Materials and Methods

Mobile element warfare via CRISPR and anti-CRISPR inPseudomonas aeruginosa

MATERIALS AND METHODSMicrobesCell culturing Pseudomonas aeruginosa strains (PAO1, PA14 and PA4386) and Escherichia coli strains (DH5a) were cultured using lysogeny broth (LB) agar or liquid media at 37 °C supplemented with gentamicin, where applicable, to maintain pHERD30T (50 μg/ml for P. aeruginosa, 30 μg/ml for E. coli). In all P. aeruginosa experiments, expression of genes of interest in pHERD30T was induced using 0.1 % arabinose.Type I-C CRISPR–Cas expression in PAO1 PAO1IC activity was induced using 1 mM IPTG. Construction of this strain is described (27) and may be referred to as LL77 (Targeting crRNA) or LL76 (non-targeting).Bacterial transformations P. aeruginosa transformations were performed using standard electroporation protocols (27). Briefly, overnight cultures were washed twice in an equal volume of 10 % glycerol and the washed pellet was concentrated tenfold in 10% glycerol. These electrocompetent cells were transformed with 20–200 ng plasmid, incubated shaking in LB for 1 h at 37 °C, plated on LB agar with appropriate selection, and incubated overnight at 37 °C. Bacterial transformations for cloning were performed using E. coli DH5α (NEB) according to the manufacturer's instructionsCRISPRi CRISPR interference transcriptional repression assays were conducted as in previous work (25). A Δcas3 strain was lysogenized with a DMS3m phage encoding an Acr of interest. This lysogen was transformed with a plasmid encoding a crRNA targeting the phzM promoter. The crRNA and cas genes (in the case of Type I-C) were induced in overnight cultures with 0.25 mM IPTG and 0.05% arabinose. Pyocyanin levels were measured using an acid extraction protocol described previously (25). Pyocyanin quantification was normalized to a strain encoding AcrIIA4, which inhibits Cas9, but not the Type I CRISPR–Cas systems included in this study, resulting in cultures lacking pyocyanin.PhagesPhage maintenance Pseudomonas aeruginosa DMS3m-like phages (including JBD30 and DMS3m engineered phages) were amplified on PA14 ΔCRISPR, PAO1, or PA4386 Δcas3 and stored in SM buffer at 4 °C.Construction of recombinant DMS3m acr phages To generate the isogenic panel of DMS3m and JBD30 anti-CRISPR phages, recombination cassettes were generated with up- and down-stream overhangs to aca1 and the acr promoter flanking the Acr of interest, as previously described (28). These genes were ordered from TWIST or IDT and were assembled into plasmids using Gibson assembly methods. Recombinant phages were generated by infecting cells transformed with the donor constructs and phages were isolated and assessed for resistance to CRISPR–Cas targeting. The presence of the anti-CRISPR gene was confirmed by PCR. To generate the virulent phages used for liquid growth curve assays, the dms3m c-repressor gene, gp1, was mutated using plasmids described previously (28).Plaque forming unit quantification Phage plaque forming units (PFU) were quantified by mixing 10 μl of phage with 150 μl of an overnight bacterial culture. The infected cells were aliquoted into 3 ml molten 0.7 % top agar and spread on an LB agar plate supplemented with 10 mM MgSO4 and appropriate inducers. After 18 hours of growth at 30 or 37 °C, individual plaques were counted. Three biological replicates were done per phage per strain.Phage spot assays 3 ml of molten 0.7 % top agar mixed with 150 μl of bacteria were spread on an LB agar plate supplemented with 10 mM MgSO4 to grow a bacterial lawn. Ten-fold serial dilutions of phage were made in SM buffer and 2 μl of each dilution was spotted on the lawn. Plates were incubated at 30 or 37 °C for 16 h and imaged.Efficiency of plaquing (EOP) EOP was calculated as the ratio of the number of plaque forming units (PFUs) that formed on a targeting strain of bacteria (PAO1IC, PA14 WT, PA4386 WT, PaLML1 plus crRNA plasmid) divided by the number of PFUs that formed on a related non-targeting strain (PAO1, PA14 ΔCRISPR, PA4386 ΔCRISPR, PaLML1 plus NT crRNA). Each PFU measurement was performed in biological triplicate. EOP data are displayed as the mean EOP ± standard deviation.Escaper phage isolation High titer phage preparations were mixed with overnight cultures and spread on an agar plate with top agar. Single plaques that formed after overnight propagation were picked with a sterile pipette tip and resuspended in SM buffer. This process was repeated two times under maintained targeting pressure. The escaper phages were ultimately titered and the protospacer region sequenced.Liquid culture phage infections PAO1IC was transformed with plasmids encoding either the Type I-C or Type I-F systems from PaLML1 plus one non-DMS3m targeting spacer (‘decoy’ surveillance complexes) to determine the effect of CRISPR–Cas system co-expression. A separate Type I-F plasmid with a Cas8 mutation (K247A) was also constructed. P. aeruginosa strains were grown overnight and diluted 100x in LB supplemented with 10 mM MgSO4, gentamicin, 1 mM IPTG, and 0.1 % arabinose. 140 μl of bacterial culture was infected with 10 μl of serially diluted virulent phage in a 96-well plate. Growth and infection was monitored for 20 h using the Synergy H1 microplate reader (BioTek) at shaking at 37°C. Phage was extracted after 20 h by mixing 100 μl of culture from each well with 20 μl chloroform, shaking at RT for 20 min, and centrifugation at 14 000 × g for 2 min.BioinformaticsNumerical data were analyzed in Excel and plotted in GraphPad Prism 6.0.Discovery of acr genes using aca1 and aca4 Anti-CRISPR searches were done as previously described (27).CRISPR array spacer analysis Spacers were derived from the van Belkum dataset (11) (18 genomes with 12 non-redundant arrays) or from Type I-C containing strains found using BLAST and CRISPRFinder (29) (12 non-redundant arrays). Spacers were analyzed using CRISPRTarget (30) using the Genbank-environmental, RefSeq-plasmid, IMG/VR, and PHAST databases.PAM analysis was done using the Berkeley Web Logo tool by submitting the upstream and downstream regions flanking the protospacer sequence. These eight nucleotide long flanking sequences are part of the CRISPRTarget output. Every matching protospacer (low cutoff of 20, no redundant matches removed) was utilized for the PAM analysis for n = 4443.To determine the types of elements targeted by the spacers in our collection, the cut-off score was increased to 30 and a PAM match score of +5 was used to narrow the total number of hits to matching elements. If a spacer had multiple matches, the match with the highest score was selected as the representative for that spacer OR the match to a phage genome. Only one match was considered per spacer. This reduced the number of spacers to 131.Matches were placed into the following categories: Myophages, Siphophages, Podophages, plasmids and assorted prophages. A hit was placed into a phage family, rather than into the prophage category, if the CRISPRTarget output included a link to a specific phage genome. Importantly, this means that being placed into a phage family does not mean that a phage is strictly lytic. Prophages were identified by considering the genes in the protospacer neighborhood.Lineage tracing For the Type I-C encoding strains from this study, WGS reads were imported from NCBI to Benchling, and the repeats were annotated using the Benchling annotation tool. Individual spacers were extracted using the CRISPRCasFinder tool by copying the entire CRISPR array region. Each spacer sequence was assigned a number, such that identical spacers in distinct strains were assigned the same number, allowing the visualization of spacer similarity across different strains. Lineages were manually curated using the 18 published CRISPR arrays (9) and the additional 12 CRISPR arrays found in this study.Anti-CRISPR phylogenetic tree generation BLAST was used to generate the tree of AcrIC5 relatives. The following parameters were selected. Tree method: fast minimum evolution. Max seq difference: 0.85. Distance: Grishin (protein).

Article TitleMobile element warfare via CRISPR and anti-CRISPR inPseudomonas aeruginosa


Anti-CRISPR and Aca protein NCBI protein accession codes are as follows: AcrIC1 ({"type":"entrez-protein","attrs":{"text":"WP_046701304.1","term_id":"818919851","term_text":"WP_046701304.1"}}WP_046701304.1), AcrIF2 ({"type":"entrez-protein","attrs":{"text":"WP_015972868.1","term_id":"506505375","term_text":"WP_015972868.1"}}WP_015972868.1), AcrIC3 ({"type":"entrez-protein","attrs":{"text":"WP_058130594.1","term_id":"959842971","term_text":"WP_058130594.1"}}WP_058130594.1), AcrIC4 ({"type":"entrez-protein","attrs":{"text":"WP_153575361.1","term_id":"1774266732","term_text":"WP_153575361.1"}}WP_153575361.1), AcrIC5 ({"type":"entrez-protein","attrs":{"text":"WP_089394111.1","term_id":"1219378944","term_text":"WP_089394111.1"}}WP_089394111.1), AcrIC6 ({"type":"entrez-protein","attrs":{"text":"WP_080050315.1","term_id":"1167616571","term_text":"WP_080050315.1"}}WP_080050315.1), AcrIC7 ({"type":"entrez-protein","attrs":{"text":"WP_003294373.1","term_id":"489387775","term_text":"WP_003294373.1"}}WP_003294373.1), AcrIC8 ({"type":"entrez-protein","attrs":{"text":"WP_074202337.1","term_id":"1123470859","term_text":"WP_074202337.1"}}WP_074202337.1), Aca10 ({"type":"entrez-protein","attrs":{"text":"WP_074980464.1","term_id":"1124702480","term_text":"WP_074980464.1"}}WP_074980464.1), AcrIE9 ({"type":"entrez-protein","attrs":{"text":"WP_101192668.1","term_id":"1311629338","term_text":"WP_101192668.1"}}WP_101192668.1). Sequences can be accessed at

Login or Signup to leave a comment
Find your community. Ask questions. Science is better when we troubleshoot together.
Find your community. Ask questions. Science is better when we troubleshoot together.

Have a question?

Contact or check out our support page.