- Research article
Biocomputational prediction of non-coding RNAs in model cyanobacteria
BMC Genomicsvolume 10, Article number: 123 (2009)
In bacteria, non-coding RNAs (ncRNA) are crucial regulators of gene expression, controlling various stress responses, virulence, and motility. Previous work revealed a relatively high number of ncRNAs in some marine cyanobacteria. However, for efficient genetic and biochemical analysis it would be desirable to identify a set of ncRNA candidate genes in model cyanobacteria that are easy to manipulate and for which extended mutant, transcriptomic and proteomic data sets are available.
Here we have used comparative genome analysis for the biocomputational prediction of ncRNA genes and other sequence/structure-conserved elements in intergenic regions of the three unicellular model cyanobacteria Synechocystis PCC6803, Synechococcus elongatus PCC6301 and Thermosynechococcus elongatus BP1 plus the toxic Microcystis aeruginosa NIES843. The unfiltered numbers of predicted elements in these strains is 383, 168, 168, and 809, respectively, combined into 443 sequence clusters, whereas the numbers of individual elements with high support are 94, 56, 64, and 406, respectively. Removing also transposon-associated repeats, finally 78, 53, 42 and 168 sequences, respectively, are left belonging to 109 different clusters in the data set. Experimental analysis of selected ncRNA candidates in Synechocystis PCC6803 validated new ncRNAs originating from the fabF-hoxH and apcC-prmA intergenic spacers and three highly expressed ncRNAs belonging to the Yfr2 family of ncRNAs. Yfr2a promoter-luxAB fusions confirmed a very strong activity of this promoter and indicated a stimulation of expression if the cultures were exposed to elevated light intensities.
Comparison to entries in Rfam and experimental testing of selected ncRNA candidates in Synechocystis PCC6803 indicate a high reliability of the current prediction, despite some contamination by the high number of repetitive sequences in some of these species. In particular, we identified in the four species altogether 8 new ncRNA homologs belonging to the Yfr2 family of ncRNAs. Modelling of RNA secondary structures indicated two conserved single-stranded sequence motifs that might be involved in RNA-protein interactions or in the recognition of target RNAs. Since our analysis has been restricted to find ncRNA candidates with a reasonable high degree of conservation among these four cyanobacteria, there might be many more, requiring direct experimental approaches for their identification.
In bacteria, non-coding RNAs (ncRNAs) are a heterogeneous group of sequence-specific regulators of gene expression, normally lacking a protein-coding function. They are typically 50–250 nucleotides in length , and regulate mRNA translation or decay but sometimes also directly modulate certain protein functions. Most stress responses in the organism best-studied in this respect, E. coli, include at least one small regulatory RNA as part of the regulon . However, their functions also include the control of plasmid and viral replication , bacterial virulence , quorum sensing , or the acquired resistance against bacteriophages .
In many cases, these ncRNAs function through sequence-specific base pairing; hence they frequently have a (partial) base complementarity to their target RNA molecules. The vast majority of known ncRNAs is encoded at genomic locations far away from their target genes. However, some ncRNAs are transcribed from the reverse complementary strand of the respective target and hence these are fully or partially overlapping with their target RNAs, constituting the class of antisense RNAs. Except for the more common types of ncRNA (ribosomal RNA, tRNA, tmRNA, 6S RNA, RNAse P RNA and ffs RNA), genes encoding ncRNAs are not annotated during standard genome analysis. The efforts to accomplish their identification in bacteria can broadly be divided into (i) sequencing the population of small RNAs or (ii) prediction by bioinformatics tools (mostly) followed by experimental verification (see  for review). As a result of such systematic searches, more than 80 ncRNAs are now known in E. coli, most of which had been overlooked by traditional genome analysis.
Cyanobacteria currently raise considerable interest as they perform oxygenic photosynthesis, fix atmospheric CO2 and nitrogen, frequently produce large quantities of bioactive secondary metabolites and due to their potential for the production of biofuels. As long as there is sufficient light available for photosynthesis, cyanobacteria populate widely diverse environments such as freshwater, the oceans, rock surfaces, desert soil or the polar regions. Their adaptation to vastly different environmental conditions suggests the existence of sophisticated regulatory mechanisms. Therefore, various types of regulatory RNA can be expected that interplay with the different signal transduction pathways and stress responses. Indeed, computational-experimental screens based on comparative genome analysis identified seven different ncRNAs in the marine cyanobacteria Prochlorococcus and Synechococcus  which were called Yfr1-7 for cY anobacterial F unctional R NA. In a follow-up study making use of high density microarrays and exploiting the genome information from meanwhile 12 different Prochlorococcus genome sequences, additional 14 ncRNAs and 24 antisense RNAs were found . Unicellular marine cyanobacteria of these genera provide an excellent dataset for computational predictions that require comparative genome information since currently 22 different genome sequences from very closely related isolates are available [10, 11]. However, a major bottleneck in the work with these marine cyanobacteria is that despite some recent progress , protocols for genetic manipulation are very slow or not available at all. Therefore, the finding that two of these ncRNAs are phylogenetically widely distributed enabled direct genetic work on their functional relevance: Yfr1 is distributed throughout the cyanobacterial radiation  and might play a role in the adaptation to redox stress or the regulation of carbon uptake , whereas Yfr7 was identified as the homolog of the 6S RNA  which is found in all eubacteria . However, for efficient genetic and biochemical analysis of cyanobacterial ncRNAs it would be very desirable to identify a set of ncRNA candidate genes in model cyanobacteria that are easy to manipulate and for which extended mutant, transcriptomic and proteomic data sets are available. In addition to Yfr1 which exists in all four unicellular cyanobacteria targeted here , the only currently known ncRNAs in model cyanobacteria are an antisense RNA covering the ferric uptake regulator gene furA in Anabaena PCC 7120 over its full length , and the antisense RNA IsrR, regulating the gene for the light-absorbing protein IsiA under conditions of iron limitation and redox stress in the unicellular Synechocystis PCC 6803 .
In recent years, comparative genomics-based prediction of ncRNA genes has become a standard method to search for such genes within bacterial genomes [8, 19–23]. Thus, the availability of genome sequences from closely enough related species is a critical factor as is the conservation of ncRNAs. In case of unicellular cyanobacterial model organisms, the lack of genome sequences from close relatives has been hampering such studies. With the recent release of the Microcystis aeruginosa NIES-843 genome , however, a cyanobacterium relatively close to Synechocystis has been sequenced.
Here we set out to identify possible ncRNA genes and other RNA elements (5' leader sequences and riboswitches) in the three unicellular model cyanobacteria Synechocystis PCC6803, Synechococcus elongatus PCC6301 and Thermosynechococcus elongatus BP1 plus the toxic Microcystis aeruginosa NIES-843 (from now on: Synechocystis, Synechococcus, Thermosynechococcus and Microcystis) by biocomputational comparative genome analysis with a focus on Synechocystis.
Results and Discussion
Computational screening for novel ncRNAs
To screen for novel RNA elements, all intergenic regions >50 nt were extracted from the four genomes and analyzed as outlined in Fig. 1, leaving out all annotated genes, including tRNA and four structural RNA genes. Although single sequence elements and families of sequence elements that are specific to a single genome cannot be found by our approach, this procedure initially returned 443 predicted clusters holding 1528 individual sequences, from which 383 belonged to Synechocystis, 168 to Synechococcus, 168 to Thermosynechococcus and 809 to Microcystis. However, these numbers were diminished in further filtering steps (see below).
The analysis was basically focused on sequence and structure similarities. Detailed information on all clusters predicted by our method including the positions of all sequences is available online . This information, which we show exemplarily in the inset in Fig. 1, includes the location within the compared genomes, flanking genes, a secondary structure prediction as well as Z-scores and probabilities in either forward (Z, P) or reverse (Z rev, P rev) orientation as computed by RNAz . Furthermore we conducted searches against Rfam , the database collecting ncRNAs, and TransTermHP , of which the results are also given in the online material.
High-scoring putative RNA elements
Filtering with P > = 0.5 or Z < = -2.0 reduced the initial number of 1528 individual sequences in the 443 predicted clusters to 113 sequence clusters with 620 individual elements, 94, 56, 64 and 406 in Synechocystis, Synechococcus, Thermosynechococcus and Microcystis, respectively. A summary of the highest scoring clusters is given in Table 1[29–31]. The Venn diagram in Fig. 2 shows that more homologs were detected in the Synechocystis/Microcystis comparison than in any other pairwise combination, reflecting the phylogenetic relationships between these species pairs.
We previously showed the existence of Yfr1 in three out of four tested marine cyanobacteria belonging to the genera Prochlorococcus and Synechococcus  and later demonstrated its existence throughout the cyanobacterial radiation, including the four unicellular cyanobacteria targeted here . It was, therefore, no surprise to find Yfr1 among the top-scoring elements (Z-score and probability of -4.340 and 1.0) in cluster 139 (Table 1). Although RNA elements in cyanobacteria are only scarcely covered by Rfam, the existing entries provided another positive control set: the thiamine riboswitch was correctly identified in three strains (cluster 149; Table 1) and also two RNA elements of unknown function were correctly found for Synechocystis and Microcystis but not for the other two cyanobacteria (cluster 216 and 107 in Table 1). However, we noted the functional role assumed for one of these conserved RNA structures, the ykkC/yxkD element (cluster 216), to switch efflux pumps and detoxification systems in response to harmful environmental molecules, may not apply to the cyanobacterial homologs since they are neither in Synechocystis nor in Microcystis located upstream of a putative transporter gene.
Synteny among high-scoring RNA elements
The genomic location of a predicted ncRNA gene or RNA element in the same sequence neighbourhood in some or all of the studied cyanobacteria can also be a powerful tool for finding related ncRNAs. Among the 25 high-scoring sequence clusters in Table 1, 9 (36%) showed at least partial synteny. The high scoring element in cluster 80 illustrates this fact. The primary annotation gives no hint about the possible relatedness of the flanking genes. The flanking gene sufR annotated in Microcystis encodes an iron-sulfur cluster biosynthesis transcriptional regulator and similarity searches revealed that sll0088, syc2358d and sufR actually are orthologs of each other (Fig. 3A). Flanking the intergenic region with the predicted RNA element on the other side, genes ycf24 and sufB are clearly homologs of each other, whereas ftrC in Synechococcus is not. Yet, ftrC has been inserted in this genomic region as the proximate gene, syc2356_c, codes for the homolog of sufB and ycf24. Thus, the synteny among neighbouring genes clearly support the element predicted in cluster 80 as an orthologous RNA element between the three species. Other cases of partial synteny in flanking genes are observed in cluster 139 since trxA is present in 3 out of 4 cases and in cluster 216 with the orthologs speB (Synechocystis) and agmatinase (Microcystis), whereas all other genes are different. Special cases of synteny are exposed in cluster 207 (rpl10 leader), 149 (thiamine riboswitch upstream of thiC), 394 (rps2 leader) and 62 (upstream groES). These four examples represent structurally conserved sequence elements upstream of a protein-coding gene to whom they are functionally connected; among them one riboswitch and two ribosomal leaders, thus this position must be conserved. The fourth example, the element upstream of groES contains the palindromic CIRCE element (Fig. 3B) thought to bind the heat-shock repressor protein HrcA . Here, we mapped the groES transcriptional start site to the first nt of the nine nt loop predicted by secondary structure analysis (Fig. 3B), confirming the previously determined start site . These examples illustrate the variety of elements that become identified by our approach.
For exemplary experimental verification of predicted ncRNA genes we chose two very different examples, one well-supported candidate with three members from cluster 159 (probability 0.933 and Z-score -2.00; Table 1) and one from cluster 294 (probability 1.0 and Z-score -2.64; Table 1). Northern hybridization of total RNA from Synechocystis using strand-specific RNA probes confirmed the existence of both ncRNAs (Fig. 4). Since we verified the existence of both ncRNAs experimentally, we decided to name these two ncRNAs SyR1 and SyR2, for Sy nechocystis ncR NA 1 and 2. SyR1 is a strongly accumulating ncRNA transcribed from a gene in the fabF – hoxH IGR in the forward direction as the preceding gene fabF (Fig. 4A). The syr1 gene corresponds with a length of ~130 nt to about two thirds of the fabF-hoxH intergenic spacer (length 206 nt). Judged by Northern hybridization, there was no evidence for a possible cotranscription with fabF. The element predicted with the CLID 294 is located 3' to a protein-coding gene, too, and is transcribed from the forward strand in Synechocystis 6803. SyR2 is an ~140 nt ncRNA transcribed from a gene in the apcC (ssr3383) – prmA (sll1909) IGR in the same forward direction as the preceding gene apcC. SyR2 is accumulated to rather high amounts, too, but these appeared lower than in case of SyR1 (Fig. 4B). The preceding apcC gene (ssr3383) encodes a short phycobilisome LC linker polypeptide and is the ultimate gene of a three-gene operon for phycobiliproteins. Cotranscription between this operon and SyR2 cannot be excluded unambiguously. However, a SyR2 transcript start was mapped within apcC, 49 nt before the end of the reading frame. This fact is less exotic than it seems. At the expected spacing six nt upstream, the transcript start is preceded by a regular TATA element (CAAAAT). Moreover, several examples indicate the location of ncRNA promoters within the protein-coding part of a gene: Transcription of the ssrS gene for 6S RNA in E. coli is initiated at two promoters, from these the distally located promoter P2 responds to σ70 and σS RNA polymerase holoenzymes and is located within the ygfE reading frame [35, 36]. An example from Synechocystis is provided with IsrR, the antisense RNA that is initiated from within the gene isiA, although from the reverse complementary strand .
In a more general sense these results demonstrate that, just judging from the prediction, both candidate ncRNA genes might have been expected to be 3'UTRs due to their close location to an mRNA 3'end. We did not investigate their origin from a specific promoter further as we did for the Yfr2a ncRNA (see below), but the results shown in Fig. 4, in particular the lack of a longer transcript signal in the respective agarose gel blots, plus specific RACE signals confirm unambiguously that they do accumulate as individual small transcripts and therefore constitute bona fide ncRNAs.
A family of ncRNAs that is widely conserved among cyanobacteria
The vast majority of the ~100 bacterial ncRNAs experimentally verified thus far have been identified in Escherichia coli  and a few other model proteobacteria and Pseudomonas species. Therefore it is not surprising that, with the exception of the four highly conserved ncRNAs 6S RNA, tmRNA, ffs and RnpB, ortholog genes for ncRNAs are known only among very closely related species such as between Salmonella sp. and Yersinia sp. . Here, with cluster 219 eight sequences were identified with high sequence and predicted secondary structure similarity to a family of ncRNAs initially found in marine Prochlorococcus . There are four such ncRNAs in Prochlorococcus MED4 which in the original publication had been named Yfr2, Yfr3, Yfr4 and Yfr5 . From the eight new members to this family in cluster 219 three belong to Synechocystis, one to Thermosynechococcus and two each are predicted in Synechococcus and in Microcystis. Since none of them has a more pronounced similarity to any of the original Yfr2-Yfr5 ncRNAs from Prochlorococcus MED4, we decided to call them all "Yfr2" according to the first member in this group and then just to add a suffix. Therefore, the three predicted candidates belonging to this ncRNA family in Synechocystis are Yfr2a, Yfr2b and Yfr2c. All three are expressed in Synechocystis (Fig. 5). The yfr2a gene is located downstream of thioredoxin A (trxA) gene sll1980. Both genes are in the same orientation but Yfr2a originates from a specific initiation site of transcription, mapped by TAP-RACE to position 1558975 in the genome (complementary strand), 94 nt 3' of the sll1980 stop codon (Fig. 5). The other two ncRNAs belonging to this family in Synechocystis, Yfr2b and Yfr2c, originate from genes directly upstream of two protein-coding genes, slr0199 and sll1477. In these two cases we mapped identical initiation sites of transcription for the ncRNAs and their respective downstream located protein-coding gene, to genomic positions 2730523 (forward strand, Yfr2b) and 3398352 (complementary strand, Yfr2c). Therefore, transcription of these two mRNAs occurs possibly by a read-through mechanism from Yfr2b or Yfr2c. Whether this type of transcriptional fusion has functional relevance is currently unknown as are the functions of slr0199 and sll1477. These genomic arrangements are not conserved as the genes adjacent to the other five candidate ncRNA genes belonging to this family differ in the other three cyanobacteria. This lack of synteny supports a function of Yfr2b and Yfr2c independent from serving merely as 5' untranslated leaders of slr0199 and sll1477.
Sequence alignments and secondary structure predictions of the 8 Yfr2-5-type ncRNAs suggest a centrally located single-stranded loop element together with a short unpaired region at the 5' end that are highly conserved (Fig. 6). The long helical stem bearing the 12 nt loop is very characteristically predicted in all sequences to be interrupted by at least one bulge at position -4 with regard to this loop (Fig. 6). Interestingly, this feature is shared with the Yfr2-Yfr5 ncRNAs from marine cyanobacteria . Bulge motifs have been recognized in a wide range of RNAs as key structural elements determining molecular recognition by other molecules . Therefore, the conserved bulges in Yfr2-type ncRNAs may indicate their interaction with proteins. Indeed, another hint comes from the unpaired regions of these ncRNAs which resemble the extended "GGA" and "ANGGA" RsmA-binding motifs. The ncRNAs RsmX, RsmY and RsmZ found in Pseudomonas species contain several GGA and extended ANGGA motifs . For RsmY, these motifs have been shown to be essential for sequestration of RsmA and its homolog RsmE in Pseudomonas fluorescens . Non-coding RNAs containing this motif frequently have a titrating role on their target protein, regulating gene expression at the translational level. It was not possible, however, to identify RsmA and RsmE homologs in cyanobacteria.
Expression analysis of Yfr2a
Starting with the mapped initiation site of Yfr2a we chose the region located immediately upstream of it in a promoter fusion experiment with luxAB genes to prove that it actually does contain a functional promoter. Moreover, if the expression of an ncRNA is regulated under certain environmental conditions this sometimes gives a hint into which processes this ncRNA might be involved in. As controls, we chose the same DNA fragment in reverse orientation and amplified and cloned the psbA2 (slr1311) promoter, again in both orientations.
The 300 nt upstream of Yfr2a provided indeed very strong expression to the reporter genes – under all tested conditions the measured fluorescence values were comparable to those obtained from the psbA2 promoter-driven luciferase gene expression, whereas the reverse orientation of the same fragments provided very low activity only (Fig 7A). Under cold temperature (12°C) and heat (43°C) stress the activity of both promoters drops (Fig 7C, D), a possible pleiotropic effect. Under light stress, however, the activity of both promoters is stimulated. Whereas the activity of the psbA2 promoter is increased up to 250%, the Yfr2a promoter becomes activated up to 300% (Fig 7B). This is all the more striking since psbA2 frequently serves as an example for a typical light-inducible promoter.
Six Clusters containing repetitive sequences
One problem when dealing with genome sequences of some cyanobacteria is the high number of repetitive sequences, mobile genetic elements and transposon-related sequences.
Indeed, the output from our prediction pipeline was contaminated by imperfect inverted repeat sequences flanking different families of IS elements, mainly in Microcystis, and to a lesser extent in Synechocystis. If subtracted from the data set, the total number of predicted RNA elements (numbers in brackets correspond to high-scoring elements) in Synechocystis, Synechococcus, Thermosynechococcus and Microcystis drops to 339 (78), 160 (53), 144 (42) and 426 (168). Such repetitive sequences were collected in six sequence clusters, namely #12, #25, #38, #68, #270 and #383 with 78, 87, 153, 83, 42 and 16 individual sequences, respectively. All 459 sequences from these clusters have been collected in a separate file and are accessible from our website .
Here we provide the first list of ncRNA and other RNA element candidates for model unicellular cyanobacteria. Surprisingly, we identified with Yfr2a-Yfr2c a family of ncRNAs which is widely conserved among cyanobacteria and which become accumulated to high concentrations.
Our experimental verification together with existing positive controls suggests a high number of positives in this candidate set. However, there are also putative 5' operon leaders, Rho-independent 3' transcriptional terminators and possibly yet unidentified riboswitches in this data set. Moreover, the output is contaminated to some extent by transposase-related sequences.
Nevertheless, by analogy to other bacteria, including the most streamlined marine cyanobacterium Prochlorococcus MED4 , this number of ncRNAs and other RNA elements is probably a grave underestimation. Therefore this analysis should be considered as a first step to become complemented by more exhaustive experimental screens, for instance by using tiling arrays or deep sequencing in the near future.
Cultures and manipulation of cyanobacteria
The Synechocystis Moscow strain was used in this study (obtained from A. Wilde, University of Giessen, Germany, originally from S. Shestakov, Moscow State University, Russia) and propagated on BG11  1% (w/v) agar (Bacto agar, Difco) plates. Liquid cultures of Synechocystis were grown at 30°C in BG11 medium under continuous illumination with white light of 50 μmol of photons•m-2•s-1 and a continuous stream of air.
As a promoter test vector we used the pILA plasmid  into which ~300 bp long promoter fragments were cloned as transcriptional fusions with the luxAB genes. After transformation, this plasmid integrates into the slr0168 gene within the chromosome of Synechocystis by homologous recombination. Transformation and analysis of correct integration and segregation was carried out as described elsewhere .
Extraction and analysis of RNA
Exponentially growing Synechocystis cultures (OD750 0,6 – 0,8) were collected by filtration (Pall Supor 800 Filter). Filters with cells were dissolved in 1 ml Trizol per 40 ml culture, immediately frozen in liquid nitrogen and incubated for 15 min at 65°C in a water bath. Further RNA isolation followed the manufacturer's protocol.
Small RNA Northern blots were prepared from the separation of 10 to 25 μg of total RNA on 10% urea-polyacrylamide gels as described by Steglich et al. . Polyacrylamide gels were stained with ethidium bromide (0.3 μg/l) in 1× TBE buffer, rinsed with 1× TBE and analyzed with an E-BOX video gel documentation system (Peqlab). Transcript sizes were determined by correlation to Fermentas' RiboRuler low range RNA marker. Blots for RNAs with higher molecular weight were prepared from the separation of 5 μg of total RNA on 1,5% denaturing agarose gels. Transcriptional start sites were determined by 5'-RACE as described .
Genome sequences were obtained from the finished microbial genomes website at Genbank  with the following accession numbers: Synechococcus elongatus PCC 6301, NC_006576; Synechocystis sp. PCC 6803, NC_000911; Thermosynechococcus elongatus BP-1, NC_004113; Microcystis aeruginosa NIES843, NC_010296.
Prediction of RNA elements in a comparative approach
We performed a comparative prediction of ncRNA elements within intergenic regions (IGRs). Therefore, all IGRs longer than 50 nt were extracted and compared among the different genomes using Blast. Intragenomic analyses with the settings given in Fig. 1 revealed a high number of repetitive sequences in some of the analyzed genomes. There were 895 intragenomic hits within the genome of Synechococcus, 2227 hits within Thermosynechococcus, 7198 in Synechocystis and 557014 mutual hits of IGRs from Microcystis with other IGRs from the same organism. A large number of repetitive elements N would produce approximately N2 hits, meaning that the square root of the number of hits gives an estimate of the number of repetitive elements. In the case of Microcystis this yields ~746 repetitive elements, which is supported by the finding of six large clusters holding 383 sequences from Microcystis. Therefore, we refrained from searching for intragenomic similarity. Based on these results homologous sequence regions got clustered together, aligned using ClustalW and analyzed for structural significance by RNAz. Alignments were postprocessed using the tool rnazSelectSeqs.pl with default parameters from the RNAz package. RNAz was applied in a sliding window approach (a step size of 10 nt and different window sizes were used, namely 80,100,120,140 and 160 nt) of which the window with highest probability was selected. Elements are termed "High-scoring" if they achieve an RNAz probability of 0.5 or more or if their Z-score is -2.0 or below. Details about the individual steps and their outcome are shown in Fig. 1. All predictions can be found at  and .
Matching predictions to Rfam and TransTermHP
All individual sequences were matched against Rfam  using the batch search feature provided by Rfam. Mapping of predicted sequences to information about Rho-independent terminators provided by TransTermHP  was done using Vmatch . Therefore, TransTermHP predictions for Microcystis were computed using TransTermHP 2.06 with default parameters, while for Synechocystis, Synechococcus and Thermosynechococcus existing predictions were downloaded from the TransTermHP website. All predictions were converted to FASTA-format and searched for at least 30 nt long hits with 100% identity to candidate sequences.
Comparative sequence/structure analysis
Multiple sequence alignments were generated using ClustalW  with default parameters for DNA. Comparative structure prediction was done with RNAlishapes , a tool which predicts a consensus structure for a set of aligned sequences by taking covariance and free energy into account. The resulting consensus structure was analysed together with the multiple sequence alignment using RALEE . The latter served also for manual optimisation of the alignment and the consensus structure, respectively, and for the production of colour annotated alignments. Colour plots of Consensus structures were generated using RNAalifold .
Oligonucleotide primers for the generation of hybridization probes (T7 promoter sequence in boldface letters):
Yfr2c_for: 5'-TAATACGACTCACTATAGG ccgccagcgccattgcttc-3'
Yfr2b_for: 5'-TAATACGACTCACTATAGG cggggagcatagaccagcttg-3'
Yfr2a_for: 5'-TAATACGACTCACTATAGG aggcaaaaaaataaggaagtccgcaag-3'
Syr1_for: 5'-TAATACGACTCACTATAGG ccgagggcatatctaggagaac-3'
Syr2_for: 5'-TAATACGACTCACTATAGG caaacaaaaaaagaggccattgctgacc-3'
Oligonucleotides for promoter fusion experiments (Kpn I restriction sites 5'-GGTACC-3' introduced for cloning are labelled in boldface letters):
Yfr2a (fw): 5'-GGTACC CTAGATGACACCGGCACG-3'
Yfr2a (rev): 5'-GGTACC CTCCTCACACACAAATAAATGTTAG-3'
psbA2 (fw): 5'-CCTTGGTACC AAGAGTAATGGCGTGC-3'
psbA2 (rev): 5'-GATTGGTACC GGAACTGACTAAACTTAGTC-3'
Oligonucleotides for specific mapping of 5'ends through 5'RACE:
Northern hybridization was performed at 62°C in hybridization buffer (50% deionized formamide, 7% SDS, 250 mM NaCl, 120 mM Na(PO4), pH 7.2) as described by Steglich et al. . Single stranded probes were generated from PCR-amplified templates incorporating the T7 promoter in one of the oligonucleotide primers, using the MAXIscript Kit (Ambion, USA) and 100 ng PCR-generated DNA template.
Vogel J, Papenfort K: Small non-coding RNAs and the bacterial outer membrane. Curr Opin Microbiol. 2006, 9 (6): 605-611.
Gottesman S: Micros for microbes: non-coding regulatory RNAs in bacteria. Trends Genet. 2005, 21: 399-404.
Wagner EG, Simons RW: Antisense RNA control in bacteria, phages, and plasmids. Ann Rev Microbiol. 1994, 48: 713-742.
Sittka A, Pfeiffer V, Tedin K, Vogel J: The RNA chaperone Hfq is essential for the virulence of Salmonella typhimurium. Mol Microbiol. 2007, 63 (1): 193-217.
Lenz DH, Mok KC, Lilley BN, Kulkarni RV, Wingreen NS, Bassler BL: The small RNA chaperone Hfq and multiple small RNAs control quorum sensing in Vibrio harveyi and Vibrio cholerae. Cell. 2004, 118: 69-82.
Brouns SJ, Jore MM, Lundgren M, Westra ER, Slijkhuis RJ, Snijders AP, Dickman MJ, Makarova KS, Koonin EV, Oost van der J: Small CRISPR RNAs guide antiviral defense in prokaryotes. Science. 2008, 321 (5891): 960-964.
Vogel J, Sharma CM: How to find small non-coding RNAs in bacteria. Biol Chem. 2005, 386 (12): 1219-1238.
Axmann IM, Kensche P, Vogel J, Kohl S, Herzel H, Hess WR: Identification of cyanobacterial non-coding RNAs by comparative genome analysis. Genome Biol. 2005, 6: R73-
Steglich C, Futschik ME, Lindell D, Voss B, Chisholm SW, Hess WR: The challenge of regulation in a minimal phototroph: Non-coding RNAs in Prochlorococcus. PLoS Genet. 2008, 4 (8): e1000173-
Dufresne A, Ostrowski M, Scanlan DJ, Garczarek L, Mazard S, Palenik BP, Paulsen IT, Tandeau de Marsac N, Wincker P, Dossat C, et al: Unravelling the genomic mosaic of a ubiquitous genus of marine cyanobacteria. Genome Biol. 2008, 9: R90-
Kettler GC, Martiny AC, Huang K, Zucker J, Coleman ML, Rodrigue S, Chen F, Lapidus A, Ferriera S, Johnson J, et al: Patterns and implications of gene gain and loss in the evolution of Prochlorococcus. PLoS Genet. 2007, 3 (12): e231-
Tolonen AC, Liszt GB, Hess WR: Genetic manipulation of Prochlorococcus strain MIT9313: Green Fluorescent Protein expression from an RSF1010 plasmid and Tn5 transposition. Appl Environ Microbiol. 2006, 72: 7607-7613.
Voss B, Gierga G, Axmann IM, Hess WR: A motif-based search identifies the ortholog of the small RNA Yfr1 in all major lineages of cyanobacteria. BMC Genomics. 2007, 8: 375-
Nakamura T, Naito K, Yokota N, Sugita C, Sugita M: A cyanobacterial non-coding RNA, Yfr1, is required for growth under multiple stress conditions. Plant Cell Physiol. 2007, 48: 1309-1318.
Axmann IM, Holtzendorff J, Voss B, Kensche P, Hess WR: Two distinct types of 6S RNA in Prochlorococcus. Gene. 2007, 406: 69-78.
Barrick JE, Sudarsan N, Weinberg Z, Ruzzo WL, Breaker RR: 6S RNA is a widespread regulator of eubacterial RNA polymerase that resembles an open promoter. RNA. 2005, 11 (5): 774-784.
Hernandez JA, Muro-Pastor AM, Flores E, Bes MT, Peleato ML, Fillat MF: Identification of a furA cis antisense RNA in the cyanobacterium Anabaena sp. PCC 7120. J Mol Biol. 2005, 355: 325-334.
Duehring U, Axmann IM, Hess WR, Wilde A: An internal antisense RNA regulates expression of the photosynthesis gene isiA. Proc Nat Acad Sci USA. 2006, 103: 7054-7058.
del Val C, Rivas E, Torres-Quesada O, Toro N, Jiménez-Zurdo JI: Identification of differentially expressed small non-coding RNAs in the legume endosymbiont Sinorhizobium meliloti by comparative genomics. Mol Microbiol. 2007, 66 (5): 1080-1091.
González N, Heeb S, Valverde C, Kay E, Reimmann C, Junier T, Haas D: Genome-wide search reveals a novel GacA-regulated small RNA in Pseudomonas species. BMC Genomics. 2008, 9: 167-
Mandin P, Repoila F, Vergassola M, Geissmann T, Cossar P: Identification of new noncoding RNAs in Listeria monocytogenes and prediction of mRNA targets. Nucleic Acids Res. 2007, 35: 962-974.
Padalon-Brauch G, Hershberg R, Elgrably-Weiss M, Baruch K, Rosenshine I, Margalit H, Altuvia S: Small RNAs encoded within genetic islands of Salmonella typhimurium show host-induced expression and role in virulence. Nucleic Acids Res. 2008, 36 (6): 1913-1927.
Ulvé VM, Sevin EW, Chéron A, Barloy-Hubler F: Identification of chromosomal alpha-proteobacterial small RNAs by comparative genome analysis and detection in Sinorhizobium meliloti strain 1021. BMC Genomics. 2007, 8: 467-
Kaneko T, Nakajima N, Okamoto S, Suzuki I, Tanabe Y, Tamaoki M, Nakamura Y, Kasai F, Watanabe A, Kawashima K, et al: Complete genomic structure of the bloom-forming toxic cyanobacterium Microcystis aeruginosa NIES-843. DNA Res. 2007, 14 (6): 247-256.
Prediction of ncRNAs in model cyanobacteria website. [http://www.cyanolab.de/paper/suppl/syncomp]
Washietl S, Hofacker IL, Stadler PF: Fast and reliable prediction of noncoding RNAs. Proc Natl Acad Sci USA. 2005, 102: 2454-2459.
Griffiths-Jones S, Moxon S, Marshall M, Khanna A, Eddy SR, Bateman A: Rfam: annotating non-coding RNAs in complete genomes. Nucleic Acids Res. 2005, 33: D121-124.
Kingsford CL, Ayanbule K, Salzberg SL: Rapid, accurate, computational discovery of Rho-independent transcription terminators illuminates their relationship to DNA uptake. Genome Biol. 2007, 8: R22-
Barrick JE, Corbino KA, Winkler WC, Nahvi A, Mandal M, Collins J, Lee M, Roth A, Sudarsan N, Jona I, et al: New RNA motifs suggest an expanded scope for riboswitches in bacterial genetic control. Proc Natl Acad Sci USA. 2004, 101: 6421-6426.
Rodionov DA, Vitreschak AG, Mironov AA, Gelfand MS: Comparative genomics of thiamin biosynthesis in procaryotes. New genes and regulatory mechanisms. J Biol Chem. 2002, 277: 48949-48959.
Zengel JM, Lindahl L: Diverse mechanisms for regulating ribosomal protein synthesis in Escherichia coli. Prog Nucleic Acid Res Mol Biol. 1994, 47: 331-370.
Oliveros JC: VENNY. An interactive tool for comparing lists with Venn diagrams. [http://bioinfogp.cnb.csic.es/tools/venny/index.html]
Nakamoto H, Suzuki M, Kojima K: Targeted inactivation of the hrcA repressor gene in cyanobacteria. FEBS Lett. 2003, 549: 57-62.
Glatz A, Horvath I, Varvasovszki V, Kovacs E, Torok Z, Vigh L: Chaperonin genes of the Synechocystis PCC 6803 are differentially regulated under light-dark transition during heat stress. Biochem Biophys Res Commun. 1997, 239: 291-297.
Hsu LM, Zagorski J, Wang Z, Fournier MJ: Escherichia coli 6S RNA gene is part of a dual-function transcription unit. J Bacteriol. 1985, 161: 1162-1170.
Kim KS, Lee Y: Regulation of 6S RNA biogenesis by switching utilization of both s factors and endoribonucleases. Nucleic Acids Res. 2004, 32: 6057-6068.
Hershberg R, Altuvia S, Margalit H: A survey of small RNA-encoding genes in Escherichia coli. Nucleic Acids Res. 2003, 31: 1813-1820.
Hermann T, Patel DJ: RNA bulges as architectural and recognition motifs. Structure. 2000, 8: R47-R54.
Babitzke P, Romeo T: CsrB sRNA family: sequestration of RNA-binding regulatory proteins. Curr Opin Microbiol. 2007, 10: 156-163.
Valverde C, Lindell M, Wagner EG, Haas D: A repeated GGA motif is critical for the activity and stability of the riboregulator RsmY of Pseudomonas fluorescens. J Biol Chem. 2004, 279: 25066-25074.
Overview on repeat elements in model cyanobacteria. [http://www.cyanolab.de/paper/suppl/syncomp/repeat_cluster.html]
Rippka R, Deruelles J, B WJ, Herdmann M, Stanier RY: Generic assignments, strain histories and properties of pure cultures of cyanobacteria. J Gen Microbiol. 1979, 111: 1-61.
Kunert A, Hagemann M, Erdmann N: Construction of promoter probe vectors for Synechocystis sp. PCC 6803 using the light-emitting reporter systems Gfp and LuxAB. Journal of Microbiological Methods. 2000, 41: 185-194.
Dienst D, Dühring U, Mollenkopf H, Vogel J, Golecki J, Hess WR, Wilde A: The RNA chaperone Hfq is essential for cell motility of the cyanobacterium Synechocystis PCC 6803. Microbiology. 2008, 154: 3134-3143.
Kurtz S: The Vmatch large scale sequence analysis software. [http://www.vmatch.de]
Thompson JD, Higgins DG, Gibson TJ: CLUSTAL W: improving the sensitivity of progressive multiple sequence alignment through sequence weighting, positions-specific gap penalties and weight matrix choice. Nucleic Acids Res. 1994, 22: 4673-4680.
Voss B: Structural analysis of aligned RNAs. Nucleic Acids Res. 2006, 34: 5471-5481.
Griffiths-Jones S: RALEE – RNA ALignment editor in Emacs. Bioinformatics. 2005, 21: 257-259.
Hofacker IL, Fekete M, Stadler PF: Secondary structure prediction for aligned RNA sequences. J Mol Biol. 2002, 319 (5): 1059-1066.
This work was supported by the Deutsche Forschungsgemeinschaft Focus program "Sensory and regulatory RNAs in Prokaryotes" SPP1258 (project HE 2544/4-1), the graduate school "Signal systems in plant model organisms" (to JG) and by the Freiburg Initiative in Systems Biology. We thank Martin Hagemann for the gift of promoter test vectors.
BV designed and carried out bioinformatic analyses and participated in drafting the manuscript, JG performed all RNA analyses in the laboratory, SU and VS constructed the promoter test constructs and performed the reporter gene assays. WRH designed research and wrote the manuscript. All authors read and approved the final manuscript.