VIVO PLASMID EDITING AND FIRST ATTEMPTS OF BACTERIAL PATHOGENS GENOME ENGINEERING BY SH95 I-F1CRISPR-CAS SYSTEM

MARÍA CAROLINA MOLINA; CECILIA QUIROGA

Chapadmalal

Congreso; XVIII SAMIGE; 2023

SAMIGE

The ability to genetically manipulate microorganisms enabled to advance thefundamental understanding in biology fields and the development of novel approaches andtherapeutics. In particular, CRISPR-Cas systems outperformed previous technologies dueto their simplicity of programmability, enabling its worldwide application. Here, we tested theprogrammability of a I-F1 CRISPR-Cas system from Shewanella xiamenensis Sh95 andtransferred its interference machinery (Sh95 I-F1CRISPR-Cas) to different model and clinicalstrains. Plasmids coding the nuclease (pFL) or without (pC14) were constructed bytraditional cloning. Mini CRISPR arrays were designed (anti-gfp-1, control guides). Guideswere generated by oligo annealing restriction and ligation (anti-gfp-2, anti-blaTEM, anti-blaKPC, anti-csgD). We used plasmid loss assays (PLA) to quantify the interferenceefficiency as a ratio of colonies recovered in Selective (S) vs Non-Selective (NS) plates (logCFU/mL). Controls, E. coli BL21(DE3) with pFL or pC14 expressing the control guide,showed NS/S=1.4. The efficiency of transformation dropped drastically when targetingAmpR of high-copy number plasmids with ~7 logs reduction and NS/S=23.75. Minimalinhibitory concentration assays (MIC) of survivor colonies revealed reversion to AmpS (2/5)after an ON culture in absence of antibiotic. Otherwise, PLA with anti-gfp showedNS/S=2.34 and 2.64. Survivor colonies recovered were AmpR and emitted no fluorescenceintensity. Unexpectedly, different plasmid editing outcomes were recovered: insertions of ~200 bp and deletions from 243 up to ~4200 bp. The pattern of deletions was bidirectionalfrom the target site and these colonies harbored smaller versions of the target plasmidmolecules. Homogeneous outcomes as well as mixed populations were detected by flowcytometry. We propose a model of 4 possible scenarios: i) Total plasmid degradation; ii)Evasion; iii) Recovery, and in vivo plasmid editing; iv) Coexistence, balance betweenplasmid replication and degradation rates. Next, we performed PLA in an attempt to quantifythe efficiency of Sh95 I-F1CRISPR-Cas against blaKPC in clinical multidrug-resistant isolates:Enterobacter cloacae, Providencia stuartii, and Klebsiella pneumoniae. Ratios (NS/S) didnot show differences. This suggests that the current methodology is not suitable to addressthis question in such strains. Additionally, anti-csgD guide was tested in E. coli BL21(DE3),E. coli K12 MG1655, and a clinical isolate Salmonella enterica. In E. coli, targeting achromosomal gene resulted in cell death, but otherwise, survivors of S. enterica wererecovered. Three biological replicates were picked for further examination and notablemacrocolony phenotypic change was observed. Overall, our results reveal the potential ofSh95 I-F1CRISPR-Cas as a plasmid gene editing platform and the importance of bothdeepening our understanding of non-model bacteria and developing novel tools that couldaid in that process.