A single plasmid based CRISPR interference in Synechocystis 6803 – A proof of concept

Summary

We developed a simple method to apply CRISPR interference by modifying an existing plasmid pCRISPathBrick containing the native S. pyogenes CRISPR assembly for Synechocystis PCC6803 and named it pCRPB1010. The technique presented here using deadCas9 is easier to implement for gene silencing in Synechocystis PCC6803 than other existing techniques as it circumvents the genome integration and segregation steps thereby significantly shortens the construction of the mutant strains. We executed CRISPR interference against well characterized photosynthetic genes to get a clear phenotype to validate the potential of pCRPB1010 and presented the work as a “proof of concept”. Targeting the non-template strand of psbO gene resulted in decreased amount of PsbO and 50% decrease in oxygen evolution rate. Targeting the template strand of psbA2 and psbA3 genes encoding the D1 subunit of photosystem II (PSII) using a single spacer against the common sequence span of the two genes, resulted in full inhibition of both genes, complete abolition of D1 protein synthesis, complete loss of oxygen evolution as well as photoautotrophic growth arrest. This is the first report of a single plasmid based, completely lesion free and episomal expression and execution of CRISPR interference in Synechocystis PCC6803.

Keywords:

Sequence motif analysis – Genetic interference – antibiotics – Polymerase chain reaction – Plasmid construction – Oxygen – CRISPR – Synechocystis

Introduction

CRISPR (clustered regularly interspaced short palindromic repeats)-Cas9 (CRISPR associated protein 9) has revolutionized genome editing due to its simple execution and precision. The system was discovered in Streptococcus pyogenes and has primarily three working components: The ‘Cas9’ protein which has a single strand nuclease activity, a ‘tracrRNA’ complementary to the palindromic repeat sequences in the crRNA, which also forms RNA loop for Cas9 binding, and ‘crRNA’, which harbors palindromic repeats separated by ‘spacer’ sequences complementary to the targeted DNA sequence [1]. These three components are transcribed from a constitutive bi-directional promoter. Once transcribed, these two RNAs form a double stranded RNA with an RNA loop for Cas9 binding, which is then processed by cellular RNaseIII resulting in activated Cas9, which could screen the target genome and search for protospacer adjacent motif (PAM) sequences. Once such sequences are found lying next to those complementary to the crRNA, Cas9 will exert its nuclease activity [1]. This tri-component system was further simplified to a bi-component system where the tracrRNA and crRNA were combined as a single effector RNA named gRNA and Cas9 was placed under inducible promoters to control its expression for genome editing [2]. This system is widely adapted for both prokaryotic and eukaryotic gene editing [3].

An inactivated Cas9, termed as deadCas9 (dCas9), was developed which preserves the specific binding capacity to DNA targets of Cas9, with a complete loss of its nuclease activity [2]. This enzyme was able to perform transcriptional repression in bacteria in a process termed as CRISPR interference (CRISPRi) [4], which was then extensively characterized in E. coli [2, 3, 5] and being used for wide variety of purposes in genetic engineering and synthetic biology [6]. The basic working principle of this system’s application for transcriptional inactivation is that dCas9 will scan the genome until it finds a PAM, and if further pairing occurs between the crRNA and genomic DNA, it will bind strongly to the DNA, thereby preventing the transcription of the gene by hindering the binding and/or movement of RNA polymerase [3]. As there is no Dicer enzyme present in bacteria this RNA interference system swiftly gained high popularity [4, 7].

CRISPRi was further developed into two different systems in E. coli. ‘System one’ was developed by using the two component dCas9 and gRNA, which was shown to be highly versatile both in repressing the target genes and in tuning the CRISPRi system [3]. This system was shown to work from two plasmids containing gRNA and dCas9, respectively, that need to be transformed into the target bacterial cells. Alternatively, the gRNA and dCas9 cassettes can be integrated into bacterial genomes [8]. ‘System two’ of dCas9 relied on the original three component system of S. pyogenes CRISPR arrangement. The whole system could be executed in E. coli from a single plasmid by using the native S. pyogenes bi-directional promoter. This system was used as repressor and also as a transcriptional activator in E. coli [5]. The system was further developed by introducing a BsaI restriction site to facilitate easy cloning of multiple spacers. The resulting plasmid was called as pCRISPathBrick and was shown to be capable of multiplex repression of genes in E. coli [9]. The same system was also used to improve E. coli strain genome stability by silencing multiple insertion elements [10].

Synechocystis PCC6803 is an important model organism in scientific research. It was the first phototrophic organism to be fully sequenced offering an opportunity to study photosynthesis in unprecedented detail. It combines features from plants with features from microorganisms in scientifically attractive proportions: it holds a photosynthetic apparatus while being uncomplicated (simple, fast, versatile) to culture and to genetically transform. These and the available genome data are the main reasons to use Synechocystis PCC6803 in photosynthetic research. Accordingly, since its sequencing in 1996 about 160 papers are published annually with Synechocystis PCC6803 acting as the phototrophic model organism. The high number of studies with the same model organism facilitates meta-analysis of scientific data, it renders reproducibility across research groups more reliable and aids the standardization of specific scientific methods. Moreover, Synechocystis PCC6803 is also getting more and more attention in biotechnological applications [11], as well as in studies in systems biology [12]. However, because of the high and dynamically changing ploidy of cyanobacterial genome, this organism is a complex system to work with [13]. Accordingly, genome modification in Synechocystis PCC6803 requires a tedious process involving several steps of re-cultivation for complete segregation of the mutated chromosome. A number of attempts were made to edit cyanobacterial genomes using the CRISPR-Cas9 [14–16] and Cpf1 (also called Cas12) [17] genome editing machinery. However there were reports about toxicity of Cas9 proteins in cyanobacteria, which needs be avoided using tight regulation of Cas9 expression [18]. On the contrary, dCas9 does not affect cyanobacterial growth and it has been shown to allow the suppression of multiple genes, resulting in phenotypic changes in Synechocystis PCC6803 [8, 19, 20]. These studies were carried out using dCas9 in the above mentioned ‘system one’ methodology, in which the dCas9 and gRNA were integrated into the genome. This very step of generating the mutant is similar to the generation of knockout mutants using classical homologous recombination and further selection of mutants using chromosomal segregation [21], which is again a long process in itself to execute. Further, the resulting Synechocystis PCC6803 genome does not remain lesion free, rather it becomes an insertional mutant.

Here we developed a simple, fast and single plasmid-based strategy of CRISPR-dCas9 system following the above mentioned ‘system two’ approach. The resulting shuttle plasmid was created by replacing the original p15 origin of replication of pCRISPathBrick with a ‘conjugating origin of replication’ derived from RSF1010 and was named pCRPB1010. In this work, we demonstrate the functionality of this plasmid as “proof of concept” by showing suppression of three well characterized genes in Synechocystis PCC6803.

Through the episomal expression of CRISPR-dCas9, we were able to achieve significant repression of psbO and psbA (psbA2 and psbA3) genes that code for the manganese-stabilizing protein and for D1 reaction center subunit of PSII, respectively_. The rescue of the mutants was fast and straightforward as it did not necessitate chromosomal segregation step that is characteristic in the other methodology. The phenotypic changes resulting from the gene suppressions were in good agreement with earlier studies [22, 23]. With this work we report for the first time a simple method to repress multiple genes using a single plasmid containing dCas9 in Synechocystis PCC6803, and demonstrate the potential of the technique to be applied in further studies requiring gene silencing.

It is noteworthy that last year a similar system has been successfully developed in Nostoc sp. PCC 7120 (Anabaena sp. PCC 7120) [24], a filamentous cyanobacterium, the most prominent model organism of cyanobacterial cell differentiation and nitrogen fixation. While here we present a system consisting of the original S. pyogenes genes and the native bidirectional promoter, they used plasmid-based dCas9-dependent silencing using a single guide RNA with engineered promoters and optimized ribosome binding site.

Materials and methods

Bacterial cultures and media

Synechocystis PCC6803 was used as the wild type strain (from here: Synechocystis). Cells kept under identical conditions in light have identical chlorophyll (Chl) to protein ratio and also identical cellular chlorophyll content even in psbO inactivated mutants [25]. Therefore throughout the present work the cultures were normalized to identical chlorophyll concentration as we found it more reliable [26] and less prone to artefacts (e.g. precipitates in the medium, differences in cell shape, interfering chemicals) than light scattering and Bradford method, respectively. Nevertheless, the growth of the cultures was monitored by reading light scattering (OD₇₂₀) for simplicity. The cells were grown in BG11 medium under 40 μmol photons m^-2 s^-1 white light intensity, 3% CO₂ enriched atmosphere, in 30°C. For maintaining the pCRPB1010-containing strains, BG11 was supplemented with chloramphenicol (25 μg mL^-1). Solidified BG-11 agar plates were supplemented with 5 mM glucose when needed and the antibiotic concentrations and other conditions were kept as above. For experiments with WT and psbO repressed cells, Synechocystis cultures were grown to mid-log phase (OD₇₂₀ 0.8) without antibiotics for two days and measurements were performed from cell suspensions adjusted to 5μg Chl mL^-1. psbA repressed cells were maintained in respective antibiotics and 5 mM glucose. Escherichia coli DH5α was grown in Luria Broth supplemented with chloramphenicol (25 μg mL^-1) for the cells harboring pCRPB1010 and with kanamycin (50 μg mL^-1) for the cells containing pRK2013 [27].

Plasmids used in this study

Plasmid pdCas9 was a gift of Luciano Marraffini (Addgene plasmid # 46569). It contains the dCas9 ORF, a tracrRNA and a crRNA resembling the native S. pyogenes organization under the control of a constitutive bi-directional promoter [5]. The spacer array of pdCas9 was modified to enable directional cloning of multiple spacers using the BsaI restriction endonuclease, as described previously [9]. The resulting plasmid, pCRISPathBrick has been characterized elsewhere to be able to suppress multiple genes in E. coli successfully [9, 10]. These plasmids contain the p15A origin of replication and are therefore unable to replicate outside E. coli.

pPMQAK1 was a kind gift of Peter Lindblad. This is a shuttle vector [28] that has pRSF1010 origin of replication, which enables its transfer to cyanobacteria through conjugation. We amplified this region by PCR with respective restriction cut sites (see Table 1) and by replacing the p15 oriC in pCRISPathBrick we created pCRPB1010 (Fig 1). The spacers were designed using CLC-main workbench following the instructions published earlier [9], and ordered as complementary oligonucleotides of 66 bp each. The oligonucleotides were phosphorylated, annealed and then cloned into pCRPB1010 using BsaI-restricted plasmids, transformed into E. coli DH5α and selected on antibiotic plate. The clones were confirmed by PCR using the corresponding primers (Table 1). The PCR fragments were confirmed by restriction fragment lengths using multi-cutter restriction endonucleases. The resulting plasmid pCRPB1010 was 14.5 kb in size and was not possible to naturally transform into the Synechocystis cells. Therefore a bi-parental mating between E.coli containing pRK2013 (helper plasmid) and pCRPB1010 and Synechocystis PCC6803 was performed [29], relying on the conjugative origin of replication RSF1010, as follows.

Transformation of Synechocystis via biparental mating

The plasmids containing different spacers were chemically transformed into helper E. coli DH5α strain containing pRK2013. The cells containing dual plasmids were rescued on dual antibiotic plates (kanamycin and chloramphenicol). From overnight grown E. coli cells 0.1 mL culture was mixed with 1 mL of mid-log phase Synechocystis PCC6803. After centrifugation at 4500 rpm for 10 min at room temperature, 1 mL supernatant was discarded. The pellet was re-suspended into remaining medium and spotted on a BG11-containing agar plate with 5% LB. After two days the spot was scraped and spread on BG11 agar plates with chloramphenicol (25 μg mL^-1) and glucose (5 mM). After 5 to 7 days a green lawn appeared. The lawn was scraped into 20 mL BG11 medium containing chloramphenicol (and 5 mM glucose in case of targeting psbA genes). After 3 days the culture became dense and from the liquid culture dilution plating was done on selective medium to isolate a single Synechocystis colony. In another 5-days, single colonies appeared on the plate. One colony was inoculated into chloramphenicol-containing (25 μg mL^-1) liquid broth and further maintained in liquid culture and also stored at -80°C in the presence of 7.5% DMSO.

Designing spacer sequences

Spacers were designed according to the published protocol [9], using Synechocystis PCC6803 genomic sequence (GCA_000009725.1) (see Table 1). For psbO, there were two spacers targeting the non-template DNA strand inside the coding region. Spacer1 and spacer2 were designed to target sequences starting at 5 bp upstream and 60 bp downstream of the start codon, respectively. For psbA2 and psbA3 a single spacer was designed targeting a 100% identical sequence span starting 4 bp upstream of each start codon of the template strand. We show the design of the targeted DNA sequences in the S1 Fig.

The spacers were synthesized by Eurofins Genomics.

Biophysical measurements

Steady-state rates of oxygen evolution were measured at 30°C using a Hansatech DW2 oxygen electrode at a light intensity of 2000 μE m^-2s^-1in the presence of 0.5 mM 2,5-dimethy-p-benzo-quinone (DMBQ) as electron acceptor. 1 mL of cells at 5 μg Chl mL^-1 was used in each measurement and the temperature of the culture was kept at 30°C by a circulator bath.

PSII activity was assessed by measuring the changes of variable Chl fluorescence values. The flash-induced Chl fluorescence measurements were performed with an FL 3000 Fluorimeter (Photon Systems Instruments Ltd.), using 1 mL samples. Measurements were repeated on three biological replicates at preset time points (see Fig 1). The data were visualized and evaluated using the Fluorwin software and Origin.

Western blot

Cell suspensions were adjusted to 5 μg Chl mL^-1. 2 mL aliquots of the suspensions were pelleted and kept at -20°C for further processing. The cell pellets were dissolved in 3X SDS-PAGE loading buffer. The 6M urea-PAGE was run following the published protocol for cyanobacteria [30]. Immunoblotting was carried out by transferring the proteins on to nitrocellulose membrane for 3 h in a cold-room at 25 V. The membrane was blocked using 10% skimmed milk in TBS. Primary antibodies against PsbO and PsbA (D1) were obtained from Agrisera AB, Sweden. Primary antibodies raised in rabbit were added to skimmed milk at a dilution of 1:2000 and incubated for 2 hr at room temperature. Goat-anti-rabbit-ALPO secondary antibody conjugate was used in 1:5000 dilution and visualized using NBT-BCIP, scanned, and analyzed using Image J software to calculate the band intensity.

Quantitative PCR

15 mL cell suspension was added to 15 mL of ice-cold ethanol containing 5% phenol, mixed immediately and centrifuged at 10000 g for 10 min at 4°C. The supernatant was discarded and the pellet was resuspended in 0.5 mL of resuspension buffer (0.3 M sucrose, 10 mM NaOAc pH 4.5) and centrifuged again, resuspended in 0.25 mL of resuspension buffer and 37.5 μL of 0.5M EDTA was added. The cells were exposed to 4 freeze-thaw cycles and collected by centrifugation as above. Total RNA was isolated from the harvested cells using Direct-Zol RNA miniprep kit (Zymo Reseach). 500 ng of RNA was reverse transcribed using High-Capacity cDNA Reverse Transcription Kit (ThermoFisher) and the residual DNA was removed using TURBO DNA-free™ Kit (Invitrogen). cDNA corresponding to 10 ng RNA was used as template in the subsequent qPCR reactions. The primers used are listed in the primer list (Table 1).

Assessment of growth rate

Cells were grown at 30±C under the conditions described above in the presence of respective antibiotics to an approximate OD₇₂₀ of 2.0. For psbA repressed cells 5 μg glucose was also supplied. The cell cultures were then diluted to OD₇₂₀ 0.2 and kept under the same conditions and the OD₇₂₀ was measured every 24 hr.

Statistical analysis

Where applicable, mean values and standard deviations are shown. The significance of statistical differences was tested using t-test in GraphPad Prism software at p≤0.05 significance level.

Results and discussion

Metabolic engineering as well as exploration of protein functions and metabolic pathways often necessitates the generation of mutants deficient in certain genes or gene sets. Due to the polyploidy of the cells, creating deletion mutants in Synechocystis PCC6803 (from here: Synechocystis) is always a tedious and time-consuming process. Moreover, classical methods of creating mutants using homologous recombination have the limitation of altering one locus at a time. CRISPR-Cas9 has revolutionized genome editing and opens up the possibility to knock out multiple genes in a single shot. Earlier studies indicated Cas9 toxicity in cyanobacteria [18], which limits its utility in Synechocystis. On the other hand, a nuclease-inactivated Cas9 protein named dCas9 has no toxicity issues and CRISPR interference for transcriptional inactivation using this protein has been successfully implemented to downregulate multiple genes in Synechocystis [8, 19, 31]. However, this process was not lesion free to the genome and initial selection was tedious as it involves homologous recombination.

Creating pCRPB1010

pCrisPathBrick was a single plasmid developed to induce CRISPR interference in E. coli using an inactivated Cas9 called dead Cas9 (dCas9), which allows cloning multiple spacers and thereby results in multiplex gene repression in E. coli [9]. Since the p15A origin of replication in this vector does not function in cyanobacteria, we replaced it with the replication origin of a broad-host range plasmid pRSF1010 that can replicate in various cyanobacterial strains [32]. This promoter was amplified from pPMQAK1, an E. coli–Synechocystis shuttle vector [28] (Fig 1). The resulting recombinant conjugative plasmid pCRPB1010 can therefore replicate both in E. coli and Synechocystis with all necessary parts of the dCas9 machinery. This allowed us to test this plasmid for transcriptional inactivation of genes in Synechocystis. We have also optimized the original conjugation method [32]. The plasmid was transferred into Synechocystis by bi-parental mating using a mutant E. coli containing the helper plasmid pRK2013 [27]. The technique developed here circumvents the above mentioned issues and demonstrates (discussed below) that multiple genes can be downregulated by using a CRISPR interference system with episomal expression from a single plasmid in Synechocystis. The time needed for creating a mutant strain using this technique is about 20 working days (Table 2).

Targeting the psbO, psbA2 and psbA3 genes

After transferring pCRPB1010 into Synechocystis, the presence of pCRPB1010 inside the cell was confirmed by colony PCR using primers designed for RSF1010 and dCas9, respectively, resulting in 5.4 kb and 700 bp bands (Fig 2A). The “empty plasmid” does not differ in this respect from the ones with spacers.

The expression of dCas9 was validated using quantitative PCR, using trpA as internal control. The results show that (1) dCas9 is expressed in all mutant strains (including the one with empty plasmid) and (2) the expression of dCas9 (C_T = 26,2 ± 0,19) is as stable as the housekeeping gene (C_T = 24,5 ± 0,24) across the strains and the expression level is comparable to moderately expressed genes (Table 3). The expression of dCas9 proved that the native bidirectional constitutive promoter of S. pyogenes between the genes of the tracrRNA and dCas9 is active in Synechocystis, as this promoter controls the expression of the whole system.

Next, spacers targeting psbO and psbA genes were designed (detailed in Experimental Procedures). The first targeted gene, psbO is present in the genome in one single copy with no paralogs. It is a well characterized non-essential photosynthetic gene, the deletion of which results in characteristic phenotype change [22]. The non-template strand of psbO was targeted and two spacers binding to two different segments inside the coding region were designed (see S1 Fig). Either one, or both of the two spacers were cloned into pCRPB1010 in order to test if the second spacer can have additional silencing effect.

As a second test we aimed to assess the capacity of the system for targeting multiple genes, and wanted to see if a single spacer can concomitantly silence multiple genes. To this end a single spacer targeting an identical segment of both psbA2 and psbA3 was designed. psbA genes code for D1 protein of the second photosystem, a protein of the highest turnover rate under most conditions. This gene has three paralogs in the Synechocystis genome, one of which (psbA1) has a somewhat divergent DNA sequence and is hardly expressed under normal conditions [33]. The two other copies (psbA2 and psbA3) [34] code for identical protein sequences and their DNA sequence similarity is so high that it was possible to design a common spacer targeting both of them (see S1 Fig).

Contrary to the spacers targeting psbO, we targeted the template strand of psbA genes to understand if there is any difference between the two approaches, as there are contradictory reports in the literature about targeting a particular DNA strand to achieve higher transcriptional repression [35].

This psbA-specific spacer alone and with single and double psbO-specific spacers, was cloned into pCRPB1010. The latter two constructs were created to validate multiple gene targeting capacity of this plasmid. The insert lengths were confirmed using PCR following the protocol published earlier [9] (Fig 2B). Since biophysical measurements (see below) were carried out using cells kept without antibiotics for two days, the presence of the dCas9 gene (and hence the plasmid) was checked from cultures grown similarly. The plasmid remained stable without antibiotics inside the Synechocystis for two days (Fig 2C). It is also noteworthy that the plasmid pCRPRB1010 in Synechocystis was still present in the cells kept in light even after a month of serial re-culturing without antibiotic.

Effect of silencing on gene expression and on the amount of synthesized proteins

In order to see the effect of the dCas9 system on gene expression, cells from exponential growth phase were sampled and the expression of psbO, psbA2 and psbA3 were assessed using quantitative real-time PCR (qPCR), using trpA gene as an internal control. The expression levels were evaluated as the percent of the ones in the wild type strain (WT). Both, in the absence and presence of the psbA-targeting spacer, the single targeting of psbO (O, AO, respectively) resulted in very significant but not complete suppression of the expression level, and the double targeting of psbO gene (OO, AOO, respectively) showed significant improvement of the repression. On the contrary, the psbA targeting resulted in complete abolishment of the expression of both psbA genes (Fig 3A).

In order to see the effect of the dCas9 system on protein synthesis level, cells from exponential growth phase were sampled for western blot analysis. The cultures were adjusted to 5 μg/mL chlorophyll-a concentration and equal amounts were separated on SDS-PAGE. Western blot results showed a decrease in the intensity of bands in the presence of single and double spacers targeting psbO (Fig 3B). As an internal control, D1 protein was analyzed from the same samples and no change in expression was observed (Fig 3B).

When both psbA and psbO genes were targeted, western blot results showed that D1 band is absent in psbA-targeted strains (Fig 3C). In cells where only psbA was targeted, the PsbO amount was unchanged, whereas in combined targeting cases single and double spacers against psbO resulted in decreased PsbO band intensity. The changes in band intensities were in good agreement with the gene expression levels of the corresponding strains.

Phenotypic assessment of the silencing

The effect of gene silencing on the photosynthetic activity was assessed via biophysical characterization of the cultures, using oxygen evolution rate and fast chlorophyll fluorescence measurements. These measurements were carried out in Synechocystis cultures balanced to 5 μg/mL chlorophyll-a concentration. PsbO is a manganese stabilizing protein characterized earlier, which has a role in molecular oxygen evolution in Synechocystis and other photosynthetic organisms. Complete disruption of psbO results in about 70% of reduction in oxygen evolution rate as reported earlier [22]. In concert with these data we found that the oxygen evolution rate of the cultures containing the plasmid with both a single and double spacer against psbO was decreased by about 40% as compared to the wild type strain (Fig 4A). In case of D1 deletion mutants, previous studies showed that disruption of D1 prevents the synthesis of PSII and leads to complete abolition of oxygen evolution and photosynthetic growth [23, 36, 37]. In concert with these data, no oxygen evolution could be observed in case of strains containing spacers targeting psbA genes.

In order to gain a better understanding on the PSII activity of the psbO- and psbA-suppressed cells, fast chlorophyll fluorescence induction measurements were performed. Variable chlorophyll fluorescence, i.e. the rise of the fluorescence yield from the initial F_o level to the maximal F_m level assesses electron transport capacity through PSII. The psbO-suppressed cells showed somewhat decreased variable fluorescence induction compared to wild type and pCRPB1010-containing cells (Fig 4B), which in concordance with the observed decrease in oxygen evolution rate, indicates some inhibition of PSII activity. In psbA-silenced cell cultures the fast fluorescence transients showed drastic changes (Fig 4C) in agreement with the absence of D1 protein and the completely ceased oxygen evolution.

In spite of the above shown decrease in photosynthetic activity of psbO-suppressed cultures, we did not observe significant difference between their growth rates. On the other hand, psbA-suppressed strains could only be maintained with 5 mM glucose added and did not grow in photoautotrophic conditions (Fig 4D), which was in agreement with previously published observations [37]. This result further corroborates the finding of the protein analysis above, showing the absence of D1 protein and hence complete inactivation of PSII caused by the activity of dCas9 expressed from a single plasmid inside Synechocystis.

Previous studies reported 50–95% repression of gene expression in Synechocystis using dCas9 [8, 19]. In those investigations, the genome integration strategy was used and the non-template strands of the genes were targeted. Here we tested the targeting of each strand. We found significant but not full repression of the targeted gene with non-template strand targeting, in agreement with the above publications. However, when the template strand was targeted, a complete loss of function was observed. Whether this complete inactivation of the photosynthetic apparatus was the consequence of the choice of the template strand targeting, is hard to tell at the current point without further investigation. Nevertheless, it is obvious that using this targeting method it is possible to achieve very strong repression in Synechocystis for complete loss of function of two genes using one single spacer.

In conclusion, the work presented here is the first report of an easy and single-plasmid based execution of CRISPR interference using dCas9 in Synechocystis, which is capable of concomitantly downregulating multiple genes in a lesion-free manner. This is a first report showing that the native CRISPR-array from S. pyogenes can also function well in Synechocystis with its fully functional bidirectional promoter between the genes of the tracrRNA and dCas9 and also the first report where dCas9 was used to modulate the photosynthetic machinery in Synechocystis.

Our findings are in good agreement with the results of Higo et al. [24] in Nostoc sp. PCC7120. Taken together this technique will facilitate and speed up the molecular level investigations and genetic engineering of these organism as well as other cyanobacterial strains.

Supporting information

S1 Fig [a]
Targeted genomic sequences of and genes.