Bacterial Syringe Delivers Cre Protein Into Mouse Lung Cells

Scientists have long dreamed of delivering therapeutic proteins into cells without ever touching the cell’s DNA. Now, a team at the University of South Alabama and the Mitchell Cancer Institute has taken a significant step toward that goal, showing that a modified strain of the opportunistic pathogen Pseudomonas aeruginosa can inject a functional genome-editing protein directly into the nuclei of living lung cells in mice. The study, published in Current Research in Biotechnology, demonstrates for the first time in an intact animal that the bacterium’s type III secretion system, a needle-like molecular machine normally used to wage war on host cells, can be repurposed to recombine genes inside mammalian lung tissue.

The type III secretion system, or T3SS, is one of the most sophisticated weapons in the bacterial arsenal. It resembles a microscopic syringe, spanning the two bacterial membranes and protruding outward to puncture the membrane of a host cell. Through this conduit, disease-causing bacteria push effector proteins that hijack cellular machinery. In P. aeruginosa, a Gram-negative pathogen responsible for ventilator-associated pneumonia, bloodstream infections and devastating wound infections, the T3SS injects four known exoenzymes: ExoU, a potent phospholipase A2 cytotoxin; ExoS and ExoT, bifunctional Rho GTPase-activating proteins with ADP-ribosyltransferase activity; and ExoY, a promiscuous nucleotidyl cyclase. Together these toxins help make P. aeruginosa one of the six leading pathogens driving antimicrobial resistance-associated mortality, blamed for an estimated 559,000 deaths worldwide each year.

Yet the very machinery that makes the bacterium dangerous also makes it an attractive delivery vehicle. Crucially, proteins targeted for T3SS export require only a short N-terminal signal, roughly the first 50 to 60 amino acids, to be recognized by the secretion apparatus. That molecular simplicity means researchers can fuse this targeting sequence to almost any protein of interest and trick the bacterium into injecting it. Unlike viral vectors, which have earned a growing list of FDA approvals but carry risks of insertional mutagenesis, unregulated transgene expression and anti-vector immune responses, bacterial protein delivery leaves no foreign DNA behind in the host genome. The effect is transient, dose-controllable and potentially reversible, features that have fueled interest in microbial-based cancer therapy since the era of Busch and Coley more than a century ago.

To prove the concept in a living animal, the team, led by Amanda N. Tuckey and Robert W. Sobol, chose the Cre-loxP recombination system as a rigorous readout. Cre recombinase, derived from bacteriophage P1, recognizes 34 base-pair loxP sequences, each composed of two 13 base-pair palindromic repeats flanking an asymmetric eight base-pair core. When Cre finds two loxP sites oriented in the same direction, it excises the intervening DNA. This makes Cre an ideal test cargo: the protein must not only be secreted by the bacteria and translocated across the host cell membrane, but also imported into the nucleus and enzymatically active on chromosomal DNA before any detectable change occurs.

The researchers first validated the readout in human cells. They engineered U2OS osteosarcoma cells to carry a LoxP-DsRed-LoxP-EGFP reporter, originally developed by Jacco van Rheenen, in which cells glow red until Cre excises the DsRed cassette, switching fluorescence to green. Confocal microscopy and flow cytometry confirmed that EGFP appeared only when Cre was delivered, never in empty-vector controls. With the cellular assay established, the team turned to the bacterium itself.

They built an arabinose-inducible expression plasmid, pUCP18-Ara, by cloning the araC regulator and araBAD promoter from E. coli into a shuttle vector that replicates in both E. coli and Pseudomonas. Into this backbone they placed fusion genes linking the first 54 amino acids of each of the four exoenzymes to a nuclear localization sequence and Cre recombinase. Secretion tests, in which culture supernatants were concentrated by trichloroacetic acid precipitation and probed with an anti-Cre antibody, revealed that the ExoU targeting sequence drove the most reliable and abundant secretion of the fusion protein. The ExoU(54)-NLS-Cre strain therefore became the workhorse for the animal experiments.

A key safety consideration shaped the entire design. The team used strain PA103, a well-characterized P. aeruginosa isolate, but specifically an isogenic mutant, PA103ΔUT, lacking functional ExoU and ExoT. This deletion substantially attenuates virulence while preserving the intact T3SS needle. The bacteria were delivered by intratracheal instillation into the airways of Ai14 reporter mice, animals whose chromosomes carry a loxP-flanked transcriptional STOP cassette upstream of the red fluorescent protein tdTomato. Any cell that receives functional Cre in its nucleus will delete the STOP cassette and light up red, providing a permanent, cell-autonomous record of successful protein delivery. A low inoculum of 2 × 10⁴ colony-forming units was chosen to minimize lung pathology, and arabinose induction was maintained by intraperitoneal injection at the time of infection and on the two following days.

Seventy-two hours after infection, the mice were euthanized and their lungs perfused, fixed and cut into 300-micrometer serial sections with a vibratome. Confocal microscopy of five fields per animal revealed a striking difference: mice infected with the empty-vector control strain showed minimal red fluorescence background, while those receiving the ExoU-NLS-Cre bacteria displayed significantly more tdTomato-positive cells throughout the lung slices. Quantification of red fluorescence normalized to DAPI-stained tissue area confirmed the difference was statistically significant, with a two-tailed Welch’s t-test yielding p = 0.035 and a large effect size. Every animal survived, with no weight loss or overt signs of stress, demonstrating that functional protein delivery to the lung cell nucleus is feasible in a living mammal without causing measurable harm.

The long-term ambition is to engineer this platform, sometimes called a Type III Protein Delivery System, to treat lung cancer, which accounts for roughly 14 percent of all new cancers and is projected by the American Cancer Society to claim more than 150,000 lives annually in the United States. Because the T3SS bypasses endosomal degradation and deposits proteins directly into the cytosol, it could deliver tumor suppressors, pro-apoptotic factors or synthetic lethal enzymes to cancer cells while sidestepping the genotoxic risks of gene transfer. Related systems built in Yersinia enterocolitica and Salmonella enterica have already been explored for anticancer payloads, and the Pseudomonas T3SS has previously been used in cell culture to inject transcription factors for cellular reprogramming and even antigens for SARS-CoV-2 vaccine development.

Significant engineering challenges remain before the platform approaches the clinic. Although PA103ΔUT is attenuated, the researchers found that residual cytotoxicity, likely from Exotoxin A, the Type II-secreted ADP-ribosyltransferase that inactivates host elongation factor 2, caused cultured lung cells to round up and detach within 48 hours of exposure, even though the effect was not apparent in the intact mouse. Future strains will need additional mutations, such as a disrupted toxA gene, and possibly auxotrophies that force the bacteria to die outside the controlled infection setting, ensuring biocontainment without relying on antibiotics. The team also acknowledges that Cre itself can cause off-target genomic effects, and plans to compare additional T3SS targeting sequences in animal models to maximize efficiency and specificity. Still, the demonstration that a disarmed pathogen can rewrite the genome of lung cells purely through protein injection marks a compelling proof of principle for protein-based therapeutics that never touch DNA.

Subject of Research: Type III secretion system-mediated protein delivery causing genomic recombination in murine lungs

Article Title: Exploiting the bacterial type III secretion system of Pseudomonas aeruginosa to impact genomic recombination in murine lungs

Article References: N.Tuckey, A., Clark, J., Q.Al-Rahahleh, R., H.Siddiqui, A., A.DeFreitas, S., P.Roos, W., T.Lin, M., Audia, J. P., & W.Sobol, R. (2026). Exploiting the bacterial type III secretion system of Pseudomonas aeruginosa to impact genomic recombination in murine lungs. Current Research in Biotechnology, Article 100420. https://doi.org/10.1016/j.crbiot.2026.100420

Image Credits: AI Generated

DOI: 10.1016/j.crbiot.2026.100420

Keywords: Pseudomonas aeruginosa, type III secretion system, Cre-loxP recombination, protein delivery, bacterial therapeutics, lung cancer, tdTomato reporter mice, ExoU, microbial cancer therapy, gene recombination, PA103ΔUT, in vivo delivery