ATTACK: AssociaTe Toxin-Antitoxin with CRISPR-Cas to Kill Multidrug-Resistant Pathogens

ATTACK: AssociaTe Toxin-Antitoxin with CRISPR-Cas to Kill Multidrug-Resistant Pathogens

As we know, bacteria can evolve resistance against any treatments that threaten their whole population, including antibiotics. Exemplifying the Red Queen Evolution dynamics, antibiotic resistance (AR) is rapidly emerging as antibiotics are being globally used (or overused) to treat infections. To circumvent this dilemma, precise antimicrobial techniques that specifically kill AR pathogens become of urgent need.

Scientists have developed CRISPR-Cas into specific antimicrobials that precisely cut the AR genes, causing specific eradication or re-sensitization of AR pathogens in a complex bacterial community1,2. The CRISPR antimicrobial consists of Cas endonuclease(s) and a crRNA that is programmed to specify AR genes as the cleavage target. However, the fast-evolving bacteria can also acquire resistance to CRISPR antimicrobials, e.g., by inactivating the Cas nuclease with Acr (Anti-CRISPR) proteins, or by breaking down their encoding genes via genomic rearrangements or transposition events. After we recently reported the CreTA (CRISPR-regulated toxin-antitoxin) RNA pairs, which lurk within CRISPR-Cas systems and prevent their inactivation by making them addictive to the bacterial cell3, I initiated the concept of integrating such ‘stabilizing’ element into CRISPR antimicrobials to counteract against these anti-CRISPR mechanisms, leading to more effective sterilization results.

The concept of ATTACK antimicrobial strategy.
Figure 1. The concept of ATTACK antimicrobial strategy.

We envision that when CRISPR antimicrobials manage to keep their activity, pathogens will die from DNA breakage or become sensitive to antibiotics after the loss or mutation of the target AR genes; when pathogens inactivate CRISPR-Cas via various anti-CRISPR mechanisms, CreTA will act as a backup, triggering cellular death and thus presenting a dilemma for the target pathogen (Figure 1). I dubbed this  strategy ‘ATTACK’ (AssociaTe Toxin-Antitoxin and CRISPR-Cas to Kill Pathogens).

To demonstrate the effectiveness of this strategy, we selected to target the carbapenem-resistant Acinetobacter baumannii (CRAB), a highly-priority antimicrobial-resistant pathogen. We considered the native I-F CRISPR-Cas machinery in Acinetobacter species to be the best candidate for developing CRISPR antimicrobials against CRAB. However, we encountered a challenge due to the highly specific nature of the CreTA modules in protecting the cognate CRISPR-Cas apparatus4 and the fact that only two CreTA modules from archaeal I-B systems have been experimentally characterized3,5. To overcome this hurdle, we screened various Acinetobacter species to locate a suitable CreTA element from Acinetobacter sp. WCHA45, and engineered it to fit the CRISPR-Cas apparatus of the A. baumannii type strain AYE. We then designed a mini-CRISPR to target the carbapenem-resistant gene (oxa23) on the chromosome of most CRAB isolates, and developed the AYE CRISPR-Cas apparatus into a targeted antimicrobial specifically against CRAB. By combining the CreTA element and the CRISPR-Cas machinery, our ATTACK strategy demonstrated a higher killing efficiency and re-sensitizing effect compared to the use of CRISPR only.

Inspired by referees, we also tested the performance of ATTACK against a plasmid-born AR gene, which typically leads to curing of antibiotic resistance rather than specific killing of the pathogen. Our lab, however, did not have an A. baumannii strain carrying plasmid-born AR genes, so we opted to demonstrate this application by targeting an engineered plasmid which contained an aac3 gene known to confer gentamycin resistance. Unexpectedly, the combination with CreTA did not seem to improve the AR curing effect of CRISPR antimicrobial when resistant colonies were subjected to screening on plates containing 8 μg/ml gentamycin. However, we observed that when selection was conducted with 4 μg/ml gentamycin, the resistant colonies were significantly reduced by ATTACK as compared to traditional CRISPR antimicrobial. Essentially, this means, ATTACK eradicates plasmid-born AR genes more efficiently and re-sensitizes pathogens more completely. From these findings, we infer that CreTA not only protects CRISPR-Cas from being inactivated, but also ensures its high-level expression and targeting activity.

Although the data we obtained demonstrated the feasibility and effectiveness of ATTACK, the precise toxicity mechanism of the newly identified bacterial CreTA remains a mystery. Through extensive genetic experiments, we were able to show that unlike other known CreT RNA toxins3,5, WCHA45 CreT is a bactericidal small RNA, which probably functions by interacting with essential RNAs such as 16S rRNA and tRNA. Bioinformatic analyses revealed its homologs scattered across dozens of A. baumannii and other species, all of which lurk within an I-F CRISPR-Cas locus. Our team is keen to explore the potential mechanisms and applications of these mysterious small RNA toxins in the future.

In a nutshell, this paper presents the characterization of the first bacterial CreTA module, which safeguards the I-F CRISPR-Cas system, and proposed the ‘ATTACK’ concept to associate toxin-antitoxin with CRISPR-Cas in order to combat multidrug-resistant pathogens. The theoretical findings highlight the prevalence and diversity of RNA toxins evolving alongside CRISPR-Cas, while the ATTACK approach demonstrates their potential in enhancing CRISPR tools, exemplified herein by the creation of an advanced class of precise antimicrobials specifically against AR pathogens.


  1. Citorik, R. J., Mimee, M. & Lu, T. K. Sequence-specific antimicrobials using efficiently delivered RNA-guided nucleases. Nature Biotechnology 32, 1141-1145 (2014).
  2. Bikard, D. et al. Exploiting CRISPR-Cas nucleases to produce sequence-specific antimicrobials. Nature Biotechnology 32, 1146-1150 (2014).
  3. Li, M. et al. Toxin-antitoxin RNA pairs safeguard CRISPR-Cas systems. Science 372, eabe5601 (2021).
  4. Cheng, F. Y. et al. The toxin-antitoxin RNA guards of CRISPR-Cas evolved high specificity through repeat degeneration. Nucleic Acids Research 50, 9442-9452 (2022).
  5. Cheng, F. et al. Divergent degeneration of creA antitoxin genes from minimal CRISPRs and the convergent strategy of tRNA-sequestering CreT toxins. Nucleic Acids Res 49, 10677-10688 (2021).

Please sign in or register for FREE

If you are a registered user on Nature Portfolio Microbiology Community, please sign in