Resequencing of a mutant bearing an iron ... - Wiley Online Library

The Plant Journal (2018) 93, 235–245

doi: 10.1111/tpj.13770

Resequencing of a mutant bearing an iron starvation recovery phenotype defines Slr1658 as a new player in the regulatory network of a model cyanobacterium Hagit Zer1,†, Ketty Margulis1,†, Jens Georg2, Yoram Shotland3, Gergana Kostova2, Laure D. Sultan1, Wolfgang R. Hess2,4 and Nir Keren1,* 1 Department of Plant and Environmental Sciences, Edmond J. Safra Campus, The Alexander Silberman Institute of Life Sciences, Hebrew University of Jerusalem, Givat Ram, Jerusalem, Israel, 2 Faculty of Biology, Genetics and Experimental Bioinformatics, University of Freiburg, Scha€ nzlestr. 1, D-79104 Freiburg, Germany, 3 Department of Chemical Engineering, Shamoon College of Engineering, Beer Sheva 84100, Israel, and 4 Freiburg Institute for Advanced Studies, University of Freiburg, Albertstr. 19, D-79104 Freiburg, Germany Received 17 March 2017; revised 26 October 2017; accepted 30 October 2017; published online 20 November 2017. *For correspondence (e-mail [email protected]). † These authors contributed equally to this work.

SUMMARY Photosynthetic microorganisms encounter an erratic nutrient environment characterized by periods of iron limitation and sufficiency. Surviving in such an environment requires mechanisms for handling these transitions. Our study identified a regulatory system involved in the process of recovery from iron limitation in cyanobacteria. We set out to study the role of bacterioferritin co-migratory proteins during transitions in iron bioavailability in the cyanobacterium Synechocystis sp. PCC 6803 using knockout strains coupled with physiological and biochemical measurements. One of the mutants displayed slow recovery from iron limitation. However, we discovered that the cause of the phenotype was not the intended knockout but rather the serendipitous selection of a mutation in an unrelated locus, slr1658. Bioinformatics analysis suggested similarities to two-component systems and a possible regulatory role. Transcriptomic analysis of the recovery from iron limitation showed that the slr1658 mutation had an extensive effect on the expression of genes encoding regulatory proteins, proteins involved in the remodeling and degradation of the photosynthetic apparatus and proteins modulating electron transport. Most significantly, expression of the cyanobacterial homologue of the cyclic electron transport protein PGR5 was upregulated 1000-fold in slr1658 disruption mutants. pgr5 transcripts in the Dslr1658 mutant retained these high levels under a range of stress and recovery conditions. The results suggest that slr1658 is part of a regulatory operon that, among other aspects, affects the regulation of alternative electron flow. Disruption of its function has deleterious results under oxidative stress promoting conditions. Keywords: cyanobacteria, iron homeostasis, oxidative stress, photosynthesis, regulatory systems, Synechocystis sp. PCC 6803.

INTRODUCTION Cyanobacteria, Gram-negative photoautotrophic prokaryotes, require substantial quantities of iron (Fe) to maintain many biological activities, including photosynthesis. Cyanobacterial iron requirements exceed those of nonphotosynthetic prokaryotes approximately 10-fold and are exceptionally high even among other photosynthetic organisms (Keren et al., 2004; Kranzler et al., 2013). Although iron is abundant in the Earth’s crust it is nevertheless a limiting factor in many marine, freshwater and © 2017 The Authors The Plant Journal © 2017 John Wiley & Sons Ltd

terrestrial environments. In the Earths’ oxidative environment the most common iron ion, Fe3+, suffers from very low solubility. In water bodies, iron bioavailability is typically in the subnanomolar range (Byrne and Kester, 1976). Surges in iron concentration occur following short lived Aeolian dust deposition or deep mixing events (Richier et al., 2012; Shaked and Lis, 2012). These aspects of iron geochemistry impose conditions in which the proliferation of photosynthetic organisms depends on their ability to 235

236 Hagit Zer et al. maintain homeostasis during the transition into limiting conditions and, equally important, during the transition out of limiting conditions. To thrive in aquatic environments photosynthetic organisms must negotiate these dynamic transitions in iron bioavailability. When transitioning into iron limitation, iron must be released from storage and safely transported to cofactor synthesis pathways, avoiding potential oxidative damages from free intracellular Fe. In the low Fe acclimated state the internal Fe quota of Synechocystis sp. PCC 6803 (a laboratory model strain, hereafter Synechocystis 6803) drops by a factor of 5 (Sharon et al., 2014), while the maximal transport rate increases 40-fold (Kranzler et al., 2011). When iron bioavailability suddenly increases internal quotas might surge and, with them, the risk of massive oxidative damages. In laboratory studies the focus was, to a large extent, on the transition into limitation, while the dynamics of transition out of limitation received less attention. In this study we probed transitions in both directions, in two bacterioferritin co-migratory peroxiredoxin deletion mutants (PRX; see Section S1 for more details on these proteins). One of which (Dslr0242) exhibited an iron starvation recovery phenotype. However, surprisingly, this phenotype was found to be derived from a different gene, unrelated to peroxiredoxins – the slr1658 gene. RESULTS Analysis of mutations in slr0242 and sll0221, encoding bacterioferritin associated PRXs Analysis of the two disruption mutants in the putative PRX genes slr0242 and sll0221 was performed during transition into and out of iron limitation. We used two medium types, 10Fe (YBGII, 10 lM Fe: 16 lM EDTA) and 0Fe (YBGII, 16 lM EDTA, no added Fe). Cultures grown in 10Fe were transitioned to 0Fe for 7 days and then back to 10Fe for up to 3 additional days. Cultures were also transferred between 10Fe medium as a control. To account for the inherent variability of growth curves we repeated each experiment four times, the statistical significance of the results is reported in Table 1 (additional details on the analysis of slr0242 and sll0221 disruption strains in Section S1). Our results suggested that slr0242 is linked to intracellular Fe homeostasis (Table 1, Section S1 and Figures S1–S3). The knockout strain did not recover properly following the transition from 0Fe to 10Fe (0Fe?10Fe). Growth did not resume, photosystem I (PSI) activity remained low and IsiA fluorescence remained high. IsiA is an iron-regulated chlorophyll a-binding protein which can act as an antenna when bound to PSI but can also dissipate absorbed light energy (Chen and Bibby, 2005; Ryan-Keogh et al., 2012; Ma et al., 2017). slr0242 is located close to slr0243, encoding a hypothetical protein (Figure S1(a)). Both genes are co-transcribed as

Table 1 Changes in culture biomass during the two phases of the growth experiments DOD730(0–7 d)

DOD730(7–10 d)

10Fe?10Fe 3.3 0.7 (100%) 3.0 0.4 (91%) 2.8 1.2 (84%)

10Fe?10Fe 2.3 1.4 (100%) 2.5 0.5 (108%) 1.5 1.0 (65%)

(a) Fe experiments (n = 4) Wild type Dsll0221 Original Dslr0242 strain (Dslr0242_slr1658m)

10Fe?0Fe Wild type 1.8 0.3 (100%) Dsll0221 2.1 0.4 (116%) Original Dslr0242 strain 1.5 0.4 (83%) (Dslr0242_slr1658m) (b) slr1658 complementation experiments (n = 10Fe?0Fe Wild type 1.24 0.34 (100%) Original Dslr0242 strain 1.29 0.24 (104%) (Dslr0242_slr1658m) Dslr0242_slr1658c 1.30 0.17 (105%)

0Fe?10Fe 2.2 1.4 (100%) 2.3 1.0 (105%) 0.2 0.2 (9.1%)** 3) 0Fe?10Fe 0.67 0.05 (100%) 0.23 0.21 (35%)* 0.75 0.19 (112%)

The protocol used in these experiments was as follows: Cultures grown in 10Fe were transitioned to 0Fe for 7 days and then back to 10Fe for up to 3 additional days: 10Fe ! 0Fe½7d ! 10Fe½3d: (The arrows represent transitions between medium types; see Figures 1 and S2 for examples.)As a control, cultures were transferred between 10Fe media: 10Fe ! 10Fe½7d ! 10Fe½3dDifferences in optical density (OD) values are reported (DOD730). Values in parentheses are percent of the corresponding wild type value. Student t-test was performed on the data. Dslr0242 cultures were found to be significantly different from the wild type following 0Fe?10Fe transition with **P < 0.005 and *P < 0.02. Significance values are in boldface.

part of the dicistronic transcriptional unit TU127, from a single transcriptional start site at position 143881 (Kopf et al., 2014). To clarify the genomic context of the phenotype we produced two additional disruption mutants in the two genes in the operon (Section S2 and Figure S4). Surprisingly, the distinct phenotype observed following the 0Fe?10Fe transition in the original Dslr0242 was completely lost in these mutants. Thus, we concluded that the originally observed Dslr0242 phenotype must have been caused by a different mutation. Whole genome resequencing of the Dslr0242 strain identifies an unrelated single frameshift mutation in the slr1658 gene To identify the genetic basis of the above described phenotype, MiSeq Illumina resequencing was carried out on DNA from the Dslr0242 and, as a control, on the ‘wild type’ that was the genetic background for the mutant strain. This procedure resulted in identification of a single point mutation in the slr1658 gene (Section S2). This mutation is a deletion of one nucleotide, causing a frameshift (Figure S5). Therefore, to properly describe the original Dslr0242 strain we renamed it as follows, Dslr0242_slr1658m, to indicate the point mutation in the slr1658 gene. slr1658 is

© 2017 The Authors The Plant Journal © 2017 John Wiley & Sons Ltd, The Plant Journal, (2018), 93, 235–245

Slr1658 as a new player in the regulatory network 237

(a)

Validation of the slr1658 phenotype in complementation and knockout strains

The slr1658 gene is part of an operon coding for putative regulatory proteins slr1658 is part of an operon transcribed from a transcriptional start site at position 3459716 (Kopf et al., 2014). This operon contains three genes, all annotated as unknown or hypothetical. Searches against the Pfam database returned no evidence for the presence of characterized domains in these proteins. However, searches against the SMART/ STRING databases (Letunic et al., 2015) revealed the presence of a histidine kinase-like ATPase domain (HATPase_c) (e-value 2 9 e10) in Slr1658. Slr1659 was predicted as its closest possible interaction partner based on multiple

Wild type Δslr0242_slr1658m 0Fe→10Fe Δslr0242_slr1658c

2.0 1.6

10Fe→0Fe

1.2 0.8 0.4 0.0

0

1

2

3

4

5

6

7

8

9

10

Time (days)

(b)

Wild type

Δslr0242_slr1658m Δslr0242_slr1658c

(c) 1200

Maximal P700 absorbance (ΔAmax705 nm)

To unequivocally link the recovery phenotype to slr1658 we complemented the original Dslr0242_slr1658m frameshifted strain with a wild type copy of slr1658 (Figure S6(a)). The resulting strain, Dslr0242_ slr1658c, exhibited growth rates identical to those of the wild type, upon transition into and out of iron limitation (Figure 1 and Table 1(b)). Maximal PSI activity in the complemented strain was also restored to wild type values during recovery (Figure 1(c)). These results verified the connection between the slr1658 mutation and the 0Fe?10Fe transition phenotype. To verify our observation with the point-mutation mutant, we reconstructed the slr1658 knockout mutant strain (Dslr1658; Figure S6(b)). The phenotypes associated with the point mutation were reproduced in the new mutant, growth recovered slowly from iron limitation (Figure 2a), PSI activity was also slow to rise following the 0Fe?10Fe transition (Figure 2b). Chlorophyll fluorescence spectroscopy indicated the continued presence of IsiA 3 days after the 0Fe?10Fe transition (Figure 2c). The effect observed during recovery was not a result of differences in the viability of wild type and mutant cells at the end of the limitation period. Sytox live/dead staining indicated that the fraction of live cells, while being lower in iron limited cultures, was similar for both wild type and Dslr1658 (Figure 2d, Section S3 and Figure S8). Iron and oxidative stress are tightly linked (Shcolnick et al., 2009) and the 0Fe?10Fe transition raises the risk of oxidative damage. Indeed, during recovery, Dslr1658 exhibited increased sensitivity to externally applied hydrogen peroxide (Figure 3). These results indicated a problem in handling and storing iron following the 0Fe?10Fe surge in Fe bioavailability.

2.4

Absorbance (OD 730 nm)

annotated as coding for a protein of unknown function, 199 amino acids in length and with a molecular mass of 22.6 kDa. As a result of the mutation, the C-terminal sequence of Slr1658 changed and became truncated by 21 AA (Figure S5(c)).

Wild type Δslr0242_slr1658m Δslr0242_slr1658c

1000 800 600 400 200 0

0

1

2

3

Time after 0Fe 10Fe transition(days) Figure 1. Growth and PSI activity of wild type, Dslr0242_slr1658m and Dslr0242_slr1658c strains through 10Fe?0Fe?10Fe transition. (a) Changes in culture density measured as OD730 throughout the experiment. Fe transition points are marked by dotted lines. Statistical analysis is included in Table 1(b). (b) Cultures at the end of the experiment. (c) PSI activity measured as absorbance changes at 705 nm. The values plotted are for the maximal absorbance change recorded in the presence of DCMU and DBMIB, DAmax (Salomon and Keren, 2011). This value represents the number of active PSI reaction centers per cell. T0 values are similar to those recorded in the past for Fe limited cells (Sharon et al., 2014; Salomon and Keren, 2015). [Colour figure can be viewed at wileyonlinelibrary.com]

evidence (Szklarczyk et al., 2015), including gene fusions between the respective homologues, e.g., in the cyanobacteria Mastigocoleus testarum (Sequence ID: ref|WP_063800790.1|) and Rivularia sp. PCC 7116 (protein Riv7116_2462). The analysis of phylogenetic relationships between Slr1658 homologues demonstrated a signal that is not completely congruent with taxonomy. Homologues are missing, for example, from all Prochlorococcus strains (Figure S7) and from photosynthetic eukaryotes. They can,


238 Hagit Zer et al.

(a)

(b)

2000

0Fe→10Fe

Wild type Δslr1658

Maximal P700 absorbance (ΔAmax705 nm)

Absorbance (OD 730 nm)

2.0

10Fe→0Fe

1.5

1.0

0.5

1200

800

400

0

0.0 0

1

2

3

4

5

6

7

8

9

0

10

Wild type Δslr1658

(d)

Day 7

2

3

10Fe transtion (days)

Sytox positive Sytox negative

Day 10

1.0

Fraction of Sytox negative or positive cells

Normalized ﬂuorescence intensity

(c)

1

Time after 0Fe

Time (days)

1.0

Wild type Δslr1658

1600

0.8

0.6

0.4

0.2

0.8

0.6

0.4

0.2

0.0 0.0 600

650

700

600

650

700

750

Wavelength (nm)

Δslr1658

Wild type

7d 0Fe

Wild type

7d 10Fe

Figure 2. Iron-related phenotype of the slr1658 knockout insertion strain. (a) Growth of wild type and Dslr1658 cultures during 10Fe?0Fe and 0Fe?10Fe transition. The Dslr1658 strain exhibits slower recovery from the iron limitation period (n = 3). (b) PSI activity measured as in Figure 1. DAmax represents the number of active PSI reaction centres per OD730. PSI activity in the mutant strains was significantly lower throughout the 3 days following the 0Fe?10Fe transition (n = 3). (c) 77 K chlorophyll fluorescence spectra measured using excitation at 430 2 nm, exciting chlorophylls preferentially. The characteristic IsiA peak at 680 nm remains clearly visible in the mutant strain 3 days after the 0Fe?10Fe transition while in the wild type a typical iron replete curve is observed. (d) Viability test at the end of iron limitation period (day 7, n = 3). As a control wild type cells were grown on 10Fe media for the same period of time.

however, be found in non-oxygenic photosynthetic organisms such as Rhodospirillum rubrum and chemoorgano-heterotrophs like Beggiatoa alba. The latter organisms represent strains with well documented oxygen tension-related shifts in their metabolic activities (Johnston and Brown, 1954; Schmidt et al., 1987). Further bioinformatics analysis revealed significant similarities to cyclic di-GMP binding proteins for Slr1657 and the presence of a STAS (Sulphate Transporter and Anti-Sigma factor antagonist) domain in Slr1659 (So€ ding et al., 2005). Additional bioinformatics details are summarized in Table S1. Slr1659 was observed in a study of the Synechocystis 6803 phospho-proteome (Mikkat et al., 2014). The operon

structure is conserved in Synechocystis sp. PCC 6714, but in other cyanobacterial strains homologues may also appear separately in the genome. Altogether, the bioinformatics data indicate that this operon is involved in a hitherto unknown signal transduction, sensory and regulatory process. Transcriptome remodeling in the slr1658 frameshifted mutant To test the effect of the mutation in slr1658 on regulatory events during the 0Fe?10Fe transition, we performed transcriptome analysis on the Dslr0242_slr1658m and the Dslr0242_slr1658c strains (detailed in Section S4). The analysis resulted in three major findings:


Slr1658 as a new player in the regulatory network 239 Control 7d 10Fe→ 2d 10Fe

Recovery 7d 0Fe→ 2d 10Fe

0

H2O2 [mM]

1 2 4 8

W ild Δs type lr1 65 8

Δs l

0F e→

2d

10 7d

10

Fe

→

2d

2d 0F e→ 7d

10

Fe

Fe 10

Fe 10 2d → Fe 10 7d

Fe

Δslr1658

Wild type

7d

W ild

r1

65

8

ty pe

16

Figure 3. Sensitivity of wild type and Dslr1658 to hydrogen peroxide. Cultures were grown for 7 days in 0Fe or in 10Fe media and then transferred to 10Fe for 2 additional days, followed by exposure to a range of H2O2 concentrations for 20 h in darkness (bottom). Aliquots from each treatment were spotted onto BG11 plates (Top). Viable cultures produced colonies on the agar plate in the picture (Top). Three independent repeats are shown for the 0Fe?10Fe recovery and one for the 10Fe?10Fe control. [Colour figure can be viewed at wileyonlinelibrary.com]

(i) The transcriptional profile of Dslr0242_ slr1658m was significantly different from that of the complemented strain already in 0Fe medium (time point 1 before the 0Fe?10Fe transition). (ii) Functional enrichment analysis indicated that the most effected clusters contained translation-associated genes and a wide range of PSI, photosystem II (PSII) and phycobilisome genes. (iii) The largest effect was detected in the TU895 transcriptional unit encompassing a single gene, ssr2016 (Table 2). ssr2016 codes for a homologue of Pgr5 (Yeremenko et al., 2005). In plants (Munekage et al., 2002) and in algae (Saroussi et al., 2016) Pgr5 was shown to be involved in the antimycin A-sensitive electron flow from PSI to the plastoquinone (PQ) pool. A ssr2016 mutant of Synechocystis 6803 was also affected in this alternative route (Yeremenko et al., 2005). Interestingly, ssr2016 was also induced during transcriptome analysis of the sml0013 mutant, encoding the NdhP subunit of the cyanobacterial NDH1 complex, which is also involved in alternative electron transport (Schwarz et al., 2013).

To verify this major result, we repeated the experiment and examined the expression profile of the ssr2016 gene in the wild type, Dslr0242_slr1658m, Dslr0242_slr1658c and Dslr1658 strains by quantitative polymerase chain reaction (PCR) (Figure 4). In the wild type strain, ssr2016 remained low through the recovery period, from 1 h to 24 h, as compared with the approximately 1000-fold (210) higher transcript levels observed in the two slr1658 disrupted strains (Dslr0242_slr1658m and Dslr1658). The complemented strain, Dslr0242_slr1658c, exhibited an only 10-fold increase in Pgr5 transcript levels (Figure 4). The approximately 100-fold difference between Dslr0242_ slr1658m and Dslr0242_ slr1658c observed here fits well with the microarray data (Table 2). These results suggest that the complementation in Dslr0242_slr1658c, which rescued the physiological phenotype, is 99% complete on the transcriptional level. Probing the regulatory function of slr1658 Our analysis provided evidence for extensive slr1658-dependent changes in the transcriptome. The most pronounced effect was observed on the transcript accumulation of pgr5 but multiple additional genes were also significantly affected. To get a clearer picture on the role of Slr1658 in the Synechocystis 6803 regulatory network we selected a number of representative transcripts and stress conditions for further analysis. In addition to pgr5 we tested the iron€ hring et al., 2006) and stress inducible transcript isiA (Du the nitrogen limitation inducible nblA transcript (Klotz et al., 2015). These three genes exhibited significantly altered expression profiles during Fe limitation (Table 2). Initially, the levels of these transcripts were quantified under control conditions in Dslr1658 versus wild type (full YBG11 medium, logarithmic cultures Figure 5(a)). The transcript levels of nblA and isiA were higher in Dslr1658 as compared with the wild type. pgr5 levels were more than 2000-fold higher. When examining the transcriptional profile of pgr5 under stress and during recovery we found that its levels are consistently higher in Dslr1658 than in the wild type with very little change over the range of different conditions (Figure 5(b)). In comparison, the levels of isiA (Figure 5(c)) exhibited a higher basal level in the mutant but a very similar response to stress and recovery to the wild type. nblA exhibited a similar trend with an identical response to nitrogen stress treatment in the wild type and the mutant (Figure 5(d)). DISCUSSION The data presented in this manuscript expose an unexpected connection between iron homeostasis and the function of Slr1658. The bioinformatics prediction together with the transcriptome analysis suggests that this protein has a regulatory role. Several regulators involved in iron sensing and control were previously described, including


240 Hagit Zer et al. Table 2 ‘Top Ten’ list of differentially regulated genes in the Dslr0242_ slr1658m and Dslr0242_ slr1658c strains during a 0Fe?10Fe transition. [Colour figure can be viewed at wileyonlinelibrary.com]


pgr5 transcript aboundance log2 [(sample/rnpB)/(Wild type–1 h/rnpB–1 h)]

Slr1658 as a new player in the regulatory network 241 14 12 10

–1 h 3h 6h 24 h

8 6 4 2 0 –2

Wild type

Δslr1658

Δslr0242_ 1658c

Δslr0242_ 1658m

Figure 4. Validation of Pgr5 phenotype by quantitative RT-PCR in wild type, Dslr0242_slr1658m, Dslr0242_slr1658c and Dslr1658. Transcriptional changes in the pgr5 gene during the 0Fe?10Fe transition (performed at T0). The sampling regime is identical to that of the microarray experiment (Section S4). Log2 values of pgr5 transcription abundance are presented. The data were calculated relative to time 1 h of the wild type strain. Transcript levels in each sample were internally normalized to the expression of the housekeeping rnpB gene. Error bars represent standard deviation for three biological repeats. A two-way analysis of variance (ANOVA) test was conducted on wild type versus Dslr1658 and wild type versus Dslr0242_slr1658m (factors: genotype and time). The genotype factor yielded statistically significant results in both cases (P-value < 0.005).

Fur proteins (Garcin et al., 2012; Fillat, 2014; Krynicka et al., 2014), the TetR family regulator PfsR (Liu et al., 2014) and the non-coding RNA IsaR1 (Georg et al., 2017). However, the pattern of regulated genes affected by Slr1658 differed substantially from the previously identified iron-stress response in Synechocystis 6803 (Hernandez-Prieto et al., 2012). It also differed from the transcriptional response to high light (Muramatsu and Hihara, 2012), phosphorus (Suzuki et al., 2004), or nitrogen starvation (Krasikov et al., 2012). The physiological phenotypes and the transcriptional profiles observed in slr1658 mutants do not fully overlap with any of these, suggesting it is a new player in the cyanobacterial regulatory network. Transcriptional profiling under stress and during recovery from a number of limiting conditions indicated that the prime target of slr1658 is the cyanobacterial homologue of Pgr5 (Yeremenko et al., 2005). Regardless of the conditions, in slr1658 mutants pgr5 transcript levels are consistently higher than those of the wild type. In chloroplasts, Pgr5 functions in the antimycin A-sensitive cyclic electron flow route. It is essential for electron transport from ferredoxin back into the PQ pool in this route, but is not sufficient (Johnson, 2011; Leister and Shikanai, 2013). It functions as part of a complex with PgrL1 and different models place it in association with PSI or with cytochrome b6f. In cyanobacteria, the Pgr5 homologue was shown to have a similar effect on alternative

electron transfer (Yeremenko et al., 2005). However, there is no homologue for PgrL1 in cyanobacteria and the identity of possible interacting proteins in alternative electron flow are still unknown (Yeremenko et al., 2005). Transcriptional changes during iron limitation and recovery were not limited to pgr5. The transcriptional profile demonstrated an increase in the expression of genes related to oxidative stress response and the degradation of photosynthetic components in Dslr0242_slr1658m. These transcriptional effects were verified in Dslr1658. Oxidative stress and iron homeostasis are tightly linked in photosynthetic organisms (Shcolnick et al., 2009) and through this link to many additional environmental stresses affecting photosynthetic and respiratory electron flow pathways. In addition to the effect on pgr5, we observed changes in the transcriptional profile of stress responsive genes under nitrogen limitation, high light treatment and during recovery from these conditions. However, for these stress response genes it seemed that the Dslr1658 mutation affected the basal level more than the extent of response to stress and recovery. Based on the data presented in this paper we propose that the function of Slr1658 is regulatory and that its primary target is Pgr5. The systemic effects observed in Dslr1658 could be the result of lesions accumulated during the limitation period or due to an inability to cope with the oxidative stress during the recovery period. Based on the similar growth rates, culture viability and physiological parameter under Fe limitation we suggest that the effect is a direct result of an inability to cope with recovery. Pgr5 function diverts electrons flow from NADPH, which can be used for reactive oxygen species (ROS) detoxification, to ATP. It is overexpression can, therefore, be detrimental under these conditions. Our findings define Slr1658 as a new player in the regulatory network controlling alternative electron flow in cyanobacteria. EXPERIMENTAL PROCEDURES Media and growth conditions Stock cultures of the glucose tolerant strain of Synechocystis sp. PCC 6803 and mutants were grown in YBG11 (Shcolnick et al., 2007) medium containing 10 lM iron (10Fe – YBG11 media). Stock cultures of Dslr0242_slr1658m, Dslr0242–2, Dslr0243 and sll0221 strains were supplemented with 50 lg/ml kanamycin. Dslr0242_slr1658c strains were supplemented with 50 lg/ml kanamycin and 25 lg/ml ampicillin. To achieve iron depletion (0Fe conditions) cells were spun down, resuspended in a MES-EDTA solution (20 mM 2-ethanesulfonic acid [MES], 10 mM ethylenediaminetetraacetic acid [EDTA], pH 5) and placed on a shaker for 20 min. The procedure was repeated twice. Subsequently, cells were resuspended in YBG11 medium with or without added iron (10Fe or 0Fe, respectively). Cultures were grown in 250 ml Erlenmeyer flasks at 30°C with constant shaking. Light intensity was set to 60 lmol photons m2 sec1. Viability was assessed by Sytox Green dead cell stain (Molecular Probes) (Cohen et al., 2014) using Becton Dickinson FACS Calibur.


242 Hagit Zer et al.

control condtions

(a)

pgr5 expression

(b) 14 12

Wild type

log2[(Δslr1658/rnpB)/(Wild type /rnpB)]

Transcript aboundance

10 8

14

6 12

4 2

10

0

8

–2

6

12

4

Δslr1658

10 8

2

6 0

4 pgr5

isiA

nblA

2

–2

0 –2

(c) 14 12

(d)

isiA expression

12

10

10

8

8

6

6

4

4

2

2

0

0

–2

–2

12

Δslr1658

12

10

10

8

8

6

6

4

4

2

2

0 –2

0Fe

0N

nblA expression

14

Wild type

HL

Wild type

Δslr1658

0 HL

0Fe

0N

–2

HL

0Fe

0N

Y axis scale (panels B-D): Transcript aboundance - log2[(treatment/rnpB)/(Wild typecontrol/rnpBcontrol)] Stress treatment HL, 0Fe, 0N.

Recovery

Figure 5. Function of slr1658 under multiple stress and recovery conditions. (a–d)The transcript levels of pgr5, isiA and nblA were tested under control growth conditions (a) as well as under stress and during recovery (b–d). Control conditions were 10Fe YBGII medium and growth light (60 lmol photons m2 sec1). For high light treatment (HL) cultures were exposed to a light intensity of 330 lmol photons m2 sec1 for 24 h (HL time point) followed by return to growth light for 6 h (recovery time point). Fe stress conditions were 0Fe medium for 7 days followed by a 2 day recovery in 10Fe. For nitrogen limitation cells were grown in 0N (no added nitrogen YBGII) for 3 days followed by 2 days in 17.5 mM N. Log2 values of the transcription abundance of the pgr5, isiA and nblA genes are presented. Transcript levels were determined by quantitative RT-PCR as in Figure 4. The data are normalized internally to rnpB and to the transcript level of the wild type under control conditions. Error bars represent standard deviation for three biological repeats.


Slr1658 as a new player in the regulatory network 243 Generation of the mutant strains

Microarray hybridization

The procedure and the primers used for generating the different mutant strains is included in Figure S1(a–c). Transformation was performed as described in Eaton-Rye (2011).

Two replicates of total RNA extractions were used in the microarray experiments. The used microarray design covered all genes and intergenic regions for which transcripts were previously detected (Georg et al., 2009; Mitschke, 2011; Kopf et al., 2014). Each time point was hybridized in duplicates. For RNA hybridizations, the DNase-treated RNA was labeled directly, without cDNA synthesis in 2 lg aliquots with Kreatech’s ULS labeling kit for Agilent gene expression arrays with Cy3 according to the manufacturer’s protocol. Fragmentation and hybridization was performed with 600 ng of labeled RNA according to the Agilent protocol for 8 9 60 k single colour microarrays. Arrays were scanned on the Agilent Technologies Scanner G2505B, using Agilent Feature Extraction Software 10.7.3.1 and the protocols G3_GX_1 colour for Cy3-labelled arrays. Raw data were processed with the R package Limma. Median signal intensity was normexp background corrected and quantile normalized. Analysis of the two replicates provided very similar data sets (Figure S10). All data resulting from microarray analysis were deposited at NCBI’s GEO database with the accession number GSE102876.

Induction of oxidative stress Cultures were grown for 7 days in 0Fe medium and then transferred to 10Fe for an additional 2 days. The cells were then washed in MES-EDTA and the optical density at 730 nm (OD730) was adjusted to a final value of 0.9 in 0Fe medium. This was followed by 20 h of incubation in the presence of 0–16 mM H2O2 in darkness. To assess viability after the H2O2 treatment, 2 ll from each treatment were spotted onto BG11 plates.

Spectroscopy Cell density was measured as OD730 (Salomon and Keren, 2011), using a Carry3000-Bio spectrophotometer (Varian, CA, USA). 77 K chlorophyll fluorescence spectra were measured using a Fluoromax-3 spectrofluorometer (Jobin Ivon, Longjumaeu, France) as described in (Salomon and Keren, 2011). PSI activity was measured as P700 photo-oxidation using a JTS-10 spectrophotometer (Salomon and Keren, 2011). In order to block electron flow to PSI, 10 lM 3-(3,4-dichlorophenyl)-1,1dimethylurea (DCMU) and 10 lM 2,5-dibromo-3-methyl-6-isopropylbenzoquinone (DBMIB) were added.

Sequencing Wild type and Dslr0242 DNA was sequenced using MiSeq Illumina. We performed paired-end sequencing of 250 nt each. To build the genome of the mutant and the wild type, the reads were mapped against the Synechocystis 6803 substrain GT-I (gi| 359270229|dbj|AP012276.1|) genome as a reference. Out of all reads, 5 098 206 reads (98.3%) were mapped as paired. The mean coverage of the genome was 358 with a confidence mean of 35.3.

Transcript analysis – RNA extraction and analysis Cultures of Dslr0242_slr1658m and Dslr0242_slr1658c used for transcript analysis were collected at several time points before and after the 0Fe?10Fe transition, centrifuged at 1700 g for 10 min at 30°C and frozen in liquid nitrogen. Samples were then resuspended in PGTX and incubated for 15 min at 65°C in a water bath (Pinto et al., 2009). After addition of 700 ll of chloroform/isoamylalcohol (24:1) and thorough agitation samples were incubated at room temperature for 10 min. Samples were centrifuged with a swing out rotor at 5000 rpm, 23°C for 5 min. The upper aqueous phase was transferred into a new vial and the same volume of chloroform/isoamylalcohol (24:1) was added and mixed. Samples were centrifuged as before and the aqueous phase transferred again and mixed with an equal volume of isopropanol. After gently inverting the tube, RNA was precipitated overnight at 20°C. RNA was pelleted through centrifugation (13 000 rpm, 4°C, 30 min). The pellet was washed with 200 ll of 70% ethanol (13 000 rpm, 4°C, 10 min), air dried for 10 min and resuspended in 20 ll H2O.

Northern blot analysis Northern blot analysis was performed using 3 lg of total RNA on 1.3 % agarose formaldehyde gels as described (Dienst et al., 2008). The primers for generating the probe to detect the cis-encoded antisense RNA IsrR are given in Figure S1(c).

Real-time quantitative PCR Samples for quantitative RT-PCR analysis were collected during the 0Fe?10Fe transition. RNA extraction was performed essentially as for the microarray analysis. DNase treatment (Turbo-DNAfree, Life Technologies, Grand Island, NY, USA) was followed by reverse transcription (Revertaid First Strand cDNA Synthesis Kit, Thermo Scientific, Wilmington, DE, USA) and the resulting cDNA was used directly for Q-RT-PCR. qPCR reactions were run on a LightCycler 480 (Roche), using 2.5 ll of LightCycler 480 SYBR Green I Mastermix and 2.5 lM forward and reverse primers in a final volume of 5 ll. Reactions were performed in triplicates in the following conditions: preheating at 95°C for 10 min, followed by 40 cycles of 10 sec at 95°C, 10 sec at 58°C, and 10 sec at 72°C. A calibration curve was generated for all samples. Primer sequences for each gene are presented in Table S1. Three independent cultures were measured with at least two technical repeats each. The results were confirmed to be aligned to a linear calibration curve. Pgr5 data points were internally normalized to rnpB levels, an established housekeeping gene in Synechocystis 6803 (Pinto et al., 2012) and to wild type time 0.

ACKNOWLEDGEMENTS This work was supported by the joint ISF-UGC grant (2733/16) awarded to NK; The Advanced School Environmental Science scholarship awarded to KM; the People Programme (Marie Curie Actions) of the European Union’s Seventh Framework Programme (grant 317184 to W.R.H).

CONFLICT OF INTEREST The authors declare no conflict of interest. SUPPORTING INFORMATION Additional Supporting Information may be found in the online version of this article. Section S1: A study of the function of bacterioferritin and bacterioferritin co-migratory proteins Figure S1. Genes, operons and primers. Figure S2. Growth of wild type and mutant strains through the 10Fe?0Fe?10Fe transition.


244 Hagit Zer et al. Figure S3. Photosynthetic performance during recovery from Fe limitation. Section 2: Identification of the slr1658 frameshift mutation Figure S4. Growth of wild type and newly constructed bactrioferritin co-migratory protein mutant strain through the 10Fe?0Fe? 10Fe transition. Figure S5. The frameshift mutation in slr1658 and its results. Figure S6. Complementation of the frameshift mutation in slr1658. Figure S7. Sequence relationships among Slr1658 homologues. Table S1. HHPred bioinformatics analysis. Section 3: Viability under Fe limitation Figure S8. FACS raw data for Sytox stained cells. Section 4: Transcriptional analysis Figure S9. Changes in the transcript levels of an iron responsive non-coding-RNA in Dslr0242_slr1658m compared with Dslr0242_slr1658c. Figure S10. PCA analysis of the transcriptome data. Figure S11. Transcriptomic differences in Dslr0242_slr1658m compared with Dslr0242_slr1658c at the 1 h time point. Figure S12. Evolution of transcriptomic differences in Dslr0242_slr1658m compared with Dslr0242_slr1658c following addition of iron.

REFERENCES Byrne, R.H. and Kester, D.R. (1976) Solubility of hydrous ferric oxide and iron speciation in seawater. Mar. Chem. 4, 255–274. Chen, M. and Bibby, T.S. (2005) Photosynthetic apparatus of antenna-reaction centres supercomplexes in oxyphotobacteria: insight through significance of Pcb/IsiA proteins. Photosynth. Res. 86, 165–173. Cohen, A., Sendersky, E., Carmeli, S. and Schwarz, R. (2014) Collapsing aged culture of the cyanobacterium synechococcus elongatus produces compound(s) toxic to photosynthetic organisms. PLoS ONE, 9, e100747. € hring, U., Mollenkopf, H.J., Vogel, J., Golecki, J., Hess, W.R. Dienst, D., Du and Wilde, A. (2008) The cyanobacterial homologue of the RNA chaperone Hfq is essential for motility of Synechocystis sp. PCC 6803. Microbiology, 154, 3134–3143. D€ uhring, U., Axmann, I.M., Hess, W.R. and Wilde, A. (2006) An internal antisense RNA regulates expression of the photosynthesis gene isiA. Proc. Natl Acad. Sci. USA, 103, 7054–7058. Eaton-Rye, J.J. (2011) Construction of gene interruptions and gene deletions in the cyanobacterium Synechocystis sp. Strain PCC 6803. Methods Mol. Biol. 684, 295–312. Fillat, M.F. (2014) The FUR (ferric uptake regulator) superfamily: diversity and versatility of key transcriptional regulators. Arch. Biochem. Biophys., 546, 41–52. Garcin, P., Delalande, O., Zhang, J.-Y., Cassier-Chauvat, C., Chauvat, F. and Boulard, Y. (2012) A transcriptional-switch model for Slr1738-controlled gene expression in the cyanobacterium Synechocystis. BMC Struct. Biol. 12, 1. Georg, J., Voss, B., Scholz, I., Mitschke, J., Wilde, A. and Hess, W.R. (2009) Evidence for a major role of antisense RNAs in cyanobacterial gene regulation. Mol. Syst. Biol. 5, 305. Georg, J., Kostova, G., Vuorijoki, L., et al. (2017) Acclimation of oxygenic photosynthesis to iron starvation is controlled by the sRNA IsaR1. Curr. Biol. 27(1425–1436), e7. Gray, J., Wardzala, E., Yang, M., Reinbothe, S., Haller, S. and Pauli, F. (2004) A small family of LLS1-related non-heme oxygenases in plants with an origin amongst oxygenic photosynthesizers. Plant Mol. Biol. 54, 39–54. € n, V., Georg, J., Barreira, L., Varela, J., Hess, Hernandez-Prieto, M.A., Scho W.R. and Futschik, M.E. (2012) Iron deprivation in Synechocystis: inference of pathways, non-coding RNAs, and regulatory elements from comprehensive expression profiling. G3 (Bethesda), 2, 1475–1495. Johnson, G.N. (2011) Reprint of: physiology of PSI cyclic electron transport in higher plants. Biochim. Biophys. Acta - Bioenerg. 1807, 906–911. Johnston, J.A. and Brown, A.H. (1954) The effect of light on the oxygen metabolism of the photosynthetic bacterium, Rhodospirillum rubrum. Plant Physiol. 29, 177–182.

Keren, N., Aurora, R. and Pakrasi, H.B. (2004) Critical roles of bacterioferritins in iron storage and proliferation of cyanobacteria. Plant Physiol. 135, 1666–1673. Klotz, A., Reinhold, E., Doello, S., Forchhammer, K. and Doello, S. (2015) Nitrogen starvation acclimation in synechococcus elongatus: redox-control and the role of nitrate reduction as an electron sink. Life, 5, 888–904. Komenda, J., Knoppova , J., Krynicka, V., Nixon, P.J. and Tichy, M. (2010) Role of FtsH2 in the repair of Photosystem II in mutants of the cyanobacterium Synechocystis PCC 6803 with impaired assembly or stability of the CaMn4 cluster. Biochim. Biophys. Acta - Bioenerg. 1797, 566–575. Kopf, M., Kla€hn, S., Scholz, I., Matthiessen, J.K.F., Hess, W.R. and Voß, B. (2014) Comparative analysis of the primary transcriptome of Synechocystis sp. PCC 6803. DNA Res. 21, 527–39. Kranzler, C., Lis, H., Shaked, Y. and Keren, N. (2011) The role of reduction in iron uptake processes in a unicellular, planktonic cyanobacterium. Environ. Microbiol. 13, 2990–2999. Kranzler, C., Rudolf, M., Keren, N. and Schleiff, E. (2013). Iron in cyanobacteria. In Advances Botanical Research (Chauvat, F. and Cassier- Chauvat, C., eds). Amsterdam: Genomics of Cyanobacteria, pp. 57–105. Krasikov, V., Aguirre von Wobeser, E., Dekker, H.L., Huisman, J. and Matthijs, H.C.P. (2012) Time-series resolution of gradual nitrogen starvation and its impact on photosynthesis in the cyanobacterium Synechocystis PCC 6803. Physiol. Plant, 145, 426–439. a, R., Boehm, M., Nixon, P.J. Krynicka, V., Tichy, M., Krafl, J., Yu, J., Kan and Komenda, J. (2014) Two essential FtsH proteases control the level of the Fur repressor during iron deficiency in the cyanobacterium S ynechocystis sp. PCC 6803. Mol. Microbiol. 94, 609–624. Leister, D. and Shikanai, T. (2013) Complexities and protein complexes in the antimycin a-sensitive pathway of cyclic electron flow in plants. Front. Plant Sci. 4, 161. Liu, L., Huang, C. and He, Z.-G. (2014) A TetR family transcriptional factor directly regulates the expression of a 3-methyladenine DNA glycosylase and physically interacts with the enzyme to stimulate its base excision activity in Mycobacterium bovis BCG. J. Biol. Chem. 289, 9065–9075. Letunic, I., Doerks, T. and Bork, P. (2015) SMART: recent updates, new developments and status in 2015. Nucleic Acids Res. 43, D257–60. Ma, F., Zhang, X., Zhu, X., Li, T., Zhan, J., Chen, H., He, C. and Wang, Q. (2017) Dynamic changes of IsiA-containing complexes during long-term iron deficiency in Synechocystis sp. PCC 6803. Mol. Plant, 10, 143–154. Mikkat, S., Fulda, S. and Hagemann, M. (2014) A 2D gel electrophoresisbased snapshot of the phosphoproteome in the cyanobacterium Synechocystis sp. strain PCC 6803. Microbiology, 160, 296–306. Mitschke, et al. (2011) An experimentally anchored map of transcriptional start sites in the model cyanobacterium Synechocystis sp. PCC 6803 Jan Mitschke. Proc. Natl Acad. Sci. USA, 108, 1–33. Munekage, Y., Hojo, M., Meurer, J., Endo, T., Tasaka, M. and Shikanai, T. (2002) PGR5 is involved in cyclic electron flow around photosystem I and is essential for photoprotection in Arabidopsis. Cell, 110, 361–371. Muramatsu, M. and Hihara, Y. (2012) Acclimation to high-light conditions in cyanobacteria: from gene expression to physiological responses. J. Plant Res. 125, 11–39. Pinto, F., Thapper, A., Sontheim, W. and Lindblad, P. (2009) Analysis of current and alternative phenol based RNA extraction methodologies for cyanobacteria. BMC Mol. Biol. 10, 79. de Porcellinis, A.J., Kla€hn, S., Rosgaard, L., Kirsch, R., Gutekunst, K., Georg, J., Hess, W.R. and Sakuragi, Y. (2016) The Non-Coding RNA Ncr0700/ PmgR1 is Required for Photomixotrophic Growth and the Regulation of Glycogen Accumulation in the Cyanobacterium Synechocystis sp. PCC 6803. Plant Cell Physiol. 57, 2091–2103. Pinto, F., Pacheco, C.C., Ferreira, D., Moradas-Ferreira, P. and Tamagnini, P. (2012) Selection of Suitable Reference Genes for RT-qPCR Analyses in Cyanobacteria B. Neilan, ed. PLoS One, 7, e34983. Pruzinska, A., Tanner, G., Aubry, S., et al. (2005) Chlorophyll breakdown in senescent Arabidopsis leaves. Characterization of chlorophyll catabolites and of chlorophyll catabolic enzymes involved in the degreening reaction. Plant Physiol. 139, 52–63. Richier, S., Macey, A.I., Pratt, N.J., Honey, D.J., Moore, C.M. and Bibby, T.S. (2012) Abundances of iron-binding photosynthetic and nitrogen-fixing proteins of Trichodesmium both in culture and in situ from the North Atlantic. PLoS ONE, 7, e35571.


Slr1658 as a new player in the regulatory network 245 Ryan-Keogh, T.J., Macey, A.I., Cockshutt, A.M., Moore, C.M. and Bibby, T.S. (2012) The cyanobacterial chlorophyll-binding-protein IsiA acts to increase the in vivo effective absorption cross-section of PSI under iron limitation. J. Phycol. 48, 145–154. Salomon, E. and Keren, N. (2011) Manganese limitation induces changes in the activity and in the organization of photosynthetic complexes in the cyanobacterium Synechocystis sp. strain PCC 6803. Plant Physiol. 155, 571–579. Salomon, E. and Keren, N. (2015) Acclimation to environmentally relevant Mn concentrations rescues a cyanobacterium from the detrimental effects of iron limitation. Environ. Microbiol. 17, 2090–2098. Saroussi, S.I., Wittkopp, T.M. and Grossman, A.R. (2016) The Type II NADPH dehydrogenase facilitates cyclic electron flow, energy-dependent quenching, and chlororespiratory metabolism during acclimation of Chlamydomonas reinhardtii to Nitrogen deprivation. Plant Physiol. 170, 1975–1988. Schmidt, T.M., Arieli, B., Cohen, Y., Padan, E. and Strohl, W.R. (1987) Sulfur metabolism in Beggiatoa alba. J. Bacteriol. 169, 5466–5472. Schwarz, D., Schubert, H., Georg, J., Hess, W.R. and Hagemann, M. (2013) The gene sml0013 of Synechocystis species strain PCC 6803 encodes for a novel subunit of the NAD(P)H Oxidoreductase or complex I that is ubiquitously distributed among Cyanobacteria. Plant Physiol. 163, 1191–1202. Sendersky, E., Kozer, N., Levi, M., Moizik, M., Garini, Y., Shav-Tal, Y. and Schwarz, R. (2015) The proteolysis adaptor, NblA, is essential for degradation of the core pigment of the cyanobacterial light-harvesting complex. Plant J. 83, 845–852. Shaked, Y. and Lis, H. (2012) Disassembling iron availability to phytoplankton. Front. Microbiol. 3, 123. Sharon, S., Salomon, E., Kranzler, C., Lis, H., Lehmann, R., Georg, J., Zer, H., Hess, W.R. and Keren, N. (2014) The hierarchy of transition

metal homeostasis: iron controls manganese accumulation in a unicellular cyanobacterium. Biochim. Biophys. Acta - Bioenerg. 1837, 1990– 1997. Shcolnick, S., Shaked, Y. and Keren, N. (2007) A role for mrgA, a DPS family protein, in the internal transport of Fe in the cyanobacterium Synechocystis sp. PCC6803. Biochim. Biophys. Acta - Bioenerg. 1767, 814– 819. Shcolnick, S., Summerfield, T.C., Reytman, L., Sherman, L.A. and Keren, N. (2009) The mechanism of iron homeostasis in the unicellular cyanobacterium synechocystis sp. PCC 6803 and its relationship to oxidative stress. Plant Physiol. 150, 2045–2056. € ding, J., Biegert, A. and Lupas, A.N. (2005) The HHpred interactive server So for protein homologuey detection and structure prediction. Nucleic Acids Res. 33, W244–W248. Suzuki, S., Ferjani, A., Suzuki, I. and Murata, N. (2004) The SphS-SphR two component system is the exclusive sensor for the induction of gene expression in response to phosphate limitation in Synechocystis. J. Biol. Chem. 279, 13234–13240. Szklarczyk, D., Franceschini, A., Wyder, S., et al. (2015) STRING v10: protein-protein interaction networks, integrated over the tree of life. Nucleic Acids Res. 43, D447–D452. Wang, Q., Jantaro, S., Lu, B., Majeed, W., Bailey, M. and He, Q. (2008) The high light-inducible polypeptides stabilize Trimeric photosystem I complex under high light conditions in Synechocystis PCC 6803. Plant Physiol. 147, 1239–1250. Yeremenko, N., Jeanjean, R., Prommeenate, P., Krasikov, V., Nixon, P.J., Vermaas, W.F.J., Havaux, M. and Matthijs, H.C.P. (2005) Open reading frame ssr2016 is required for antimycin A-sensitive photosystem I-driven Cyclic electron flow in the cyanobacterium Synechocystis sp. PCC 6803. Plant Cell Physiol. 46, 1433–1436.
