Assembly CRISPR-Cpf1 in a pSEVA
To improve CRISPR tools in cyanobacteria, the CRISPR-Cpf1 cassette from pSL2680 was transferred to a new and more versatile vector. For potential candidates, we chose vectors from the SEVA repository as this collection is particularly interesting owing to the modular and interchangeable structure of their plasmids (32, 33). The Standard European Vector Architecture (SEVA, http://seva.cnb.csic.es) platform is a large repository of plasmids that are formed by three modules: an antibiotic marker, a replication origin and a cargo. These modules are separated by three permanent regions: an origin of transfer region (oriT) and two transcriptional terminators designated as T1 and T0 (Figure 1A) (32, 34). The SEVA platform uses a simple plasmid design that facilitates the swapping of functional modules and the extension of genome-engineering options beyond common bacterial laboratory strains (34). Also, the SEVA vectors used here are relatively small (5100-5200 bp pSEVA351-451 respectively) compared to other vectors used in cyanobacteria, such as pPMQAK1 (8372 bp) (35), pSCB (6592 bp) (36) or pANS based vectors (7842 bp) (37), making transference easier.
SEVA vectors have been successfully used in CRISPR editing processes in gram-negative bacteria (38-40) but they have not been previously tried in gene-editing processes in cyanobacteria. So far it has been reported that in one strain, Synechocystis 6803, SEVA vectors containing RSF1010 or RK2 origins can be successfully transformed by natural transformation, electroporation and conjugation to express heterologous genes (41, 42). The potential use of SEVA vectors for gene editing in cyanobacteria is therefore highly promising but it will be necessary to prove first that they can be used for transformation processes in more than one cyanobacterial strain. The SEVA vectors chosen for this study contained the RSF1010 origin that has been reported to work well on Synechocystis and other cyanobacteria and that have a higher copy number than those of RK2 origin (Table 1) (41, 42).
Table 1. Plasmids used in this study
Plasmid
|
Description
|
Reference
|
pSL2680
|
12684 bp, RSF1010 ori, kanamycin resistance (KmR), CRISPR (cpf1 and CRISPR array)
|
(28)
|
pSEVA351
|
5120 bp, RSF1010 ori, chloramphenicol resistance (CmR)
|
(34)
|
pSEVA451
|
5334 bp, RSF1010 ori, spectinomycin resistance (SptR)
|
(34)
|
pSEVA351-Cpf1
|
pSEVA351 with the CRISPR machinery from pSL2680.
|
This study
|
pSEVACpf1nblA
|
pSEVA351Cpf1 with the CRISPR array targeting nblA and with a homologous template for making a knock-out
|
This study
|
pSEVA451-Cpf1
|
pSEVA451 with the CRISPR machinery from pSL2680.
|
This study
|
pRK2013
|
Conjugative plasmid KmR, provide tra genes and nicking function from RK2
|
(43)
|
pRL623
|
Helper plasmid CmR, carries the genes for M.AvaI, M.Eco47II and M.EcoT22I and mob gene from ColK
|
(44)
|
Then, we evaluated the use of SEVA plasmids together with pSL2680 vector in three phylogenetically diverse cyanobacterial strains with different morphologies, growth and metabolic characteristics, which are important for its potential in biotechnological applications or basic research (Figure 2A). The unicellular Synechocystis 6803 was the first photosynthetic organism to have its genome sequenced. It rapidly became a model strain, being widely used in biotechnology as a photoautotrophic chassis. The second strain chosen was Anabaena 7120, a filamentous cyanobacterium that is interesting due to its capability to fix atmospheric nitrogen using the specialized cells called heterocyst when growing in a medium lacking combined nitrogen. Finally, we also chose a coccoidal cyanobacteria species of the genus Chroococcidiopsis. This genus includes extremophile strains that are a reference in the study of the resistance to desiccation, irradiation and DNA repair in cyanobacteria, among other extreme conditions (45-47). Concretely for our study, we used Chroococcidiopsis sp. B13, a strain isolated from a solar panel in Spain that can resist desiccation and UV-C exposure (laboratory collection, manuscript in preparation).
Vectors were transferred into the different strains by triparental mating. E. coli HB101 strain bearing the conjugative plasmid pRK2013 was used as conjugative strain and HB101 harbouring the pSEVA vector was the cargo strain. pSEVA351 was used in Synechocystis 6803 and Chroococcidiopsis B13. To avoid antibiotic incompatibilities for Anabaena conjugation, as it needs the helper plasmid pRL623 (CmR), pSEVA451 was alternatively used for this strain. As shown in Figure 2B, all the conjugations yielded colonies after 1-2 weeks. Transformation was confirmed by the acquisition of resistance to the respective antibiotic and the presence of the plasmid was verified by PCR (Figure 2C). pSL2680 has been used so far for genome editing in Synechococcus 11801, Synechococcus 2973, Synechocystis 6803 and Anabaena 7120 (27-29) but this is the first successful attempt to transform this vector into a Chroococcidiopsis strain.
Furthermore, RSF1010 pSEVA plasmids were successfully functional in all the cyanobacteria tested (Figure 2). Whereas most of the genetic tools for cyanobacteria reported are frequently focused on Synechococcus or Synechocystis (18), the SEVA vectors described in this work may broaden the potential applications for other strains, such as Anabaena and Chroococcidiopsis tested here and, by extension, to other cyanobacteria.
Next, we checked if SEVA vectors containing the CRISPR-Cpf1 endonuclease and the CRISPR array could still be transferable. To construct the pSEVA351-Cpf1 and pSEVA451-Cpf1 plasmids, the whole CRISPR system from pSL2680 was PCR-amplified and cloned into pSEVA451 or pSEVA351 (Figure 1). The CRISPR array kept the same characteristics as the one in the original pSL2680, where the gRNA sequence targeting the gene of interest could be cloned in AarI sites, replacing the lacZα gene. Figure 2 showed that pSEVA-Cpf1 plasmids could be transformed successfully into the three chosen cyanobacteria genera: Synechocystis, Chroococcidiopsis and Anabaena. As has been previously described for Synechococcus 2973 (28), no toxicity was observed in the presence of Cpf1 nuclease in all the strains tested.
Some genome editing vectors based on both Cas9 and Cpf1 have been built for Anabaena (26, 28, 48, 49). On the other hand, Chroococcidiopsis is an extremophile strain for which just a few references relating to its genetic modification can be found but none related to CRISPR editing (50, 51). This strain is becoming more and more relevant as it has been recently used in European Space Agency experiments due to its high resilience properties (45, 46). Therefore, this SEVA vector opens a door to genome editing in this strain and it could be a valuable tool for researching the resistance properties to different stresses that this strain displays.
Natural transformation of Synechocystis 6803 using SEVA-derived vectors
SEVA vectors containing the RSF1010 broad-host-range replicon, but harbouring different antibiotic markers (kanamycin, pSEVA251; chloramphenicol, pSEVA351; and spectinomycin/streptomycin, pSEVA451), have been successfully transformed in Synechocystis 6803 by natural transformation (a process in which bacteria actively take up and maintain extracellular DNA) (41). As protocols for natural transformation are generally simple and straightforward, the use of this technique in editing processes could facilitate the rapid employment of organisms capable of natural transformation in Biotechnology (52-54). From the three strains chosen in this work, only Synechocystis 6803 is able to be transformed by natural transformation, and therefore, we evaluated if the SEVA-Cpf1 vector could be also useful for natural transformation in this cyanobacterium.
We tested the transfer of pSEVA351-Cpf1 into Synechocystis 6803, using the protocol described by Sebesta et al. (2019) (55). pSEVA351 as a positive control and pSL2680, which contains the CRISPR machinery (like pSEVA-Cpf1), were also included in the study for comparison reasons. After a week of growing on BG11 with appropriate antibiotics, transformants were obtained only when pSEVA351 and pSEVA351-Cpf1 vectors were used (Figure 3A). Cultures of different colonies were made in BG11 with fresh antibiotics and the presence of the plasmid was verified by PCR (Figure 3B). This experiment confirms that SEVA-Cpf1 vectors can be also used for natural transformation. On the other hand, no growth was observed in cultures transformed with pSL2680 within two weeks after transformation (data not shown). Therefore, the use of pSL2680 seems to be inconvenient for natural transformation and maybe for this reason it is advised to be used in conjugation processes instead (Addgene #85581, supplemental material) (28).
Genomic Editing of Synechocystis 6803 using pSEVA351-Cpf1: nblA gene deletion as a proof of concept
To validate the pSEVA351-Cpf1 system for genome editing processes, Synechocystis 6803, a cyanobacterium of industrial importance, was chosen. The genomic target selected was the deletion of the nblA gene coding for the non-bleaching protein A. This mutation confers an easily observable phenotype since it prevents depigmentation developed under nitrogen deprivation due to phycobilisome degradation (20). WT phenotype bleaches under lack of nitrogen by the degradation of antenna complexes, while nblA mutants remain green in these conditions. For these reasons, the nblA gene knock-out has often been used as a reporter of editing processes in cyanobacteria (2, 20, 28).
Synechocystis 6803 genome contains two adjacent copies of the nblA gene (Figure 4A). To introduce the deletion, a synthetic template for gRNA was designed to bind between the two copies of nblA in a similar way as described by Ungerer and Pakrasi (2016) (28). The directed break is subsequently repaired by double homologous recombination using a repair template cloned right after the CRISPR array (Figure S2). Natural transformation was used to introduce the pSEVACpf1nblA into Synechocystis 6803. The colonies obtained after transformation (Figure 4B) were submitted to three rounds of repatching onto BG11 Cm10 before analysing them by PCR (Figure 4C). This step seems to be necessary due to the high degree of ploidy and the segregation process of Synechocystis. After the last patching, all the mutants selected showed a segregated selection of the nblA1/2 gene (Figure 4C). In parallel, we transferred the pSEVACpf1nblA by triparental mating, following the procedure used by Ungerer and Pakrasi (2016) (28) (Figure S2). Colonies appeared after 7 days and they were also patched three times onto BG11 with Cm. PCR confirmed the deletion segregation in these mutants (Figure S2B).
We proceeded to test the non-bleaching phenotype of the nblA mutant’s deletion obtained after 3 patches on selective media. To further verify segregation, we performed bleaching tests using wild-type and the selected ΔnblA Synechocystis 6803. After 10 days of additional growth, qualitative differences in colour were registered as shown in Figure 4D (natural transformation) and Figure S2C (triparental mating). The mutant colonies of Synechocystis 6803 did not display the colour of the WT that was bleached upon removal of nitrate, indicating that they were fully segregated for the nblA deletion. Therefore, pSEVA-Cpf1 is useful for editing processes on Synechocystis and it could also be useful for editing processes on other cyanobacteria.
We have proved that pSEVA-Cpf1 vectors can be used in the transformation of several genera of cyanobacteria. Using pSEVA-Cpf1, we successfully edited the Synechocystis 6803 genome by deleting the two copies of nblA. Moreover, the pSL2680 failed when using natural transformation but not the pSEVA vectors, which could be assimilated by this procedure even when the genetic editing of Synechocystis 6803 was carried out. Apart from Synechocystis 6803, in this work we have shown for the first time the replication of pSEVA-Cpf1 vectors in two cyanobacterial species different from those previously reported: Anabaena 7120 and Chroococcidiopsis sp. These results demonstrate that pSEVA and derived vectors are a promising tool for the transformation of other biotechnological relevant cyanobacteria, for instance extremophile bacteria. Considering the SEVA design, our plasmids would broaden the scope of cyanobacteria genera able to be CRISPR edited. Additionally, SEVA vectors could be also the basis for further development of CRISPR technology, e.g. they could be modified to include inducible CRISPRi gene repression elements.
What is more, if no CRISPR-mediated gene editing were pursued, considering the modular architecture of SEVA vectors, different combinations of promoters, gene reporters or the chance to clone any desired gene into the SEVA backbone would provide a source of multiple possibilities for obtaining genetically modified cyanobacteria. Additionally, as pSEVA has shown to be functional in other bacteria, these plasmids could be transferred to different microorganisms, increasing the synthetic biology tools available.
However, there are still future challenges that must be considered. For example, in order to allow more than one round of genome editing, the editing plasmid must be cured unless the new editing vector must contain a different selection marker. To remove the plasmids, cells are grown in an antibiotic-free medium for many generations. Recently Niu et al. (2019) (26) described the use of the counter selection marker sacB, which encodes the Bacillus subtilis levansucrase (56) for curing the pCpf1 editing vector (a pSL2680-derived plasmid) in Anabaena. In the presence of 5% sucrose, sacB expression is lethal and therefore, survival depends on recombination events or alternatively on the spontaneous loss of the sacB-containing vector, facilitating the curing of the editing plasmids from cyanobacteria cells (26, 57). Due to the modular backbone of SEVA vectors, these developments can easily be obtained in the near future.