNA templates. (A) Schematic representation of in vitro transcription strategy using streptavidin bead bound linear DNA template, designed to detect transcript release. (B) In vitro transcription assay was carried out with biotinylated linear DNA templates containing non-canonical intrinsic terminators immobilized on streptavidin beads using experimental protocol as described in Materials and Methods. Transcription reaction mixture was separated into supernatant (s) and washed-beads (wb). The appearance of RNAs in the supernatant fraction indicates transcript release which contains run-off (C) and (Term. Prod.) termination product (B).
with the run-off product, indicating that the mycobacterial noncanonical terminators function as true intrinsic terminators. RNA release was not observed with templates kdpC, glnA1, and rrf (Fig. S3) indicating that they may not function as terminators under these experimental conditions.
G/C-rich stem is necessary to bring about termination. From these results it is apparent that U residues downstream of hairpin structure are not absolutely necessary and they do not appear to influence significantly termination of the intrinsic terminators in mycobacteria.
3.3. U-tract is not necessary for intrinsic termination
3.4. Transcription termination in vivo with U-tract deficient terminators
From the results described in the Figs. 1 and 2 it is apparent that transcription termination can take place with the suboptimal Utract, raising the possibility that a few U residues may be necessary and sufficient for termination in mycobacteria. Thus, to examine whether the U residues downstream of a G/C-rich stem are integral constituents of intrinsic termination event, experiments were carried out using constructs where U’s were replaced with C’s. In parallel, as control, G/C-rich stem mutants were also used (Fig. 3). In vitro transcription with Rv3183 and Lsr2 terminator mutants with no U-tract showed the termination product of same size as wild type templates (Fig. 3). MsRNAP could terminate with comparable efficiency with both wild type and mutant templates. However, termination assays with intrinsic terminators having destabilized stem-loop structure did not show the release of termination product (Fig. 3). Thus even for non-canonical intrinsic terminators,
To test whether the non-canonical intrinsic terminators function in vivo, a representative set of four in vitro validated terminators were chosen. Two of them - Rv3183 and Lsr2 have suboptimal U-tracts (3 U’s) and the other two e mkl and bfrB do not possess U’s at all following the hairpin. To assess termination in vivo, RNA was quantified by qRT-PCR of sequence from upstream and downstream of the hairpin structure as described in Materials and Methods and Fig. 4A. Standard curve was generated using known concentration of in vitro synthesized RNA harboring the test intrinsic terminator. Using this standard curve, absolute copy number of test RNA was calculated. In all the four genes used for the in vivo study, efficient termination was observed. Quantitation revealed percentage termination to be around 80% and read-through is about 10e20% (Fig. 4B). Thus from both in vitro and in vivo experiments, it is
Please cite this article as: E. Ahmad et al., Revisiting intrinsic transcription termination in mycobacteria: U-tract downstream of secondary structure is dispensable for termination, Biochemical and Biophysical Research Communications, https://doi.org/10.1016/j.bbrc.2019.11.062
E. Ahmad et al. / Biochemical and Biophysical Research Communications xxx (xxxx) xxx
5
Fig. 3. U-tracts of intrinsic terminators are dispensable for termination. (A and B) RNA structures and sequences for the derivatives of predicted intrinsic terminator of Rv3183 and Lsr2 is depicted either having destabilized hairpin structures or reduced U-tract with appropriate DG. Transcription assays carried out with Rv3183-1 and Lsr2-1 having destabilized hairpin stem resulted in the absence of termination products. Experiments with Rv3183-2 and Lsr2-2 where the U-tract was reduced did not show any significant difference in termination.
apparent that non-canonical terminators function as terminators in M. tuberculosis.
3.5. Tuberculosis Rho-independent terminator (TRIT) motif is unlikely to function as a terminator Extracting genome sequences from 20 to þ 80 nt around gene termini from M. tuberculosis CDC1551, and using RNAfold routine from Vienna package [13], a new transcription termination motif in mycobacteria termed TRIT has been described [5]. TRITs account for ~81% of the significantly stable terminator sequences in mycobacteria [5]. Examination of a few of these sequences in the list revealed absence of similarity with text book defined intrinsic terminators or the non-canonical variants described above. Moreover, these predicted motifs have not been experimentally verified by carrying out standard termination assays either in vitro or in vivo. Hence, we examined the sequences having TRIT motif and compared them with canonical as well as non-canonical intrinsic terminators predicted by GeSTer and experimentally verified. Expectedly, we did not find high degree of similarities between the intrinsic terminators verses TRIT motif. Among all the predicted terminators, only tuf, Rv1324 and rpoC (3 out of 149) were validated as terminators (Table S2). Since most of the predicted or experimentally validated terminators do not have TRIT motif, we suggest that the main function of the motif is in some other cellular process and not transcription termination (See Discussion).
4. Discussion The textbook definition of intrinsic transcription termination is primarily based on E.coli paradigm. A G/C-rich stem loop structure, immediately followed by a stretch of 7e9 ‘U’ constituted the canonical intrinsic terminator [3,14]. For a long time, this notion of the termination structure prevailed as the subsequent analysis supported the model [15,16]. However, with the sequencing of a number of genomes, it was apparent that such canonical structures were under-represented in many organisms [6,7]. Most of the earlier algorithms that identified the terminators relied on the presence of these canonical motifs i.e. stem loop followed by Utract in RNA and hence failed to detect alternate terminator sequences in the genomic sequences leading to the suggestion for the operation of an altogether different termination mechanism [17]. However, GeSTer algorithm led to the identification of noncanonical intrinsic terminators that lacked or under-represented ‘U’ residues [7,9]. Now we have revisited the analysis of intrinsic terminators in mycobacteria to address the apparent differences arisen after Czyz et al. showed that the non-canonical terminators do not function as terminators [11]. They suggested following reasons for the discrepancy. 1) The RNAP used by Unniraman et al., 2001 was likely to have contained additional protein factors. 2) The transcript release assays would be necessary to assess actual terminators. Hence, they proposed that the observed termination products could be because of stalled complexes or nuclease contamination in the RNAP preparations. Now, using a pure
Please cite this article as: E. Ahmad et al., Revisiting intrinsic transcription termination in mycobacteria: U-tract downstream of secondary structure is dispensable for termination, Biochemical and Biophysical Research Communications, https://doi.org/10.1016/j.bbrc.2019.11.062
6
E. Ahmad et al. / Biochemical and Biophysical Research Communications xxx (xxxx) xxx
Fig. 4. In vivo assessment of termination efficiency for intrinsic terminators lacking U-tracts (A) Schematic depiction of primers designed to absolute quantify the RNA copy number for both terminated and read-through transcript by RT-qPCR. (B) Using the known concentration of the RNA standard curve was generated to estimate and compare the transcripts that represent the terminated and read-through products for bfrB, lsr2, mkl, and Rv3183. Termination efficiency was calculated using RNA copy number [Termination efficiency ¼ Termination (X-Y)/(Termination (X-Y) þ Read-through (Y))]. The data represented is mean ± SD from the three independent experiments.
preparation of the enzyme and two different bead assays for RNA release, we show that non-canonical terminator structures indeed function as true terminators. The GeSTer algorithm predicted terminators could terminate transcription irrespective of whether templates were immobilized or enzyme is immobilized. Moreover, the non-canonical terminators functioned as terminators both in vitro and in vivo. Notably, at least seven of the tested terminators were found to dissociate transcription elongation complexes in immobilized assays. However, not all predicted non-canonical terminator structures could terminate transcription. For example, as Czyz et al. showed [11], our termination assay with gyr A do not appear to release RNA in vitro (data not shown). It may be noted that for the current analysis RNAP used has been purified through different steps including affinity chromatography [12]. Hence the possibility of either having additional factors or nuclease contamination has been considered in the present experimental design. The differences between two reports could also be due to source of the RNAP (M. bovis verses M. smegmatis). Given the importance of transcription termination, new algorithms have been proposed to predict their occurrence. One such algorithm RNIE describes a new motif TRIT for transcription termination in mycobacteria [5]. Although nearly 150 terminator sequences have been identified by this approach, these predictions
were not subjected to experimental validation. Our present experimental analysis indicates that TRIT motif is unlikely to act as a true terminator. Very few of the GeSTer predicted, experimentally validated terminators have TRIT motif. Further even with the terminators studied by Czyz et al., only a very small subset of validated sequences have TRIT motif [11]. Although these motifs have the propensity for strong secondary structures, they lack the core features required for an intrinsic terminator in eubacteria. From all the studies so far, it is apparent that not all secondary structures would function as terminators. We have previously shown that stem loop structures with very low DG or very high DG are unlikely to function as terminators [10]. Thus, in order to function as terminators, species specific optimized DG of the structures appears to be necessary [10]. Although these predictive algorithms are important, the experimental validation of predicted structure is necessary. As the minimal requirement for an intrinsic terminator is a G/C-rich stem, while the U-tract facilitates the release of RNA, we note that very few of the TRIT motif appear to have G/C-rich stem which is an essential feature for termination hairpin loop within the termination bubble. Hence, it is not surprising that the structures with TRIT motif did not function as terminators in our analysis. We suggest that TRIT motif may function as some other important cellular signal instead of acting as terminator. RNA secondary
Please cite this article as: E. Ahmad et al., Revisiting intrinsic transcription termination in mycobacteria: U-tract downstream of secondary structure is dispensable for termination, Biochemical and Biophysical Research Communications, https://doi.org/10.1016/j.bbrc.2019.11.062
E. Ahmad et al. / Biochemical and Biophysical Research Communications xxx (xxxx) xxx
structures are known to function in generating stable 3’ ends of RNA after a termination step occurring downstream [18,19]. In concluding from our re-investigation of intrinsic termination in mycobacteria, it appears that some of the predicted noncanonical intrinsic terminators function as terminators both in vitro and in vivo. It is also evident that not all the GeSTeR predicted terminators function as terminators by themselves as suggested earlier [11]. Thus, intrinsic termination in eubacteria has retained the core characteristics across bacterial kingdom with species-specific optimized stem loop structures. Although U-tract present are important, they are dispensed away in organisms such as mycobacteria, which are G/C-rich, and have a paucity of canonical terminators. Experimental validation of the predicted structures would be important to ascertain their function as true terminators. Declaration of competing interest The authors declare no conflict of interest with the contents of this article. Acknowledgements The central facility of phosphor imaging of Indian Institute of Science, supported by Department of Biotechnology, Government of India is acknowledged. Financial support for the project is from a grant to V.N. from the Department of Biotechnology (DBT), Government of India, DBT sponsored Life Science Research, Education and Training at JNCASR, and DBT-IISc partnership program. V.N. is a J.C. Bose Fellow of the Department of Science and Technology, Government of India. Appendix A. Supplementary data Supplementary data to this article can be found online at https://doi.org/10.1016/j.bbrc.2019.11.062. References [1] I. Gusarov, E. Nudler, The mechanism of intrinsic transcription termination, Mol. Cell 3 (1999) 495e504, https://doi.org/10.1016/s1097-2765(00)80477-3. [2] R. Reynolds, R.M. Bermudez-Cruz, M.J. Chamberlin, Parameters affecting transcription termination by Escherichia coli RNA polymerase. I. Analysis of 13 rho-independent terminators, J. Mol. Biol. 224 (1992) 31e51, https://doi.org/
7
10.1016/0022-2836(92)90574-4. [3] J.W. Roberts, Mechanisms of bacterial transcription termination, J. Mol. Biol. (2019), https://doi.org/10.1016/j.jmb.2019.04.003. [4] M.D. Ermolaeva, H.G. Khalak, O. White, H.O. Smith, S.L. Salzberg, Prediction of transcription terminators in bacterial genomes, J. Mol. Biol. 301 (2000) 27e33, https://doi.org/10.1006/jmbi.2000.3836. [5] P.P. Gardner, L. Barquist, A. Bateman, E.P. Nawrocki, Z. Weinberg, RNIE: genome-wide prediction of bacterial intrinsic terminators, Nucleic Acids Res. 39 (2011) 5845e5852, https://doi.org/10.1093/nar/gkr168. [6] A. Mitra, K. Angamuthu, H.V. Jayashree, V. Nagaraja, Occurrence, divergence and evolution of intrinsic terminators across eubacteria, Genomics 94 (2009) 110e116, https://doi.org/10.1016/j.ygeno.2009.04.004. [7] S. Unniraman, R. Prakash, V. Nagaraja, Conserved economics of transcription termination in eubacteria, Nucleic Acids Res. 30 (2002) 675e684, https:// doi.org/10.1093/nar/30.3.675. [8] A. Mitra, A.K. Kesarwani, D. Pal, V. Nagaraja, WebGeSTer DBea transcription terminator database, Nucleic Acids Res. 39 (2011) D129eD135, https:// doi.org/10.1093/nar/gkq971. [9] S. Unniraman, R. Prakash, V. Nagaraja, Alternate paradigm for intrinsic transcription termination in eubacteria, J. Biol. Chem. 276 (2001) 41850e41855, https://doi.org/10.1074/jbc.M106252200. [10] A. Mitra, K. Angamuthu, V. Nagaraja, Genome-wide analysis of the intrinsic terminators of transcription across the genus Mycobacterium, Tuberculosis 88 (2008) 566e575, https://doi.org/10.1016/j.tube.2008.06.004. [11] A. Czyz, R.A. Mooney, A. Iaconi, R. Landick, Mycobacterial RNA polymerase requires a U-tract at intrinsic terminators and is aided by NusG at suboptimal terminators, mBio 5 (2014) e00931, https://doi.org/10.1128/mBio.00931-14. [12] A. China, V. Nagaraja, Purification of RNA polymerase from mycobacteria for optimized promoter-polymerase interactions, Protein Expr. Purif. 69 (2010) 235e242, https://doi.org/10.1016/j.pep.2009.09.022. [13] A.R. Gruber, R. Lorenz, S.H. Bernhart, R. Neubock, I.L. Hofacker, The Vienna RNA websuite, Nucleic Acids Res. 36 (2008) W70eW74, https://doi.org/ 10.1093/nar/gkn188. [14] Y. d’Aubenton Carafa, E. Brody, C. Thermes, Prediction of rho-independent Escherichia coli transcription terminators. A statistical analysis of their RNA stem-loop structures, J. Mol. Biol. 216 (1990) 835e858, https://doi.org/ 10.1016/s0022-2836(99)80005-9. [15] W.S. Yarnell, J.W. Roberts, Mechanism of intrinsic transcription termination and antitermination, Science 284 (1999) 611e615, https://doi.org/10.1126/ science.284.5414.611. [16] M.J.L. de Hoon, Y. Makita, K. Nakai, S. Miyano, Prediction of transcriptional terminators in Bacillus subtilis and related species, PLoS Comput. Biol. 1 (2005) e25, https://doi.org/10.1371/journal.pcbi.0010025. [17] T. Washio, J. Sasayama, M. Tomita, Analysis of complete genomes suggests that many prokaryotes do not rely on hairpin formation in transcription termination, Nucleic Acids Res. 26 (1998) 5456e5463, https://doi.org/ 10.1093/nar/26.23.5456. [18] D. Dar, R. Sorek, High-resolution RNA 3’-ends mapping of bacterial Rhodependent transcripts, Nucleic Acids Res. 46 (2018) 6797e6805, https:// doi.org/10.1093/nar/gky274. [19] X. Wang, M.P.A. N, H.J. Jeon, Y. Lee, J. He, S. Adhya, H.M. Lim, Processing generates 3’ ends of RNA masking transcription termination events in prokaryotes, Proc. Natl. Acad. Sci. U. S. A 116 (2019) 4440e4445, https://doi.org/ 10.1073/pnas.1813181116.
Please cite this article as: E. Ahmad et al., Revisiting intrinsic transcription termination in mycobacteria: U-tract downstream of secondary structure is dispensable for termination, Biochemical and Biophysical Research Communications, https://doi.org/10.1016/j.bbrc.2019.11.062
Contents lists available at ScienceDirect
Biochemical and Biophysical Research Communications journal homepage: www.elsevier.com/locate/ybbrc
Revisiting intrinsic transcription termination in mycobacteria: U-tract downstream of secondary structure is dispensable for termination Ezaz Ahmad a, Shubhada R. Hegde b, Valakunja Nagaraja a, c, * a
Department of Microbiology and Cell Biology, Indian Institute of Science, Bengaluru, 560012, India Institute of Bioinformatics and Applied Biotechnology, Bengaluru, 560100, India c Jawaharlal Nehru Centre for Advanced Scientific Research, Bengaluru, 560100, India b
a r t i c l e i n f o
a b s t r a c t
Article history: Received 6 November 2019 Accepted 9 November 2019 Available online xxx
In Escherichia coli, G/C-rich hairpin structure followed by a U-tract in the 30 region of the nascent RNA are crucial determinants for intrinsic or factor independent transcription termination. In mycobacteria, there is a scarcity of such intrinsic terminators. However, secondary structures having G/C-rich stem devoid of any U’s or with suboptimal U-tracts were identified earlier as terminators and found to be functional both in vitro and in vivo. Two different observations - that a mycobacterial RNA polymerase (RNAP) does not function at intrinsic terminators devoid of U-tracts and the identification of an altogether new motif for termination in mycobacteria necessitated re-examining a number of putative terminators for their function as terminators. When these in silico identified non-canonical terminators were subjected to experimental validation, they were found to dissociate RNA from the elongating RNAP. Termination is observed when the U-tracts were reduced, or totally absent both in vitro and in vivo. Our results, thus indicate that the presence of U-tract following the G/C-rich stem in an intrinsic terminator may not be an essential determinant for transcription termination in mycobacteria. © 2019 Elsevier Inc. All rights reserved.
Keywords: Transcription Intrinsic termination Mycobacteria Gene expression RNA polymerase
1. Introduction Transcription termination is a necessary step to end RNA synthesis. Of the two-transcription termination mechanisms in eubacteria, the intrinsic or factor independent termination has been studied in detail. The canonical intrinsic termination signal in DNA is composed of a G/C-rich palindromic element, immediately followed by a T-stretch. A weaker RNA/DNA hybrid generated after transcribing the T-stretch in the template causes transcriptional pause, and stable stem-loop formation in the nascent transcript to facilitate the release of RNA polymerase (RNAP) [1e3]. Intrinsic termination process thus, is considered economical for the cell as no additional protein factors or energy-involving processes are required to terminate transcription. However, the sequencing of a number of genomes indicated that many of them lack canonical
Abbreviations: RNAP, RNA Polymerase; TEC, Transcription elongation complex; EC, Elongation complex; GeSTer, Genome Scanner for Intrinsic Terminators; TRIT, Tuberculosis Rho-independent terminator; MsRNAP, Mycobacterium smegmatis RNA polymerase; SigA, Sigma A. * Corresponding author. Department of Microbiology and Cell Biology, Indian Institute of Science, Bengaluru, 560012, India. E-mail address: [email protected] (V. Nagaraja).
intrinsic terminators towards the 3’ end of the genes [4e7] hinting at the presence of alternative features as yardsticks for termination. To address, GeSTer algorithm (Genome Scanner for Intrinsic Terminators) was developed for in silico analysis, which could identify non-canonical terminators in addition to canonical terminators [7,8]. Unlike a canonical intrinsic terminator, a non-canonical variant has hairpin stem followed by a suboptimal U-tract (<3) or no U residues at all in the nascently transcribed RNA. A few of these predicted terminators were shown to be functional in mycobacteria both in vivo and in vitro [9]. Nearly 90% of in silico predicted intrinsic terminators in a number of mycobacterial species lack U-tracts entirely or possess a mixed A/U trail [10]. Using a probabilistic approach to predict intrinsic terminators, yet another algorithm termed RNIE has been developed [5]. In this in silico prediction model, Gardner et al. claimed that Tuberculosis Rho-independent terminator (TRIT) motif accounts for about 80% of significantly stable terminator structures in Mycobacterium tuberculosis. A cursory examination of some of these sequences indicated that they do not have features of either a typical intrinsic terminator or a non-canonical intrinsic terminator. Notably, none of these predicted terminators were experimentally validated either in vitro or in vivo. In yet another study, but with experimental
https://doi.org/10.1016/j.bbrc.2019.11.062 0006-291X/© 2019 Elsevier Inc. All rights reserved.
Please cite this article as: E. Ahmad et al., Revisiting intrinsic transcription termination in mycobacteria: U-tract downstream of secondary structure is dispensable for termination, Biochemical and Biophysical Research Communications, https://doi.org/10.1016/j.bbrc.2019.11.062
2
E. Ahmad et al. / Biochemical and Biophysical Research Communications xxx (xxxx) xxx
approach using Mycobacterium bovis RNAP, Czyz et al. showed that only canonical structures function as intrinsic terminators in mycobacteria. Their data indicated that non-canonical intrinsic terminators that lack U-tract in RNA do not function as terminators, but, transcription factor NusG promotes novel terminationstimulating activity on intrinsic terminators with suboptimal Utracts [11]. Given these diverse reports with seemingly disparate observations and predictions, it was imperative to reexamine the potential of predicted non-canonical intrinsic terminators to effect termination by carrying out transcription termination assays. Here, several non-canonical structures, especially those with 3 U’s immediately downstream of G/C-rich hairpin stem, predicted by the algorithm GeSTer, were tested experimentally for their ability to terminate transcription and release the transcript from transcription elongation complexes (TEC). 2. Materials and Methods 2.1. Bacterial strains, plasmids, construction of clones, chemicals and enzyme E. coli strains DH10B and BL21 lDE3 were used for cloning and protein overexpression, respectively. Selected putative intrinsic terminators were cloned in pUC18 plasmid containing T7A1 promoter. Oligonucleotides were synthesized by Sigma (Table S1). a32P UTP was purchased from Board of Radiation and Isotope Technology (BRIT), Mumbai. DNA modifying enzymes were purchased either from NEB or Roche. Streptavidin coated agarose beads and Ni2þ-nitrilotriacetic acid (Ni-NTA) beads were from GE healthcare and G-Biosciences respectively. Sigma A (SigA) enriched His-tagged RNAP was purified from Mycobacterium smegmatis SM07 strain as described before [12] and was used for in vitro transcription assays. The enzyme (MsRNAP) was purified by polyethyleneimine precipitation, heparin sepharose chromatography and Ni-NTA affinity chromatography. 2.2. In silico analysis Putative non-canonical intrinsic terminators from M. tuberculosis were identified using GeSTer [7,8]. RNIE algorithm was employed to derive the prediction of these in vitro validated intrinsic terminators in mycobacteria [5]. The Infernal package version infernal-1.0.2 was used to execute the RNIE algorithm. The genome sequences and the gene annotations were downloaded from the NCBI database (https://www.ncbi.nlm.nih.gov/) for both of the analyses. 2.3. In vitro transcription Sequences containing non-canonical intrinsic terminators were cloned in pUC18 between XbaI and HindIII sites. DNA templates for transcription reactions were prepared by PCR. Each DNA template contained pT7A1 promoter fused with a putative non-canonical intrinsic terminator with imperfect or no U-tracts. Putative terminators of mutT1, xerC, Rv3183, Lsr2, hpt, ung, kdpC, metE, mkl, fadD16, cyp144, bfrB, sdaA, rpsD, glnA1, and rrf genes were selected using GeSTer. In vitro transcription was carried out by incubating 10 nM DNA template with 200 nM RNAP in T-Buffer (Tris HCl pH8 50 mM, MgOAc 3 mM, K-Glutamate 100 mM, DTT 0.1 mM, EDTA 0.1 mM, BSA 0.1 mg/ml, Glycerol 5%), first on ice followed by 10 min at 37 C. Transcription was initiated with the addition of 200 mM rNTPs (200 mM mix of rATP, rCTP, rGTP, and 20 mM rUTP) and 10 mCi a-32PUTP, incubated at 37 C for 30 min and terminated by the addition of phenol-chloroform. Following centrifugation, an equal volume of
gel-loading dye (95% deionized Formamide, 0.05% Bromophenol Blue and 0.05% Xylene cyanol) was added to the aqueous phase, heated at 90 C for 2 min, and resolved on 8% 8 M urea-PAGE. The gel was scanned and quantified using Typhoon 9500(GE) Phosphorimager and Multi Gauge V2.3 software respectively. 2.4. Termination assays using immobilized RNAP or DNA template Assays were carried out using immobilized RNAP as described previously [11]. Reactions were set in duplicate. Briefly, 25 nM of DNA template was incubated with 200 nM RNAP in transcription buffer (Tris HCl pH8 20 mM, NaCl 20 mM, MgCl2 10 mM, b-mercaptoethanol 1 mM, BSA 0.1 mg/ml, and Glycerol 2%), first on ice followed by at 37 C for 10 min. The template contains only A and G from þ1 to þ19 positions downstream of the T7A1 promoter. Elongation complexes (ECs) with stalled RNAP were generated by the addition of 40 mM rATP and 40 mM rGTP and incubation at 37 C for 10 min. These complexes were incubated in transcription buffer preequilibrated Ni-NTA beads at room temperature for 30 min with gentle tumbling. Transcription was restarted by the addition of 200 mM rNTPs with supplement of tRNA (final concentration 50 mg/ ml) and 10 mM MgCl2 at room temperature for 30 min. First set of reactions were directly stopped by adding an equal volume of 2 STOP buffer (125 mM Tris HCl pH8, 15 mM EDTA, 333 mM NaCl, 1.25% SDS). The other set was separated into supernatant and beads and processed as described [11]. For in vitro transcription assays using immobilized DNA, templates were generated by PCR amplification using biotinylated pUC18 forward primer. 10 nM biotinylated templates were incubated with streptavidin coated agarose beads preequilibrated with T-buffer as described [11] at 37 C for 15 min. In vitro transcriptions were carried out at 37 C for 30 min. Reactions were separated into supernatant and beads and processed as described above. 2.5. RNA isolation and qRT-PCR Two primer pairs were designed for four of the in vitro tested genes, bfrB, lsr2, mkl, and Rv3183. These primer pairs were positioned to analyse the presence of transcripts upstream and downstream of the predicted terminators. RNA was isolated by using RNAZol-RT reagent (Sigma) from exponentially growing M. tuberculosis H37Ra cells, incubated at 37 C till A600 ¼ 0.5. RNA concentration and purity were checked spectrophotometrically followed by digestion with DNase I (Fermentas) to remove traces of chromosomal DNA. Samples were tested for DNA contaminations using PCR. cDNA was synthesized using Applied Bioscience cDNA synthesis kit following manufacturer’s instructions using gene specific reverse primer. Quantitative PCR (qPCR) was carried out using Bio Rad CFX96 Touch Real Time PCR Machine. RNA copy number of terminated and read-through transcripts were interpolated from a standard curve generated against known copy number of in vitro transcribed RNA for each gene. 3. Results 3.1. Transcription termination at non-canonical intrinsic terminators Our WebGeSTer analysis showed underrepresentation of intrinsic terminators in different species of mycobacteria [10]. Among the predicted structures, a high percentage were of noncanonical nature. For example, in M. tuberculosis genome, 89% were of non-canonical intrinsic terminators [10]. Typically, these comprise of G/C-rich hairpin followed by imperfect U-tracts or no ‘Us’ at all. These structures were localized downstream of the stop
Please cite this article as: E. Ahmad et al., Revisiting intrinsic transcription termination in mycobacteria: U-tract downstream of secondary structure is dispensable for termination, Biochemical and Biophysical Research Communications, https://doi.org/10.1016/j.bbrc.2019.11.062
E. Ahmad et al. / Biochemical and Biophysical Research Communications xxx (xxxx) xxx
codon anywhere between 9 nts to 200 nts. Sequences from the 30 UTR of mutT1, Rv3183, Lsr2, bfrB, cyp144, metE, mkl, kdpC, rrf, glnA1, xerC, hpt, ung, fadD16, sdaA, and rpsD were chosen for in vitro transcription analysis (Fig. 1A and S1). Linear DNA templates harboring these putative intrinsic terminators downstream of T7A1 promoters were generated (Fig. 1B) and transcription assays were carried out with Ni-NTA affinity purified MsRNAP (Fig. 1C and S1). The transcripts in the Figs. 1C and S1 corresponds to the GeSTer predicted termination products for mutT1, Rv3183, Lsr2, bfrB, cyp144, metE, mkl, kdpC, rrf, and glnA1, indicating termination event. 3.2. Release of RNA from MsRNAP at non-canonical terminator templates In the above assays, the truncated RNA products seen could be paused intermediates, other RNA truncations, in addition to actual termination products. Termination assays that rely on the release of RNA from the TEC can differentiate between true termination
3
stimulated by destabilization of the EC, and the stalled RNAP or transcriptional pause. The terminators used in the above assays were tested for their ability to dissociate ECs by detecting the RNA release. For this, in vitro transcriptions were carried out with NiNTA bead bound His-tag MsRNAP stalled complexes. In these assays RNA release was seen for mutT1, Rv3183, Lsr2 (Fig. S2), indicating that they are indeed true terminators. When plasmids harboring putative terminators fused with T7A1 promoter were also used for in vitro transcription assay, the same pattern of termination was observed irrespective of the topological state of the DNA (Fig. S2B). To further validate that the released transcripts are indeed terminated products, RNA release assays were carried out differently. Instead of immobilizing RNAP, the templates were immobilized on streptavidin beads as described in Materials and Methods. In vitro transcription with immobilized DNA templates resulted in a similar pattern to what was observed with transcription assays with immobilized RNAP (Fig. 2). In these reactions also, the terminated RNAs were recovered from the supernatant
Fig. 1. Transcription termination at non-canonical intrinsic terminators. (A) Putative terminator sequences of mutT1, Rv3183 and lsr2 obtained by GeSTer analysis are shown as RNA with a horizontal line and half-ovals to indicate RNA stem-loop structures. (B) Schematic depiction of linear DNA template used for in vitro transcription assay harboring the above sequences fused with the T7A1 promoter. (C) Transcription was carried out using 10 nM of linear DNA templates and MsRNAP (200 nM) at 37 C, and the products were resolved on 6% 8 M urea-PAGE. Terminated products and run-off products are marked with open circles (B) and closed circle (C) respectively with size marker to determine the length of the transcripts. Termination efficiencies were calculated (Termination efficiency: (Intrinsic termination)/(Intrinsic termination þ run-off) with standard deviations from 3 experimental replicates (shown below the gel lanes).
Please cite this article as: E. Ahmad et al., Revisiting intrinsic transcription termination in mycobacteria: U-tract downstream of secondary structure is dispensable for termination, Biochemical and Biophysical Research Communications, https://doi.org/10.1016/j.bbrc.2019.11.062
4
E. Ahmad et al. / Biochemical and Biophysical Research Communications xxx (xxxx) xxx
Fig. 2. Detection of released RNA products in transcription with immobilized DNA templates. (A) Schematic representation of in vitro transcription strategy using streptavidin bead bound linear DNA template, designed to detect transcript release. (B) In vitro transcription assay was carried out with biotinylated linear DNA templates containing non-canonical intrinsic terminators immobilized on streptavidin beads using experimental protocol as described in Materials and Methods. Transcription reaction mixture was separated into supernatant (s) and washed-beads (wb). The appearance of RNAs in the supernatant fraction indicates transcript release which contains run-off (C) and (Term. Prod.) termination product (B).
with the run-off product, indicating that the mycobacterial noncanonical terminators function as true intrinsic terminators. RNA release was not observed with templates kdpC, glnA1, and rrf (Fig. S3) indicating that they may not function as terminators under these experimental conditions.
G/C-rich stem is necessary to bring about termination. From these results it is apparent that U residues downstream of hairpin structure are not absolutely necessary and they do not appear to influence significantly termination of the intrinsic terminators in mycobacteria.
3.3. U-tract is not necessary for intrinsic termination
3.4. Transcription termination in vivo with U-tract deficient terminators
From the results described in the Figs. 1 and 2 it is apparent that transcription termination can take place with the suboptimal Utract, raising the possibility that a few U residues may be necessary and sufficient for termination in mycobacteria. Thus, to examine whether the U residues downstream of a G/C-rich stem are integral constituents of intrinsic termination event, experiments were carried out using constructs where U’s were replaced with C’s. In parallel, as control, G/C-rich stem mutants were also used (Fig. 3). In vitro transcription with Rv3183 and Lsr2 terminator mutants with no U-tract showed the termination product of same size as wild type templates (Fig. 3). MsRNAP could terminate with comparable efficiency with both wild type and mutant templates. However, termination assays with intrinsic terminators having destabilized stem-loop structure did not show the release of termination product (Fig. 3). Thus even for non-canonical intrinsic terminators,
To test whether the non-canonical intrinsic terminators function in vivo, a representative set of four in vitro validated terminators were chosen. Two of them - Rv3183 and Lsr2 have suboptimal U-tracts (3 U’s) and the other two e mkl and bfrB do not possess U’s at all following the hairpin. To assess termination in vivo, RNA was quantified by qRT-PCR of sequence from upstream and downstream of the hairpin structure as described in Materials and Methods and Fig. 4A. Standard curve was generated using known concentration of in vitro synthesized RNA harboring the test intrinsic terminator. Using this standard curve, absolute copy number of test RNA was calculated. In all the four genes used for the in vivo study, efficient termination was observed. Quantitation revealed percentage termination to be around 80% and read-through is about 10e20% (Fig. 4B). Thus from both in vitro and in vivo experiments, it is
Please cite this article as: E. Ahmad et al., Revisiting intrinsic transcription termination in mycobacteria: U-tract downstream of secondary structure is dispensable for termination, Biochemical and Biophysical Research Communications, https://doi.org/10.1016/j.bbrc.2019.11.062
E. Ahmad et al. / Biochemical and Biophysical Research Communications xxx (xxxx) xxx
5
Fig. 3. U-tracts of intrinsic terminators are dispensable for termination. (A and B) RNA structures and sequences for the derivatives of predicted intrinsic terminator of Rv3183 and Lsr2 is depicted either having destabilized hairpin structures or reduced U-tract with appropriate DG. Transcription assays carried out with Rv3183-1 and Lsr2-1 having destabilized hairpin stem resulted in the absence of termination products. Experiments with Rv3183-2 and Lsr2-2 where the U-tract was reduced did not show any significant difference in termination.
apparent that non-canonical terminators function as terminators in M. tuberculosis.
3.5. Tuberculosis Rho-independent terminator (TRIT) motif is unlikely to function as a terminator Extracting genome sequences from 20 to þ 80 nt around gene termini from M. tuberculosis CDC1551, and using RNAfold routine from Vienna package [13], a new transcription termination motif in mycobacteria termed TRIT has been described [5]. TRITs account for ~81% of the significantly stable terminator sequences in mycobacteria [5]. Examination of a few of these sequences in the list revealed absence of similarity with text book defined intrinsic terminators or the non-canonical variants described above. Moreover, these predicted motifs have not been experimentally verified by carrying out standard termination assays either in vitro or in vivo. Hence, we examined the sequences having TRIT motif and compared them with canonical as well as non-canonical intrinsic terminators predicted by GeSTer and experimentally verified. Expectedly, we did not find high degree of similarities between the intrinsic terminators verses TRIT motif. Among all the predicted terminators, only tuf, Rv1324 and rpoC (3 out of 149) were validated as terminators (Table S2). Since most of the predicted or experimentally validated terminators do not have TRIT motif, we suggest that the main function of the motif is in some other cellular process and not transcription termination (See Discussion).
4. Discussion The textbook definition of intrinsic transcription termination is primarily based on E.coli paradigm. A G/C-rich stem loop structure, immediately followed by a stretch of 7e9 ‘U’ constituted the canonical intrinsic terminator [3,14]. For a long time, this notion of the termination structure prevailed as the subsequent analysis supported the model [15,16]. However, with the sequencing of a number of genomes, it was apparent that such canonical structures were under-represented in many organisms [6,7]. Most of the earlier algorithms that identified the terminators relied on the presence of these canonical motifs i.e. stem loop followed by Utract in RNA and hence failed to detect alternate terminator sequences in the genomic sequences leading to the suggestion for the operation of an altogether different termination mechanism [17]. However, GeSTer algorithm led to the identification of noncanonical intrinsic terminators that lacked or under-represented ‘U’ residues [7,9]. Now we have revisited the analysis of intrinsic terminators in mycobacteria to address the apparent differences arisen after Czyz et al. showed that the non-canonical terminators do not function as terminators [11]. They suggested following reasons for the discrepancy. 1) The RNAP used by Unniraman et al., 2001 was likely to have contained additional protein factors. 2) The transcript release assays would be necessary to assess actual terminators. Hence, they proposed that the observed termination products could be because of stalled complexes or nuclease contamination in the RNAP preparations. Now, using a pure
Please cite this article as: E. Ahmad et al., Revisiting intrinsic transcription termination in mycobacteria: U-tract downstream of secondary structure is dispensable for termination, Biochemical and Biophysical Research Communications, https://doi.org/10.1016/j.bbrc.2019.11.062
6
E. Ahmad et al. / Biochemical and Biophysical Research Communications xxx (xxxx) xxx
Fig. 4. In vivo assessment of termination efficiency for intrinsic terminators lacking U-tracts (A) Schematic depiction of primers designed to absolute quantify the RNA copy number for both terminated and read-through transcript by RT-qPCR. (B) Using the known concentration of the RNA standard curve was generated to estimate and compare the transcripts that represent the terminated and read-through products for bfrB, lsr2, mkl, and Rv3183. Termination efficiency was calculated using RNA copy number [Termination efficiency ¼ Termination (X-Y)/(Termination (X-Y) þ Read-through (Y))]. The data represented is mean ± SD from the three independent experiments.
preparation of the enzyme and two different bead assays for RNA release, we show that non-canonical terminator structures indeed function as true terminators. The GeSTer algorithm predicted terminators could terminate transcription irrespective of whether templates were immobilized or enzyme is immobilized. Moreover, the non-canonical terminators functioned as terminators both in vitro and in vivo. Notably, at least seven of the tested terminators were found to dissociate transcription elongation complexes in immobilized assays. However, not all predicted non-canonical terminator structures could terminate transcription. For example, as Czyz et al. showed [11], our termination assay with gyr A do not appear to release RNA in vitro (data not shown). It may be noted that for the current analysis RNAP used has been purified through different steps including affinity chromatography [12]. Hence the possibility of either having additional factors or nuclease contamination has been considered in the present experimental design. The differences between two reports could also be due to source of the RNAP (M. bovis verses M. smegmatis). Given the importance of transcription termination, new algorithms have been proposed to predict their occurrence. One such algorithm RNIE describes a new motif TRIT for transcription termination in mycobacteria [5]. Although nearly 150 terminator sequences have been identified by this approach, these predictions
were not subjected to experimental validation. Our present experimental analysis indicates that TRIT motif is unlikely to act as a true terminator. Very few of the GeSTer predicted, experimentally validated terminators have TRIT motif. Further even with the terminators studied by Czyz et al., only a very small subset of validated sequences have TRIT motif [11]. Although these motifs have the propensity for strong secondary structures, they lack the core features required for an intrinsic terminator in eubacteria. From all the studies so far, it is apparent that not all secondary structures would function as terminators. We have previously shown that stem loop structures with very low DG or very high DG are unlikely to function as terminators [10]. Thus, in order to function as terminators, species specific optimized DG of the structures appears to be necessary [10]. Although these predictive algorithms are important, the experimental validation of predicted structure is necessary. As the minimal requirement for an intrinsic terminator is a G/C-rich stem, while the U-tract facilitates the release of RNA, we note that very few of the TRIT motif appear to have G/C-rich stem which is an essential feature for termination hairpin loop within the termination bubble. Hence, it is not surprising that the structures with TRIT motif did not function as terminators in our analysis. We suggest that TRIT motif may function as some other important cellular signal instead of acting as terminator. RNA secondary
Please cite this article as: E. Ahmad et al., Revisiting intrinsic transcription termination in mycobacteria: U-tract downstream of secondary structure is dispensable for termination, Biochemical and Biophysical Research Communications, https://doi.org/10.1016/j.bbrc.2019.11.062
E. Ahmad et al. / Biochemical and Biophysical Research Communications xxx (xxxx) xxx
structures are known to function in generating stable 3’ ends of RNA after a termination step occurring downstream [18,19]. In concluding from our re-investigation of intrinsic termination in mycobacteria, it appears that some of the predicted noncanonical intrinsic terminators function as terminators both in vitro and in vivo. It is also evident that not all the GeSTeR predicted terminators function as terminators by themselves as suggested earlier [11]. Thus, intrinsic termination in eubacteria has retained the core characteristics across bacterial kingdom with species-specific optimized stem loop structures. Although U-tract present are important, they are dispensed away in organisms such as mycobacteria, which are G/C-rich, and have a paucity of canonical terminators. Experimental validation of the predicted structures would be important to ascertain their function as true terminators. Declaration of competing interest The authors declare no conflict of interest with the contents of this article. Acknowledgements The central facility of phosphor imaging of Indian Institute of Science, supported by Department of Biotechnology, Government of India is acknowledged. Financial support for the project is from a grant to V.N. from the Department of Biotechnology (DBT), Government of India, DBT sponsored Life Science Research, Education and Training at JNCASR, and DBT-IISc partnership program. V.N. is a J.C. Bose Fellow of the Department of Science and Technology, Government of India. Appendix A. Supplementary data Supplementary data to this article can be found online at https://doi.org/10.1016/j.bbrc.2019.11.062. References [1] I. Gusarov, E. Nudler, The mechanism of intrinsic transcription termination, Mol. Cell 3 (1999) 495e504, https://doi.org/10.1016/s1097-2765(00)80477-3. [2] R. Reynolds, R.M. Bermudez-Cruz, M.J. Chamberlin, Parameters affecting transcription termination by Escherichia coli RNA polymerase. I. Analysis of 13 rho-independent terminators, J. Mol. Biol. 224 (1992) 31e51, https://doi.org/
7
10.1016/0022-2836(92)90574-4. [3] J.W. Roberts, Mechanisms of bacterial transcription termination, J. Mol. Biol. (2019), https://doi.org/10.1016/j.jmb.2019.04.003. [4] M.D. Ermolaeva, H.G. Khalak, O. White, H.O. Smith, S.L. Salzberg, Prediction of transcription terminators in bacterial genomes, J. Mol. Biol. 301 (2000) 27e33, https://doi.org/10.1006/jmbi.2000.3836. [5] P.P. Gardner, L. Barquist, A. Bateman, E.P. Nawrocki, Z. Weinberg, RNIE: genome-wide prediction of bacterial intrinsic terminators, Nucleic Acids Res. 39 (2011) 5845e5852, https://doi.org/10.1093/nar/gkr168. [6] A. Mitra, K. Angamuthu, H.V. Jayashree, V. Nagaraja, Occurrence, divergence and evolution of intrinsic terminators across eubacteria, Genomics 94 (2009) 110e116, https://doi.org/10.1016/j.ygeno.2009.04.004. [7] S. Unniraman, R. Prakash, V. Nagaraja, Conserved economics of transcription termination in eubacteria, Nucleic Acids Res. 30 (2002) 675e684, https:// doi.org/10.1093/nar/30.3.675. [8] A. Mitra, A.K. Kesarwani, D. Pal, V. Nagaraja, WebGeSTer DBea transcription terminator database, Nucleic Acids Res. 39 (2011) D129eD135, https:// doi.org/10.1093/nar/gkq971. [9] S. Unniraman, R. Prakash, V. Nagaraja, Alternate paradigm for intrinsic transcription termination in eubacteria, J. Biol. Chem. 276 (2001) 41850e41855, https://doi.org/10.1074/jbc.M106252200. [10] A. Mitra, K. Angamuthu, V. Nagaraja, Genome-wide analysis of the intrinsic terminators of transcription across the genus Mycobacterium, Tuberculosis 88 (2008) 566e575, https://doi.org/10.1016/j.tube.2008.06.004. [11] A. Czyz, R.A. Mooney, A. Iaconi, R. Landick, Mycobacterial RNA polymerase requires a U-tract at intrinsic terminators and is aided by NusG at suboptimal terminators, mBio 5 (2014) e00931, https://doi.org/10.1128/mBio.00931-14. [12] A. China, V. Nagaraja, Purification of RNA polymerase from mycobacteria for optimized promoter-polymerase interactions, Protein Expr. Purif. 69 (2010) 235e242, https://doi.org/10.1016/j.pep.2009.09.022. [13] A.R. Gruber, R. Lorenz, S.H. Bernhart, R. Neubock, I.L. Hofacker, The Vienna RNA websuite, Nucleic Acids Res. 36 (2008) W70eW74, https://doi.org/ 10.1093/nar/gkn188. [14] Y. d’Aubenton Carafa, E. Brody, C. Thermes, Prediction of rho-independent Escherichia coli transcription terminators. A statistical analysis of their RNA stem-loop structures, J. Mol. Biol. 216 (1990) 835e858, https://doi.org/ 10.1016/s0022-2836(99)80005-9. [15] W.S. Yarnell, J.W. Roberts, Mechanism of intrinsic transcription termination and antitermination, Science 284 (1999) 611e615, https://doi.org/10.1126/ science.284.5414.611. [16] M.J.L. de Hoon, Y. Makita, K. Nakai, S. Miyano, Prediction of transcriptional terminators in Bacillus subtilis and related species, PLoS Comput. Biol. 1 (2005) e25, https://doi.org/10.1371/journal.pcbi.0010025. [17] T. Washio, J. Sasayama, M. Tomita, Analysis of complete genomes suggests that many prokaryotes do not rely on hairpin formation in transcription termination, Nucleic Acids Res. 26 (1998) 5456e5463, https://doi.org/ 10.1093/nar/26.23.5456. [18] D. Dar, R. Sorek, High-resolution RNA 3’-ends mapping of bacterial Rhodependent transcripts, Nucleic Acids Res. 46 (2018) 6797e6805, https:// doi.org/10.1093/nar/gky274. [19] X. Wang, M.P.A. N, H.J. Jeon, Y. Lee, J. He, S. Adhya, H.M. Lim, Processing generates 3’ ends of RNA masking transcription termination events in prokaryotes, Proc. Natl. Acad. Sci. U. S. A 116 (2019) 4440e4445, https://doi.org/ 10.1073/pnas.1813181116.
Please cite this article as: E. Ahmad et al., Revisiting intrinsic transcription termination in mycobacteria: U-tract downstream of secondary structure is dispensable for termination, Biochemical and Biophysical Research Communications, https://doi.org/10.1016/j.bbrc.2019.11.062