• We are sorry, but NCBI web applications do not support your browser and may not function properly. More information
Logo of pnasPNASInfo for AuthorsSubscriptionsAboutThis Article
Proc Natl Acad Sci U S A. May 2, 2006; 103(18): 7054–7058.
Published online Apr 24, 2006. doi:  10.1073/pnas.0600927103
PMCID: PMC1459017
From the Cover

An internal antisense RNA regulates expression of the photosynthesis gene isiA


Small regulatory noncoding RNAs exist in both eukaryotic and prokaryotic organisms. Most of these RNA transcripts are trans-encoded RNAs with short and only partial antisense complementarity to their target RNAs, which regulate gene expression by modifying mRNA stability and translation. In contrast, reports on the function of cis-encoded, perfectly complementary antisense RNAs in eubacteria are rare. Cyanobacteria respond to iron deficiency by expressing IsiA (iron stress-induced protein A), which forms a giant ring structure around photosystem I. Here, we show that this process is controlled by IsrR (iron stress-repressed RNA), a cis-encoded antisense RNA transcribed from the isiA noncoding strand. Artificial overexpression of IsrR under iron stress causes a strongly diminished number of IsiA–photosystem I supercomplexes, whereas IsrR depletion results in premature expression of IsiA. The coupled degradation of IsrR/isiA mRNA duplexes appears to be a reversible switch that can respond to environmental changes. IsrR is the only RNA known so far to regulate a photosynthesis component.

Keywords: cyanobacteria, iron stress, redox stress, regulation of gene expression, light harvesting

Biological carbon assimilation and oxygen production through photosynthesis depend significantly on phytoplankton such as cyanobacteria. In aquatic ecosystems, particularly in the oceans, iron is often the limiting factor for cyanobacterial growth and photosynthetic activity (1). Cyanobacteria respond to iron deficiency by expressing iron stress-induced protein A (IsiA) (2, 3). In sequence and structure, IsiA is closely related to the photosystem (PS) II antenna protein CP43 and therefore also is named CP43′. IsiA forms a giant additional antenna ring around PSI, thus enhancing light absorption and helping to compensate for a reduction in the number of PSI trimeric complexes that occurs under iron limitation (35). Activation of its expression under iron limitation is mediated by the ferric uptake regulator Fur (6). IsiA also functions to dissipate excess light energy under high light and oxidative stress (79), and thus the precise cellular role of IsiA and the regulation of its gene expression under these conditions have remained enigmatic (6, 1013). Here, we report the identification of a unique regulatory mechanism acting at the posttranscriptional level by targeted degradation of duplexes formed between the isiA mRNA and its antisense regulator, IsrR (iron stress-repressed RNA).


Studying gene expression of isiA in the cyanobacterium Synechocystis sp. PCC 6803 (hereafter called Synechocystis 6803) by Northern blot hybridization, we detected a small RNA whose accumulation is inversely regulated to the mRNA level of its host gene isiA (Fig. 1). The length of IsrR was determined by Northern blots and RACE experiments at 177 nt, and its 5′ end was mapped to nucleotide 1.518.034 in the totally sequenced genome of this organism. Thus, IsrR extends from position 395 to position 571 with regard to the coding sequence of isiA (Fig. 1A). Strand-specific deoxyoligonucleotide probes identified this RNA as an antisense RNA produced from the complementary strand of isiA. We named this antisense RNA IsrR, because it is present in high amounts under iron-replete conditions but not under iron limitation. In contrast, isiA mRNA was not detectable under iron-replete conditions but became detectable after 17 h in iron-depleted media (Fig. 1B). The isiA gene is known to respond to high-light-induced stress (10, 11). Thus, we transferred cells from medium light (40 μmol of photons·m−2·s−1) to high light (170 μmol of photons·m−2·s−1) for 2–41 h and examined isiA expression. During this time course, we detected a rise in the level of isiAB dicistronic precursor transcript and mature isiA mRNA from initially undetectable amounts to very high amounts, whereas the accumulation of IsrR decreased in an exactly inverse fashion (Fig. 1C). Another short transcript from this region was previously suggested to represent a 5′ UTR transcript (6). Our data confirm the existence and orientation of such a transcript of ≈160 nt, which appears to coaccumulate with isiA mRNA to some extent.

Fig. 1.
Characterization of an antisense RNA in Synechocystis 6803. (A) Location of the isiA gene within the genome. The downstream gene isiB encodes flavodoxin; the distance between the transcription initiation site (red arrow) and translation start site indicates ...

Thus, at least four different transcripts originate in this region: the unprocessed isiAB precursor transcript, the isiA monocistronic transcript, IsrR with maximum accumulation under iron-replete and low-light conditions (when no isiA mRNA was detected), and the 5′ UTR transcript that was observed previously (6) but was not further analyzed.

IsrR can be folded into two extended stem regions, each finishing with a terminal loop (Fig. 1D). Such loop structures are frequently involved in RNA–RNA interactions and thus can be functionally relevant for a hypothetical trans-acting function. However, a direct effect on the isiA host gene’s mRNA level appeared much more likely.

To test this hypothesis, a Synechocystis WT strain was supplied with either one of three self-replicating plasmids so that normal, high, and very low intracellular levels of IsrR could be generated. To get high levels of IsrR, we transferred a plasmid containing the isrR sequence under control of the strong RNase P (rnpB) promoter into Synechocystis 6803. To deplete IsrR without introducing changes in the isiA coding sequence, we used a strategy similar to the one used by Udekwu et al. (14) to eliminate MicA RNA in Escherichia coli. The isrR sequence was cloned in reverse complementary orientation under control of the rnpB promoter and the oop terminator from bacteriophage lambda as a signal for termination of transcription, in essence expressing an anti-IsrR RNA. The third plasmid served as a control and carried no additional insertion. Northern blot hybridizations verified successful overexpression or down-regulation of the IsrR RNA in mutant cells of Synechocystis 6803 (Fig. 2). These effects could best be studied in cells treated with H2O2 because here the induction of isiA on mRNA level is much faster than the response to high-light stress or iron limitation. All these conditions are known to induce oxidative stress. Synechocystis WT cells treated with H2O2 responded within 1 h with the accumulation of isiA mRNA (Fig. 2). The rnpB-driven IsrR overexpression resulted in a significantly reduced level of isiA mRNA, which appeared at a later point in the time course. In the opposite experiment, expression of the anti-IsrR RNA led to a diminished amount of the isrR transcript and a substantially increased level of isiA mRNA under oxidative stress conditions (Fig. 2). Furthermore, anti-IsrR mutant cells responded very rapidly to the oxidative stress; the isiA mRNA became detectable only 0.3 h after the application of H2O2. In a series of replicate experiments, including a set of independently created mutant lines, similar results were obtained.

Fig. 2.
Impact of modulating the amount of IsrR on the accumulation of isiA mRNA under oxidative stress. Northern blots showing the effects of normal (WT), high (Mu:IsrR[+]), and very low (Mu:IsrR[−]) expression of isrR on the level of isiA mRNA in mutant ...

Intriguingly, comparable results were obtained when strains expressing these constructs were subjected to iron limitation (Fig. 3). We checked whether the modulation of IsrR amount would have an impact on the concentration of the IsiA protein and the formation of PSI–IsiA supercomplexes under iron stress. In comparison with WT control, amounts of isiA mRNA and protein were significantly lower in cells that overexpress IsrR, whereas both were enhanced in IsrR knocked-down cells after 2 days under iron-depleted conditions (Fig. 3 A and B). When native complexes were examined from cell membranes by using blue-native PAGE (Fig. 3C), we observed that starvation for iron induced the formation of PSI–IsiA supercomplexes at the expense of trimeric PSI complexes in WT cells. In contrast, overexpression of IsrR interfered with the formation of PSI–IsiA supercomplexes severely, whereas the IsrR knock-down experiment resulted in an even further reduced level of trimeric PSI complexes in favor of high-molecular-weight PSI–IsiA supercomplexes. The effect of IsrR knock-down was also detectable in absorption spectra of whole cells, indicated by the 6-nm shift of the red chlorophyll absorption peak toward the blue part of the spectrum compared with untreated WT cells (Fig. 3D). A blue shift is characteristic of IsiA-accumulating cells; hence, it was also observed for WT cells grown under iron-depleted conditions. In that case, however, we observed a less prominent blue shift of only 3 nm. The amounts of both dimeric and monomeric PSII complexes were consistently found to be diminished to a different degree in thylakoid proteins isolated from the WT −Fe and the two mutants. This finding is in agreement with a previously observed reduction in the total amount of PSII under iron limitation (15).

Fig. 3.
Influence of IsrR on the accumulation of the IsiA protein, trimeric PSI, and PSI–IsiA supercomplexes. (A) Northern blot showing the expression of isiA in cells with normal (wild type), high (Mu:IsrR[+]), and very low (Mu:IsrR[−]) levels ...


Short noncoding RNAs (ncRNAs) are increasingly being recognized as being crucial for the regulatory network of all organisms. In enterobacteria, trans-encoded RNAs with short and only partial antisense complementarity to their target RNAs have been shown to be involved in all major stress responses (16). The results shown here provide strong evidence for the cis-encoded antisense RNA IsrR acting on the expression of isiA. We show an inverse relationship between the presence of IsrR and isiA mRNA under three different stress conditions: iron limitation, high light, and oxidative stress. The tuning of isiA expression by IsrR might have tremendous physiological relevance because isiA in Synechocystis 6803 is the single most strongly expressed gene upon treatment by 1.5 mM H2O2 (12) and becomes strongly activated also under salt and heat stress.

At the level of the isiA transcript, apparently the response to iron limitation is much slower than to treatment with hydrogen peroxide. This repression of isiA transcription under iron-replete conditions has previously been suggested to be under control of the ferric uptake regulator Fur (6). However, the different slopes of isiA derepression point to the possible involvement of additional factors in addition to Fur and the posttranscriptional control through IsrR described in this report.

IsrR is the only antisense RNA identified so far that regulates a component involved in photosynthesis. The known cases of trans-encoded regulatory RNAs in bacteria mostly interfere with the initiation of translation, either through competition with 16S rRNA for the Shine–Dalgarno sequence or by selectively exposing or masking crucial regulatory regions in the mRNA 5′ UTRs (16). However, for the vast majority of bacterial regulatory RNAs, the targets have yet to be identified. In contrast, IsrR has a well defined target in its host gene’s mRNA. It is, over its full length, complementary to the central third of the isiA mRNA, which does not include possible interactions with the Shine–Dalgarno sequence and making an interference with initiation of translation at the ribosome very unlikely. Rather, the transient formation of IsrR–isiA RNA duplexes makes them a target for selective degradation. The rapid disappearance/reappearance of IsrR versus its target mRNA under transient oxidative stress (Fig. 2) indicates their coupled degradation. An antisense RNA that is degraded together with its target makes a perfect reversible switch to respond to environmental changes. Indeed, in the case of trans-encoded ncRNAs with imperfect complementarity, a growing number of examples indicate that they can target specific mRNAs for degradation by RNase III; examples include the IstR RNAs in E. coli (17) and the spa mRNA in Staphylococcus aureus (18). Thus, the formation of ncRNA–protein complexes containing RNase III and, in some cases, RNase E (19) could be a general way by which mRNAs are destabilized by small RNAs in bacteria in a manner similar to that in higher organisms (19).

The activation of isiA under iron limitation has been demonstrated to be primarily under control of the ferric uptake regulator Fur (6). This level of control functions independently from IsrR; we did not detect isiA expression under iron-replete conditions in the IsrR knock-down strain. However, once the transcription of isiA becomes activated under the conditions tested here, IsrR controls a premature expression of isiA at the posttranscriptional level, likely by targeting isiA mRNA for degradation. This control appears to be effective as long as the isiA mRNA is present in substoichiometric levels. If the transcriptional activation of isiA continues, its concentration finally starts to exceed the amount of IsrR, and the mRNA becomes accumulated in a significant amount. It is very possible that the transcription of IsrR continues under these conditions, indicated by the trace amounts of IsrR visible in Fig. 1 B and C after 24 h and later, but that its turnover is drastically enhanced as a consequence of RNA:RNA degradation. However, at present it cannot be excluded that the expression of IsrR is under transcriptional control as well.

We currently do not know the functional relevance of the 5′ UTR transcript, but it is likely to result from the attenuation of transcription. If so, the expression of isiA would be controlled at the RNA level by at least three different mechanisms: transcriptional control mediated by Fur, attenuation, and targeted degradation by the antisense RNA IsrR.

Why are there so few cis-acting antisense regulators known in the bacterial chromosome? Most methods used so far in the identification of regulatory RNAs in bacteria (2022), including cyanobacteria (23), were focused on intergenic regions. Thus, there could have been a methodological problem in finding small cis-encoded antisense RNAs such as IsrR.

Alternatively, the vast majority of data on bacterial riboregulators stem from the analysis of enterobacteria, whereas a single systematic search exists for noncoding RNAs in cyanobacteria (23). It is very likely that cyanobacteria, which grow under extremely diverse environmental conditions and have an evolutionary history of >3 billion years, have developed a complex RNA-based regulatory mechanism to deal with changes in their environment.

Materials and Methods

Bacterial Strains and Growth Conditions.

Liquid cultures of Synechocystis 6803 WT and mutant strains were grown at 30°C in BG11 medium under continuous illumination with white light of 50 μmol of photons·m−2·s−1 (or 170 μmol of photons·m−2·s−1 for high-light conditions) and a continuous stream of air. The medium for the mutant strains was supplemented with 60 μg/ml kanamycin.


The plasmid used for overexpression of IsrR was constructed by fusing the rnpB promoter fragment and the isrR gene in a two-step PCR process using primers 5′rnpB, 3′rnpB/isrR, 5′isrR/rnpB, and 3′isrR (for sequences, see Table 1). The plasmid expressing the IsrR complement anti-IsrR was constructed by fusion of the rnpB promoter fragment to the complementary isrR sequence and the oop terminator from bacteriophage lambda by using PCR products amplified with the 5-rnpB, 3′rnpB/anti-isrR, and 5′anti-isrR/rnpB primers and the 3′anti-isrR/oop terminator primer. The fragments were subcloned into the pDrive vector (Qiagen, Valencia, CA) and then excised by PstI/SalI digestion and inserted into the chloramphenicol resistance gene cartridge of the pVZ321 vector (24). The pVZ321 vector harboring the different fragments was transferred to WT cells by conjugation (24), and exconjugants were selected on BG11 agar plates containing 60 μg/ml kanamycin.

Table 1.

Immunoblot Analyses.

Proteins were separated by SDS/PAGE and characterized as described in ref. 25. After electrophoresis, proteins were electrophoretically transferred onto nitrocellulose membranes and immunodecorated with specific antibodies. Immunolabeled bands were visualized with a goat anti-rabbit IgG-peroxidase conjugate and SuperSignal West Pico as chemiluminescent substrate (Pierce). A polyclonal antiserum against IsiA was generated by coupling the synthetic peptide APADLSTLKGQGKKFHFTWENPQQ (position 114–137) by means of an additional N-terminal cysteine residue to keyhole limpet hemocyanin as a carrier protein and injecting it into rabbits.

Blue-Native PAGE.

Thylakoid membranes were isolated as described in ref. 25 and resuspended in 40 μl of ACA buffer (50 mM Bis-Tris·HCl, pH 7.0/750 mM ε-amino-n-caproic acid/0.5 mM EDTA). Membrane proteins were solubilized by the addition of 5 μl of 10% (wt/vol) n-dodecyl-β-d-maltoside solution. After centrifugation (15,000 × g for 20 min), the supernatants were supplemented with 8 μl of a Coomassie blue solution [5% (wt/vol) Serva Blue G/750 mM ε-amino-n-caproic acid] and loaded onto a gel. One-dimensional blue-native PAGE was carried out as described in ref. 25.

Spectral Measurements.

Absorbance spectra of whole cells were measured with a UV-2401 PC spectrophotometer (Shimadzu).

RNA Analysis.

Total RNA was isolated with TRIzol reagent (Invitrogen) and separated on 10% polyacrylamide-urea or 1.3% agarose formaldehyde gels. Northern blots and hybridizations were performed as described in ref. 23. Verification of equal loading was achieved by measurement of RNA concentrations, direct comparison of rRNA band intensities after staining by ethidium bromide, or hybridization with a 16S rRNA probe. Mapping of RNA 5′ and 3′ ends was performed by rapid amplification of cDNA ends as described in ref. 20. All DNA and RNA oligonucleotides are listed in Table 1. Signals were detected and analyzed on a Personal Molecular Imager FX system with quantity one software (Bio-Rad).


We thank M. Hagemann (Rostock, Germany) for the isiAB mutant and J. Vogel, D. Scanlan, S. Porter, and M. Hagemann for comments. This work was supported by grants from Fonds der Chemischen Industrie and Deutsche Forschungsgemeinschaft (to A.W. and W.R.H.), the European Union (MARGENES project QLRT-2001-01226) (to W.R.H.), and Landesstiftung Baden-Wuerttemberg (to W.R.H.).


iron stress-induced protein A
iron stress-repressed RNA

Conflict of interest statement: No conflicts declared.

This paper was submitted directly (Track II) to the PNAS office.

See Commentary on page 6781.


1. Behrenfeld M. J., Bale A. J., Zbigniew S. K., Aiken J., Falkowski P. G. Nature. 1996;383:508–511.
2. Burnap R. L., Troyan T., Sherman L. A. Plant Physiol. 1993;103:893–902. [PMC free article] [PubMed]
3. Laudenbach D. E., Straus N. A. J. Bacteriol. 1988;170:5018–5026. [PMC free article] [PubMed]
4. Bibby T. S., Nield J., Barber J. Nature. 2001;412:743–745. [PubMed]
5. Boekema E. J., Hifney A., Yakushevska A. E., Piotrowski M., Keegstra W., Berry S., Michel K. P., Pistorius E. K., Kruip J. Nature. 2001;412:745–748. [PubMed]
6. Vinnemeier J., Kunert A., Hagemann M. FEMS Microbiol. Lett. 1998;169:323–330. [PubMed]
7. Cadoret J. C., Demouliere R., Lavaud J., van Gorkom H. J., Houmard J., Etienne A. L. Biochim. Biophys. Acta. 2004;1659:100–104. [PubMed]
8. Ihalainen J. A., D’Haene S., Yeremenko N., van Roon H., Arteni A. A., Boekema E. J., van Grondelle R., Matthijs H. C., Dekker J. P. Biochemistry. 2005;44:10846–10853. [PubMed]
9. Yeremenko N., Kouril R., Ihalainen J. A., D’Haene S., van Oosterwijk N., Andrizhiyevskaya E. G., Keegstra W., Dekker H. L., Hagemann M., Boekema E. J., et al. Biochemistry. 2004;43:10308–10313. [PubMed]
10. Havaux M., Guedeney G., Hagemann M., Yeremenko N., Matthijs H. C., Jeanjean R. FEBS Lett. 2005;579:2289–2293. [PubMed]
11. Jeanjean R., Zuther E., Yeremenko N., Havaux M., Matthijs H. C., Hagemann M. FEBS Lett. 2003;549:52–56. [PubMed]
12. Li H., Singh A. K., McIntyre L. M., Sherman L. A. J. Bacteriol. 2004;186:3331–3345. [PMC free article] [PubMed]
13. Yousef N., Pistorius E. K., Michel K. P. Arch. Microbiol. 2003;180:471–483. [PubMed]
14. Udekwu K. I., Darfeuille F., Vogel J., Reimegard J., Holmqvist E., Wagner E. G. Genes Dev. 2005;19:2355–2366. [PMC free article] [PubMed]
15. Park Y. I., Sandstrom S., Gustafsson P., Oquist G. Mol. Microbiol. 1999;32:123–129. [PubMed]
16. Gottesman S. Annu. Rev. Microbiol. 2004;58:303–328. [PubMed]
17. Vogel J., Argaman L., Wagner E. G., Altuvia S. Curr. Biol. 2004;14:2271–2276. [PubMed]
18. Huntzinger E., Boisset S., Saveanu C., Benito Y., Geissmann T., Namane A., Lina G., Etienne J., Ehresmann B., Ehresmann C., et al. EMBO J. 2005;24:824–835. [PMC free article] [PubMed]
19. Morita T., Maki K., Aiba H. Genes Dev. 2005;19:2176–2186. [PMC free article] [PubMed]
20. Argaman L., Hershberg R., Vogel J., Bejerano G., Wagner E. G., Margalit H., Altuvia S. Curr. Biol. 2001;11:941–950. [PubMed]
21. Rivas E., Klein R. J., Jones T. A., Eddy S. R. Curr. Biol. 2001;11:1369–1373. [PubMed]
22. Wassarman K. M., Repoila F., Rosenow C., Storz G., Gottesman S. Genes Dev. 2001;15:1637–1651. [PMC free article] [PubMed]
23. Axmann I. M., Kensche P., Vogel J., Kohl S., Herzel H., Hess W. R. Genome Biol. 2005;6:R73. [PMC free article] [PubMed]
24. Zinchenko V. V., Piven I. V., Melnik V. A., Shestakov S. V. Russ. J. Genet. 1999;35:228–232.
25. Dühring U., Irrgang K. D., Lünser K., Kehr J., Wilde A. Biochim. Biophys. Acta. 2006;1757:3–11. [PubMed]

Articles from Proceedings of the National Academy of Sciences of the United States of America are provided here courtesy of National Academy of Sciences
PubReader format: click here to try


Related citations in PubMed

See reviews...See all...

Cited by other articles in PMC

See all...


Recent Activity

Your browsing activity is empty.

Activity recording is turned off.

Turn recording back on

See more...