Sense overlapping transcripts in IS1341-type transposase genes are functional non-coding RNAs in archaea

Abstract

The existence of sense overlapping transcripts that share regulatory and coding information in the same genomic sequence shows an additional level of prokaryotic gene expression complexity. Here we report the discovery of ncRNAs associated with IS1341-type transposase (tnpB) genes, at the 3'-end of such elements, with examples in archaea and bacteria. Focusing on the model haloarchaeon Halobacterium salinarum NRC-1, we show the existence of sense overlapping transcripts (sotRNAs) for all its IS1341-type transposases. Publicly available transcriptome compendium show condition-dependent differential regulation between sotRNAs and their cognate genes. These sotRNAs allowed us to find a UUCA tetraloop motif that is present in other archaea (ncRNA family HgcC) and in a H. salinarum intergenic ncRNA derived from a palindrome associated transposable elements (PATE). Overexpression of one sotRNA and the PATE-derived RNA harboring the tetraloop motif improved H. salinarum growth, indicating that these ncRNAs are functional.

Introduction

Overlapping genomic signals have been discovered across all domains of life, showing that DNA sequences are multifunctional, harboring regulatory and coding information simultaneously. Transcriptome complexity includes not only interweaving promoter elements and intergenic transcription but also a plethora of transcripts that overlap annotated genes.^1,2 Sense and antisense non-coding (ncRNA) transcripts have been identified, with a clear predominance of antisense transcripts characterization. Antisense ncRNAs are ubiquitous and affect diverse cellular processes, from transcription initiation to translation efficiency³; they have also been shown to regulate transposition events.⁴ Even with the widespread use of strand-specific sequencing technologies, the identification and characterization of sense overlapping transcripts have expanded more modestly, since the signals for these molecules are convoluted with mRNA signals and thus, can be difficult to detect.

Recently, we have identified sense ncRNAs overlapping the 5′ end of coding sequences in archaea with focus in the haloarchaeon Halobacterium salinarum NRC-1.⁵ This extremophile thrives in environments with 4.5M NaCl concentration and has been used as a model organism for systems biology studies. As such, a massive amount of microarray and tiling array-based transcriptome data in diverse environmental conditions is currently available for this organism.^6-11 In the present work, by revisiting the dataset from Koide et al.⁸, where transcriptome dynamics was assessed over a growth curve using high resolution tiling array data, we detected signals that could indicate the presence of sense overlapping transcripts inside IS1341-type transposase genes in H. salinarum.

IS1341-type transposases genes are also known in the literature as tnpB or orfB. They are part of the widespread IS200/605 family, the most ancient family in the archaeal domain of life.¹² Based on gene composition, IS200/605 family can be classified into 3 distinct groups: (a) IS200 group, with only the tnpA gene (also known as the IS200 gene); (b) IS605 group, where tnpA and tnpB are found adjacent and (c) IS1341 group, where tnpB is found alone.^13,14 Transposition of IS200/605 members requires the protein TnpA, which binds conserved palindromic sequences at the left end (LE) and right end (RE) of the insertion element.^15,16 The mechanism of transposition is unique among IS elements since it requires single stranded DNA as substrates.^15,17 The tnpB gene is not required for transposition^14,18, however, it is known to play a regulatory role by inhibiting excision and insertion of ISDra2 in Deinococcus radiodurans.¹⁹ Mechanisms of TnpB action are currently unknown.

Sense overlapping transcripts in IS1341-type transposase genes have been detected in other archaea as part of studies for the identification of ncRNAs. In Sulfolobus solfataricus, 2 ncRNAs overlapping transposases in the same strand were identified and in Thermococcus kodakarensis, 3 RNAs transcribed from transposase loci were detected.^20,21 In Pyrococcus abyssi, ncRNAs are transcribed from intergenic repetitive loci and have sequence similarity to the 3′ end of IS1341-type transposases.²² This apparent conservation of sense overlapping RNAs and similar sequences in intergenic ncRNAs prompted us to investigate not only the presence of these transcripts in H. salinarum NRC-1, but also to evaluate if they have a functional role. In the present work, we show that all IS1341-type transposases of H. salinarum possess sense overlapping transcripts, explore their expression profiles in diverse conditions, and show a conserved tetraloop motif that might be relevant for function of these ncRNAs.

Results and Discussion

H. salinarum has a large number of insertion sequences (114 according to ISfinder and ISbrowser^23,24, https://www-is.biotoul.fr), which are believed to be responsible for the unusual genome plasticity observed in this organism.²⁵ In H. salinarum NRC-1, reannotation based on R1 strain identified 10 IS1341-type transposases.²⁶ Five of these IS1341-type transposases are located adjacent to tnpA, 1 in divergent and 4 in convergent orientation (Table 1).

Table 1.

IS1341-type transposases (tnpB genes) analyzed in this work and their respective sense overlapping transcripts (sotRNAs)

Gene	IS200/605 family group	chromosome	tnpB start codon	tnpB stop codon	strand	sotRNA transcript start	sotRNA transcript end	sotRNA size [nt]	R1 strain equivalent genes
VNG0013C	IS1341	chr	12734	11478	reverse	11641	11428	214	OE1019R
VNG0026C	IS1341	chr	21789	20533	reverse	20701	20498	204	OE1040R
VNG0042G	IS1341	chr	35931	34651	reverse	34775	34570	206	OE1070R
VNG0044H	IS605 (VNG0043H – VNG0044H), convergent	chr	36715	37971	forward	37813	38042	230	OE1074F
VNG0286C	IS605 (VNG0285C – VNG0286C), convergent	chr	229443	230597	forward	230444	230629	186	OE1440F
VNG2652H	IS605 (VNG2653C – VNG2652H), divergent	chr	1989188	1987884	reverse	1988005	1987819	187	OE4727R
VNG6181H	IS605 (VNG6182H – VNG6181H), convergent	pNRC200	155603	154368	reverse	154508	154259	250	OE5062R
VNG6221H	IS605 (VNG6222H – VNG6221H), convergent	pNRC200	183054	181795	reverse	182008	181724	285	OE5102R
VNG6361G	IS1341	pNRC200	287504	288844	forward	288696	288915	220	OE6034F
VNG6406H	IS1341	pNRC200	316931	315786	reverse	315833	315631	203	OE6089R

Analyzing publicly available tiling microarray data obtained in reference growth condition for H. salinarum NRC-1⁸, we discovered a conserved transcriptional signature in all 10 IS1341-type transposases genes. This signature features a change in the expression signal inside the insertion sequence near the 3′ end, indicating the existence of possible sense overlapping transcripts (sotRNAs). Figure 1A-I illustrates this feature for VNG0042G. When the dynamics of these transcripts were analyzed over a growth curve, the putative sense overlapping transcripts showed an increase in relative expression levels, whereas the remaining of IS1341-type transposase gene showed decreasing levels (Fig. 1A–II). Tiling array data from TfbD transcription factor overexpression showed that sotRNAs presented decreasing expression levels, whereas IS1341-type genes presented increasing levels, stating an opposite regulation in varied conditions (Fig. 1A–III).⁸ Transcription signatures in the reference condition, together with the dynamical information from the differential expression profiles provided evidence for the existence of sense overlapping transcripts at the 3′ end of IS1341-type transposases genes in Halobacterium salinarum NRC-1 (Fig. 1A, B).

Figure 1.

Identification of sotRNAs in IS1341-type transposases using tiling array data. (A) (I) Tiling array signal in reference condition for VNG0042G and VNG_sot0042; (II) expression profiles over the growth curve⁸; (III) expression profiles during TfbD overexpression.⁸ Heatmaps are color-coded according to log₁₀ expression ratios between each of the 13 time points relative to reference condition. (B) Tiling array signal in reference condition and expression profiles of the remaining 9 IS1341-type transposases (arrows in yellow for genes in forward strand and in orange for genes in reverse strand) and their sotRNAs (light blue arrows) over the growth curve and TfbD overexpression, as described in (A).

To validate their existence and define their boundaries with precision, we analyzed previously published small RNA-seq data (sRNA-seq).⁵ Figure 2A shows sRNA-seq based discovery of boundaries for VNG0042G associated sotRNA. An extended analysis (Fig. S1A) showed that all IS1341-type transposases in Halobacterium salinarum NRC-1 have an associated sotRNA. In addition, it was possible to map the 3′ and 5′ ends of all sotRNAs (Table 1) (Fig. S1A). We named these discovered features using “VNG_sot” as prefix and making reference to their cognate gene in the suffix. All sotRNAs mapped in this study start inside the coding sequence of IS1341-type transposase genes, at approximately 1100 nt from the start codon, with an average size of 218 nt and end, on average, 74 nt after gene's stop codon (Table 1).

Figure 2.

Analysis of sRNA-seq and expression profiles for VNG0042G and VNG_sot0042. (A) Strand-specific sRNA-seq data⁵ was used to map the 5' end of sotRNAs in IS1341-type transposases of H. salinarum NRC-1. sRNA-seq data is visualized as log₂ of total reads aligned in each genomic position for VNG0042G and VNG_sot0042. Enrichment of 5' ends of mapped reads are visualized as peaks right below small RNA-seq coverage. Light blue arrow: sotRNA. Dark orange arrows: genes annotated on reverse strand. (B) Expression profiles of VNG0042G (orange) and VNG_sot0042 (blue) in different environmental and genetic backgrounds. Each pair of orange/blue columns represent one published experimental condition. Data for the remaining genes and for each individual experiment is available in Supplementary Figure 1. Data for intergenic ncRNAs can be found in Figure 4 and Supplementary Figure 7.

Two targets were selected for further experimental validations: VNG_sot0042 and VNG_sot2652. C-RACE experiment for VNG_sot0042 confirmed its 5′ and 3′ ends (Fig. S2) and primer extension for VNG_sot2652 confirmed its 5′ end (Fig. S3).

To further investigate the expression profiles of sotRNAs and IS1341-type transposases, we considered all tiling microarray data sets available in public databases (see Materials and Methods) and focused on probes designed to measure expression levels in the same region defined by sRNA-seq analysis. All sotRNAs were found to be expressed with diverse patterns of modulation, as would be expected for a typical functional transcript (Fig. S1B). Most of the sotRNAs were differentially expressed relative to the remaining signal of IS1341-type transposase cognate gene in different environmental conditions, which could possibly indicate some kind of regulation between these elements. Figure 2B shows the data for VNG0042G and its sotRNA. There is one case where the sotRNA (VNG_sot0286) is located antisense to the probable 3′ UTR of VNG0287a (a putative RNA-binding protein). Analysis of their expression profiles over different conditions show that VNG_sot0286 and VNG0287a have anti-correlated expression profiles in some conditions, which could suggest a cis-antisense interaction (Fig. S4). None of the other 9 sotRNAs mapped in H. salinarum NRC-1 have this putative pattern for cis-antisense regulation.

The available data cannot conclusively differentiate between the 2 most probable sotRNA biogenesis hypotheses: primary transcription or processing products. When the upstream regions of the sotRNAs were analyzed, no classical BRE-TATA promoter sequence was found. Northern-blot experiments for 2 sotRNAs (VNG_sot0042 and VNG_sot2652, Fig. S5) might argue for processing, since the full-length transcript is not detected. Also, deep-sequencing data from primary transcript enriched libraries (dRNA-seq)⁵ only shows evidence of primary transcript enrichment for VNG_sot0026. However, it was previously reported that dRNA-seq data is prone to false negative results²⁷ which reinforces the finding that VNG_sot0026 is a primary transcript but does not necessarily disprove the others. In addition, some sotRNAs were also shown as primary transcripts in T. kodakarensis.²¹ Chromatin immunoprecipitation data (ChIP-chip) points to transcription factor binding sites at upstream vicinity of some sotRNAs along with statistically derived H. salinarum gene regulatory elements (Table S1).^6,7,41 Finally, the distinct expression patterns of sotRNAs and their cognate genes across several experiments are unlikely to be generated by processing a single precursor molecule.

The presence in H. salinarum NRC-1 of a probable sotRNA in a nonfunctional IS1341-type transposase gene (Fig. S1C), a truncated small 120 bp long pseudogene identical to OE5220R, suggests that sotRNA presence may somehow be the reason why these defective elements have not been lost from the genomes. Mining public available transcriptome data for other archaea and bacteria, we found that transcripts overlapping 3′ ends of tnpB are more prevalent than previously appreciated. Besides the previously described S. solfataricus P2²⁰ and T. kodakarensis KOD1²¹ cases, we found examples in all archaea that contain IS1341 group genes and RNA-seq data available at NCBI's SRA database: Methanopyrus kandleri AV19, S. acidocaldarius DSM639 and P. furiosus DSM3638. Moreover, we found bacterial examples in Helicobacter pylori 26695 and Escherichia coli K12 data (Figure S6, File S1). These observations suggest a conserved role for sotRNAs in prokaryotes. Computational secondary structure analysis reveals a very distinctive feature among almost all sotRNAs in H. salinarum: the presence of a stable hairpin motif near the end of the sequence. Using five different secondary structure prediction algorithms on each RNA sequence, it was possible to identify: (i) a clear tetraloop motif comprising an 8 bp long stem with a UUCA loop (VNG_sot0042, VNG_sot0044, VNG_sot6181 and VNG_sot6221); (ii) single nucleotide variants of this tetraloop (VNG_sot0286 with UUUA, VNG_sot0013 with CUCA and VNG_sot0026 with GUCA) (Fig. 3); and (iii) stem-loop structures created by other sequences (VNG_sot6406 with a 3 nt long loop in a 6 bp stem, VNG_sot2652 with a 6 nt long loop in a 10 bp stem, and VNG_sot6361 with a 3 nt long loop in a 4 bp stem) (File S2). While the 5 algorithms are considerably different in their approaches and underlying assumptions to predict secondary structures²⁸, they all made consistent predictions of tetraloops and collectively added to the significance of our discovery of this feature.

Figure 3.

Tetraloop motif found in most sotRNAs in H. salinarum NRC-1. Colors represent base pairing groups and color saturation corresponds to conservation, parenthesis correspond to pairing and dots to unpaired regions. Gray blocks below the sequences represent a histogram of the most conserved nucleotide. Raw data is available at Supplementary File 2.

One of the key characteristics of IS200/605 family is the presence of conserved imperfect palindromic sequences that form a hairpin-like structure at DNA level at both left and right ends of the insertion elements, sometimes just upstream of CDS stop codons.^15,16,19 Therefore, IS1341-type transposase genes should harbor the right end (RE) of the IS element, which may explain the sotRNA motifs found as a retained structure formed at DNA level, also stable at RNA level. We named the motif found, the RE-like tetraloop.

Similarly, in P. abyssi, sequence alignments of intergenic ncRNAs sRK48 and sRK52 with other thermococcal genomes showed similar conserved structures and the most conserved one (P1 loop)²² corresponds to the RE-like tetraloop, reinforcing the probable functionality of this region in archaea. A sequence similarity search for the RE-like tetraloops in H. salinarum NRC-1 genome showed that, besides all sotRNAs regions, 3 intergenic regions also have this secondary structure: 14335 to 14359 on main chromosome, 268252 to 268281 on pNRC200 plasmid, and 19532 to 19556 on main chromosome (File S3). From these 3, only the first 2 had sufficient sRNA-seq read coverage signal: VNG_R0052 and VNG_R6334 (Table 2). We determined the 5′ and 3′ end of these transcripts using RNA-seq data (Fig. 4A and Fig. S7A) and verified the expression profiles over diverse environmental and genetic perturbations. Intergenic ncRNAs showed differential expression, with trends similar to sotRNAs (Fig. 4B and Fig. S7B).

Table 2.

Intergenic RNAs harboring the RE-like tetraloop motif

ncRNA	VNG_R0052	VNG_R6334
Start	14397	268030
End	14299	268304
Length	99	275
Chr	chr	pNRC200
Strand	reverse	forward
RE-like Start	14359	268256
RE-like End	14340	268277

Figure 4.

(A) Strand-specific sRNA-seq data⁵ was used to map 5' ends of intergenic RNAs in H. salinarum NRC-1. sRNA-seq data is visualized as log₂ of total reads aligned in each genomic position for VNG_R0052. Enrichment of 5' ends of mapped reads are visualized as peaks right below small RNA-seq coverage profile. (B) Expression profiles of VNG_R0052 in different environmental and genetic backgrounds.

When we analyze the sequence of the intergenic ncRNA VNG_R0052, we can find not only the stem loop corresponding to the RE of the IS element, but also, fused to it at the 5′ end, we can find part of the left end (LE) sequence of IS200/605 family (File S2). This phenomena has been previously observed in Haloquadratum walsbyi as a result of transposase deletion, generating elements that are denominated PATEs (palindrome associated transposable elements).²⁹ We show that a PATE element is indeed expressed at the RNA level and similarity search in the NCBI database shows that VNG_R0052 is not an exception: fused LE and RE are common among haloarchaea (File S4). We predict that, as in VNG_R0052 case, PATEs might also be functionally transcribed in other organisms. Moreover, since PATEs have counterparts in bacteria (REPINS, found for example in Pseudomonas fluorescens³⁰), we extend this prediction to bacteria.

Interestingly, the RE-like tetraloop is a previously unannotated structural motif in the HgcC ncRNA family (RFAM v12 accession code RF00062) (Fig. S8). HgcC family was first discovered in the genome of Methanococcus jannaschii by an in silico analysis designed for the identification of high guanine-cytosine content regions.³¹ Yet, functions of this ncRNA family are still unknown, thus the presence of the RE-like tetraloop inside RNAs from this family might help further studies focusing on elucidating their function.

To test if RNAs harboring this conserved structural motif may have some functional role, we overexpressed one sotRNA (VNG_sot0042) and one intergenic RNA (VNG_R0052), and analyzed the consequences on growth dynamics in standard batch culture. Remarkably, both strains harboring the RE-like tetraloops had improved growth relative to the control strain providing preliminary evidence that this conserved motif is somehow functional in H. salinarum NRC-1 (Fig. 5). In the future, the specific mechanism by which the tetraloop acts can potentially be investigated by analyzing the effect of systematically disrupting its stem and loop structures on this growth phenotype.

Figure 5.

Growth curve of 4 different strains of H. salinarum NRC-1 under reference conditions. Strains overexpressing RNAs (VNG_R0052 and VNG_sot0042) that harbors the RE-like tetraloop present higher growth when compared to control strains (NRC-1, wild type and PMTF-0, which is the wild type strain with an empty overexpression vector).

The exact role of IS1341-type transposases is still an open problem but recent reports indicate that they have an inhibitory effect on IS excision and insertion.¹⁹ Given that all transcripts analyzed in this work (sotRNAs and a PATE-derived RNA) harbor similar motifs, we might speculate whether these molecules are related to this mechanism. A possibility is that this motif could work as a competitor for TnpA binding, and thus, making it less available for binding at single stranded DNA regions. PATE-derived RNAs could also present similar functionality. Thus, IS1341 group of transposases could act as reservoirs of stem loops, as previously hypothesized.³² Further experimental work would be required to test this hypothesis, nevertheless, our data point to the functionality of ncRNAs derived from IS1341-type transposases that harbor the conserved RE-like tetraloop motif in their sequences.

Conclusions

In the present work, we report the discovery of sense overlapping transcripts in IS1341-type transposases in H. salinarum NRC-1 and show that they are generally differentially expressed relative to their cognate transposase gene in diverse environmental conditions. In addition, we detected the presence of a conserved tetraloop motif in several sotRNAs and in intergenic ncRNAs that correspond to palindrome associated transposable elements (PATEs). We provide evidence that transcripts harboring this conserved tetraloop motif, which are very similar to right end (RE) of the insertion sequence, are probably functional since their overexpression improved H. salinarum NRC-1 growth.

Materials and Methods

H. salinarum NRC-1 transcriptome analysis

Publicly available RNA-seq data was used to locate ncRNA transcripts with high spatial resolution. Publicly available tiling microarray data was used to interrogate expression dynamics over time or combinatorial experimental conditions for the ncRNAs found in this work. The RNA-seq data is composed by deep-sequencing of replicated sRNA enriched libraries (<200 bp) obtained from standard H. salinarum NRC-1 growth conditions⁵ available at NCBI's SRA database (SRX433542). Tiling microarray data is a compendium of all publicly available datasets (Aug/2014) obtained by different platforms (NCBI's GEO database accession numbers: GPL17005, GPL13426, GPL7369, GPL7255 and GPL8468). The data set used were: 13 points of a standard growth-curve (GSE12923); overexpression of the TfbD transcription factor (GSE15788)⁸; knockouts of transcription factor Tfb and overexpression of synthetically enhanced versions of it (GSE31308)⁹; knockout of ribonuclease VNG2099C (GSE45988)¹⁰; variation of extracellular salinity concentrations for 3 different salts (GSE53544).¹¹ Normalized expression ratios appropriate for each study were obtained directly from GEO records as published. Growth curve dynamics and TfbD overexpression tiling array data were visualized in detail at probe level using Gaggle Genome Browser.³³

Sequence similarity search in public databases

Sense overlapping transcripts for specific tnpB genes were found in different organism using standard BLASTn-SRA suite sequence similarity search. The RNA-seq datasets considered comprised archaeal and bacterial data: S. acidocaldarius DSM639 (SRA accession SRX327494), Methanopyrus kandleri AV19 (accession SRP019041), H. pylori 26695 (accessions SRX014018, SRX014030 to SRX014033 and SRX014050) and E. coli K12 (accession SRX254762). H. salinarum NRC-1 intergenic ncRNA containing RE-like tetraloop motifs where found through standard BLASTn searches using all 10 sotRNAs sequences as queries.

RNA secondary structure computational analysis

RNA sequences of all 10 sotRNAs and 2 intergenic ncRNAs were used for structural predictions using 5 different methods: RNAfold-MFE, which predicts the Minimum Free Energy (MFE) structure; RNAfold using thermodynamic partition function centroids algorithm³⁴; ContextFold, which uses a very large number of parameters in a machine learning approach³⁵; Vsfold5, which allow the prediction of pseudoknots by using sequential (5′ to 3′) folding and thermodynamically most-probable folding pathways³⁶ and Cylofold, which simulates the 3D folding process in a coarse-grain model.³⁷ All predictions are shown in extended dot-bracket format to include pseudoknots. All 5 methods were run using their respective web-server versions and default parameters.

Strains and culturing

H. salinarum NRC-1 was grown in enriched complex media (CM) consisting of 25% NaCl, 2% MgSO₄.7H₂O, 0.2% KCl, 0.3% Na-citrate and 1% peptone, at 37°C with constant light and agitation of 225 rpm. When required, the media was supplemented with 20 μg/mL of mevinolin (A.G. Scientific, San Diego, CA). E. coli strain DH5α was grown at 37°C with air shaking at 225 rpm in Luria-Bertani media and supplemented with 50 μg/mL of carbenicilin (Sigma) when necessary.³⁸

RNA preparation

Total RNA was prepared using the MirVana RNA extraction kit (Ambion) and treated with Turbo DNAse I (Fermentas). Samples were analyzed by PCR to rule out DNA contamination.

Primer extension

10 μg of RNA was hybridized with 1pmol of 5′ fluorophore (VIC or NED) labeled oligonucleotide and 1 μl of 10 mM dNTPs in a total volume of 10μl (Temperature cycle for annealing: 95°C 5′, 62°C 15′, 59°C 15′, 53°C 15′, 50°C 15′). After annealing, a Synthesis Mix 2 μl 10x RT Buffer (200 mM Tris-HCl (pH 8.4), 500 mM KCl); 4 μl MgCl₂(25 mM); 2 ul 0.1 M DTT; 1 μl RNAse Out; 1 μl SuperScript III (Invitrogen) was added and the reaction was incubated at 50°C for 50′. The cDNA was purified and analyzed in an ABI 3500XL (Applied Biosystems). Data processing was made using Gene Mapper V5.0 (Life Technologies).

Northern blot

For Northern blot analyzes, 50 μg of total RNA treated with RNase-free DNAseI (Fermentas) was separated on polyacrylamide gel (8% acrylamide:bisacrylamide [29:1], 8M urea, 1xTris–borate–EDTA buffer). RNAs were transferred to Hybond-N+ membranes (GE Healthcare) and hybridized with ³²P-labeled oligonucleotides (5′-GACGTAGCCTGTTCTTGAGGTGCGAACGCACCAGTTTCAT-3′ for VNG0042G and VNG_sot0042; 5′-CCCTGTTCCGCGAGGGAACGGGATGAGGTTGGGCGGCTTA-3′ for VNG2652H and VNG_sot2652) using Rapid-hyb buffer (GE Healthcare). Signals were detected by autoradiography using a M35A X-Omat Processor (Kodak).

Circularization and RACE (Rapid Amplification of cDNA Ends)

5′ and 3′ ends of RNAs were determined by RNA circularization.³⁹ Briefly, RNA free of DNA contamination was treated with Tobacco Acid Phosphatase (Epicentre) following manufacturer's instructions and treated with T4 RNA ligase (Thermo Scientific) for intramolecular ligation of 5′ and 3′ ends. cDNA was synthesized using SuperScript III First Strand Synthesis System (Life Technologies) using random hexamers. Divergent PCR (CircVNG0042-F —- 5′-ATGGTTTACGACGTAGCC-TGTTC −3′, CircVNG0042-R —- 5′-GTTCCTGTTTGACAATGAAACTG -3′) using circular cDNA was made using Kapa Taq ReadyMix (Kapa Biosystems). Amplification products were purified from agarose gels and cloned into pGEM T-easy Vector (PROMEGA) for sequencing. RACE assays were performed as previously described⁴⁰ using 5′/3′ RACE Kit 2nd Generation (Roche) for VNG_R0052 using the following primer sequences: RSP1 — 5′ - GAAGGGCGAGGCTTTCTTCTC –- 3′, RSP2 –- 5′ – CGAGGCTTTCTTCTCGATTC –- 3′, RSP3 –- 5′ - CTGTTACGGAGAATCGAGAAG –- 3′, RSP4 –- 5′ GAATCGAGAAGAAAGCCTCG — 3′.

General cloning procedures

VNG_R0052 and VNG_sot0042 were cloned into pMTF overexpression vector for H. salinarum NRC-1.⁴¹ DNA fragments were amplified by PCR using 50 ng of H. salinarum NRC-1 genomic DNA, 10pmol of specific oligos, 6.25 μl of 2x KAPA2G Fast ReadyMix (with loading dye) (Product Number: KK5102), 0.25 μl of MgCl₂ (25 mM), 3.875 μl of DMSO and DEPC treated water to a final volume of 12.5μl. The oligos used for VNG_R0052 were 5′-GACATATGCGGAGAATCGAGAAGA AAGCC-3′ and 5′-GGGGATCCTCAGTCTAAACT-TCTGGGGTGG-3′; and for VNG_sot0042 were 5′-GCCGGGCATATGACGTATCTCCGAGTCCCGTCA-3′ and 5′-GCGGGGGGATCCACAGGATAGCTGAGACTGCAGCAT-3′. The selected oligos added a BamHI site located at the 5′ region and a NdeI site located at 3′ region. PCR products were purified from 1% agarose gel by illustra GFX PCR DNA and Gel Extraction Kit (GE Healthcare). All fragments were cloned as BamHI/NdeI fragments into pMTF vector and transformed in chemically competent E. coli DH5α. Plasmidial DNA was extracted using Wizard Plus SV Miniprep Kits (Promega) and sequenced using BigDye Terminator v3.1 Cycle Sequencing Kit (Applied Biosystems). The confirmed plasmids were used to transform H. salinarum NRC-1. Transformant selection was obtained by plating the cells into solid CM supplemented with mevinolin (20 μg/ml).

Growth curves

For phenotypical characterization, isolated colonies of each H. salinarum strain were grown in liquid CM, supplemented with 20μg/ml of mevinolin when needed, until they reached OD600 = 0.5. Each culture was diluted to OD600 = 0.05 and grown under standard conditions for 8 days, all growth curves were made in 2 biological and technical replicates.⁴²

Disclosure of Potential Conflicts of Interest

No potential conflicts of interest were disclosed.

Acknowledgments

We thank Silvia Epifânio for technical support. We thank reviewers for constructive criticisms that improved our work.

Funding

This work was supported by Projeto Jovem Pesquisador em Centros Emergentes da Fundação de Amparo à Pesquisa do Estado de São Paulo (FAPESP, http://fapespbr/en/) [09/09532–0 to TK]; Edital Universal do Conselho Nacional de Desenvolvimento Científico e Tecnológico (CNPq) [473660/2013–0 to TK, 470120/2009–6 to TK, 476724/2013–9 to RZNV]; Fundação de Apoio ao Ensino, Pesquisa e Assistência do Hospital das Clínicas da Faculdade de Medicina de Ribeirão Preto da Universidade de São Paulo (FAEPA) [1640/2009 to TK]; Núcleo de Pesquisa em Ciência Genômica (NAP-CG) da Universidade de São Paulo; and fellowships FAPESP [11/07487–7 to LSZ, 2013/21522–5 to JVGF, 10/10195–5 to VCSI].

Supplemental Material

Supplemental data for this article can be accessed on the publisher's website.