Genome Engineering by RNA-Guided Transposition for Anabaena sp. PCC 7120

In genome engineering, the integration of incoming DNA has been dependent on enzymes produced by dividing cells, which has been a bottleneck toward increasing DNA insertion frequencies and accuracy. Recently, RNA-guided transposition with CRISPR-associated transposase (CAST) was reported as highly effective and specific in Escherichia coli. Here, we developed Golden Gate vectors to test CAST in filamentous cyanobacteria and to show that it is effective in Anabaena sp. strain PCC 7120. The comparatively large plasmids containing CAST and the engineered transposon were successfully transferred into Anabaena via conjugation using either suicide or replicative plasmids. Single guide (sg) RNA encoding the leading but not the reverse complement strand of the target were effective with the protospacer-associated motif (PAM) sequence included in the sgRNA. In four out of six cases analyzed over two distinct target loci, the insertion site was exactly 63 bases after the PAM. CAST on a replicating plasmid was toxic, which could be used to cure the plasmid. In all six cases analyzed, only the transposon cargo defined by the sequence ranging from left and right elements was inserted at the target loci; therefore, RNA-guided transposition resulted from cut and paste. No endogenous transposons were remobilized by exposure to CAST enzymes. This work is foundational for genome editing by RNA-guided transposition in filamentous cyanobacteria, whether in culture or in complex communities.

Supplemental Figure S1 Relative toxicity of the CAST plasmids in Anabaena.Exconjugant strains derived from CSVT15 generated by conjugation contained the following constructs: pAzUT.3, with only the YFP and erythromycin-resistance cassettes expressed constitutively; pAzUT.7 with, in addition to the components of pAzUT.3, the P glnA -driven cas12k expression induced under nitrogen deprivation, and the strongly and constitutively expressed sgRNA scaffold (without the target specific sequence); pAzUT.9, with P glnA -driven expression of all the CAST genes (tnsB, tnsC, tniQ, and cas12k), the cargo transposon flanked by the LE and RE, and the strongly and constitutively expressed sgRNA-scaffold without a target sequence.The strain carrying pAzUT.14*,used as a control, was homozygous for the cargo transposon inserted into the gfp and was cured of the pAzUT.14plasmid that contained, in addition to what is encoded by pAzUT.9, the sgRNA targeting sense sequence of gfp.Strains with the plasmids and the controls, the parental strain and pAzUT.14 were grown in successive dilutions of starting material, measured in chlorophyll a equivalents (Chl in ng), for 8, 12 and 28 days on either Streptomycin/ Spectinomycin (Sm/Sp) or Sm/Sp and Erythromycin (Em) antibiotic selection in BG11 medium (with nitrogen) and BG11 0 medium (without combined nitrogen).

GTT Sense
Supplemental Figure S2 The three different sgRNAs targeting gfp.A, Schematic representation of the coding region of gfp-mut2 showing the base at which the different target-specific sequences start that are encoded by the sgRNA.For example, the target specific sequence of the sgRNA encoded in pAzUT.14starts 163 bp inside the 717 bp-long gfp-mut2 sequence with the GTT PAM sequence and encodes the sense strand.B, Primers used for cloning of the target specific sequences into the LguI restriction site of the scaffold sgRNA from pAzU1.3 (depicted in Fig. 1).Forward (Fwd) and reverse (Rev) primers were annealed, then in one pot with pAzU1.3,digested and ligated to obtain the target specific sgRNA from the pAzUT.10,12 and 14 as indicated.LguI restriction site (red), orientation specific overhangs (green), PAM (blue).Supplemental Figure S4 Rapid conjugation protocol for RNA-guided transposition using the suicide plasmid pAzUT.17.pAzUT.17 is identical to pAzUT.14 except in the plasmid backbone region where the cyanobacterial replication elements were removed.In the rapid conjugation protocol with the parental strain CSV15, material spread in a filter on BG11 0 medium was transferred to BG11 0 medium during the first 48 h after of conjugation.Thereafter, transconjugants were selected on BG11 medium supplemented with erythromycin.A, Scheme of the gfp after insertion of RNA-guided cargo transposon and the PCR used to detect the insertion.B, PCR detection of the insertion at the LE of the cargo transposon.M, DNA molecular weight markers; 1-4, transconjugants with pAzUT.17derived from CSV15 ; P, parental strain, CSVT15 (amt1::gfp).In UU1-1 and UU1-8, reads were cut sharply at several locations of the plasmid sequences that were inserted when CSVT15 was generated by single cross-over recombination, these are the deletions for indel 7. Penetrance of these events were weak, however.The alignment of the reads is cut sharply and the location of the cargo transposon insertion inside the GFP sequence with 100% penetrance in UU1-1,4 and 8 but not in the parental, this is the true positive insertion from indel 8.
targeting at locus alr3727 in wild-type Anabaena.A, PCR strategy to detect RNA-guided transposition of the cargo transposon encoded in pAzUT.18into the locus alr3727 of Anabaena sp.PCC 7120.(B) Fragments amplified from transconjugant clones obtained using a rapid process as follows: conjugation, subsequent transfers for 48 h on BG11 0 medium, followed by sonication and then selection on BG11 0 medium with erythromycin.1-10, randomly selected transconjugant clones; WT, Anabaena sp.PCC 7120; W, water (negative control); M, DNA molecular weight standards.
Visualization in IGV of the chromosome loci for Indels 1 (contig_1_86_Sniffles2_INS_0M4) and 2 (contig_1_4114_Sniffles2_INS_1M4).These insertions are at the border of the chromosome and have no distinct border, therefore, they are a likely result of assembly issues at the end of the chromosome. .The v-shape lack of coverage for this deletion in all strains sequenced suggests a problem with assembly of the chromosome in the parental strain, possibly due to a region difficult to sequence.False positive indel.SupplementalFigure S10 Visualization in IGV of the amt1::gfp locus for indels 7 (contig_1_5088816_Sniffles2_DEL_AM4) and 8 (contig_1_5089223_Sniffles2_INS_9M4).
Supplemental Figure S12 Visualization in IGV of the locus for indel 10 (contig_3_26226_Sniffles2_INS_0M1), typical of the indels detected by Sniffles 2 in the plasmid sequences.The indels in plasmids more often than not were located in repetitive sequences encoding transposase.
Supplemental TableS2CASTGATE vectors transferred to and tested in Anabaena wild-type, and the CSVT15 and CSAM137 strains in this study.CSVT15 had been generated by single cross-over recombination so as to generate the amt1::gfp fusion (Merino-Puerto et al., 2010) and grew on BG11 0 medium without nitrogen.Similarly, CSAM137 had been generated by recombination so as to generate the sepJ::gfp fusion(Flores et al., 2007); CSAM137 grew only slowly if at al on the BG11 0 medium.
Supplemental TableS4Primers used for PCR assays and key cloning steps in this study.