Comparison of Galleria mellonella, Epithelial Cell Cytotoxicity, and Mouse Model of Bacteremia to Measure Pseudomonas aeruginosa Virulence

This study demonstrates that the Galleria mellonella infection model serves as a reliable, scalable, and cost-effective alternative to mouse models for assessing population-level virulence trends in Pseudomonas aeruginosa, showing a strong correlation with mouse bacteremia results, whereas epithelial cell cytotoxicity assays exhibited a weaker correlation.

The Gist

Imagine you are a detective trying to figure out which of 105 different "criminals" (bacteria) are the most dangerous. In the world of science, these criminals are strains of Pseudomonas aeruginosa, a tough bacterium that can cause serious infections.

To catch these criminals, scientists usually use the "Gold Standard" test: infecting mice. But mice are expensive, require special care, and many people feel uncomfortable using them for research. So, scientists have been looking for cheaper, easier, and more ethical alternatives. Two popular "substitute detectives" are:

1. The Wax Moth Larva (Galleria mellonella): A caterpillar that looks like a worm but has an immune system that works somewhat like a human's.
2. The Petri Dish (Cell Culture): Growing human lung cells in a dish and seeing if the bacteria kill them.

This paper asks a simple question: Do these two substitutes tell the same story as the mice?

The Experiment: A Three-Way Race

The researchers took 105 different bacterial strains and ran them through all three tests:

• The Mouse Test: How many bacteria does it take to make a mouse sick? (The "Gold Standard").
• The Caterpillar Test: How long does it take for the bacteria to kill a caterpillar?
• The Cell Test: How much damage do the bacteria do to human lung cells in a dish?

The Results: Who Passed the Test?

1. The Caterpillar (G. mellonella): The Reliable Sidekick
The caterpillar model turned out to be a great substitute.

• The Analogy: Think of the mouse test as a high-stakes marathon. The caterpillar test is like a 5K run. They aren't exactly the same distance, but if a runner is fast enough to win the marathon, they are almost certainly fast enough to win the 5K.
• The Finding: There was a strong link between how dangerous a bacteria was to mice and how dangerous it was to caterpillars. If a strain was a "super-villain" to mice, it was also a "super-villain" to caterpillars.
• The Verdict: Scientists can use caterpillars to screen hundreds of bacteria quickly and cheaply to find the most dangerous ones. It's a perfect "first line of defense" before moving to mice.

2. The Cell Dish: The Misleading Clue
The cell culture test was not a good substitute.

• The Analogy: Imagine trying to judge a boxer's strength by watching them punch a pillow. Some boxers are strong but punch the pillow gently; others are weak but hit the pillow hard. The pillow (the cell dish) didn't tell the whole story.
• The Finding: Many bacteria that were very dangerous to mice did not kill the cells in the dish. Conversely, some that killed the cells weren't that dangerous to mice. The test failed to distinguish between the "real" bad guys and the "fake" ones.
• The Verdict: Just because bacteria kill cells in a dish doesn't mean they will make a whole person (or mouse) sick. This test is too simple to predict real-world infections.

The Big Discoveries

Using these models, the researchers found some surprising things about the bacteria:

• The "Resistant" Paradox: Usually, we think bacteria that are hard to kill with antibiotics are also super dangerous. But this study found the opposite! The bacteria that were resistant to the most antibiotics were actually less virulent (less deadly) in the mice and caterpillars.

  • Why? It's like a criminal who spends all their time building a fortress (resistance) and has no energy left to commit crimes (virulence). Being tough against drugs seems to come at the cost of being deadly.

• No "Super-Bugs" in the Family Tree: Scientists often worry about specific "High-Risk Clones" (families of bacteria) that are known to cause outbreaks. They expected these families to be the most deadly.

  • The Twist: The study found that these "High-Risk Clones" were not inherently more deadly than other strains. Their bad reputation likely comes from them being found in sick patients or being hard to treat with drugs, not because they are naturally more lethal.

The Bottom Line

If you want to know if a new strain of Pseudomonas is dangerous:

1. Don't rely on the cell dish test; it's too misleading.
2. Do use the caterpillar test as a fast, cheap, and ethical way to get a good guess.
3. But, if you need to be 100% sure for a medical decision, you still need to check with the mice (the Gold Standard).

In short: The caterpillar is a trustworthy partner for the first round of detective work, but the cell dish is a red herring that leads you down the wrong path.

Technical Summary

1. Problem Statement

Pseudomonas aeruginosa is a major opportunistic pathogen causing severe infections, particularly in immunocompromised patients and those with cystic fibrosis. The mouse bacteremia model is currently considered the "gold standard" for assessing bacterial virulence due to its complex immune system and physiological relevance. However, mouse models are costly, ethically regulated, and low-throughput, making them unsuitable for screening large collections of clinical isolates.

Alternative models, specifically the Galleria mellonella (wax moth) larva and mammalian epithelial cell lines (e.g., A549), are widely used as cheaper, faster, and more humane alternatives. Despite their popularity, there is a significant knowledge gap regarding the concordance of these alternative models with the mouse model. It remains unclear whether they accurately reflect the virulence hierarchy of diverse P. aeruginosa clinical isolates or if they can reliably answer population-level questions about pathogenesis (e.g., the relationship between antibiotic resistance and virulence, or the existence of hypervirulent clones).

2. Methodology

The study utilized a diverse collection of 105 clinical P. aeruginosa isolates obtained from patients in the US, Spain, and Taiwan. These isolates were genetically diverse, representing 59 different Multilocus Sequence Types (STs) and both major phylogenetic clades (Group A: exoS+; Group B: exoU+).

The researchers compared virulence measurements across three distinct models:

1. Mouse Bacteremia Model (Gold Standard): Isolates were injected intravenously into BALB/c mice. Virulence was quantified by calculating the 50% pre-lethal dose (LD50).
2. Galleria mellonella Model: Larvae were injected with bacteria, and virulence was measured by the 50% lethal time (LT50). To standardize comparisons across isolates with varying killing rates, LT50 values were extrapolated to a standardized dose of 3,000 CFU using linear regression of dose-response data.
3. Epithelial Cell Cytotoxicity Model: Human A549 lung epithelial cells were co-cultured with bacteria for 3 hours. Virulence was quantified by measuring Lactate Dehydrogenase (LDH) release (a proxy for cell lysis).

Key Analyses Performed:

• Correlation Analysis: Spearman's rank correlation coefficients ($\rho$) were calculated to compare the rankings of isolates across models.
• Population-Level Questions: The study investigated two specific hypotheses using both mouse and G. mellonella data:
  • Is there an inverse relationship between antibiotic resistance and virulence?
  • Do "High-Risk Clones" (HRCs) or specific genetic clusters exhibit hypervirulence?
• Mechanistic Investigation: The role of Type III Secretion System (T3SS) effectors (ExoS vs. ExoU) was analyzed to explain discrepancies between models.

3. Key Contributions

• Quantitative Concordance: The study provides the first large-scale, quantitative comparison (n=105) of G. mellonella, cell culture, and mouse models for P. aeruginosa.
• Model Validation: It establishes that G. mellonella is a robust surrogate for mouse bacteremia models in identifying virulence trends, whereas cell cytotoxicity assays are poor predictors of in vivo virulence for this pathogen.
• Re-evaluation of High-Risk Clones: The study challenges the assumption that genetically defined "High-Risk Clones" are inherently hypervirulent, suggesting their poor clinical outcomes may stem from antibiotic resistance rather than intrinsic virulence factors.
• Insight into T3SS Dynamics: It reveals differential impacts of ExoS and ExoU effectors depending on the host model, highlighting that virulence mechanisms are host-specific.

4. Key Results

A. Correlation Between Models

• G. mellonella vs. Mouse: A semistrong positive correlation was observed between G. mellonella LT50 and mouse LD50 values ($\rho = 0.75$, $P < 0.0001$). Isolates ranked as highly virulent in mice were consistently highly virulent in larvae, and vice versa.
• Cell Cytotoxicity vs. Mouse: A weak negative correlation was found between A549 cell lysis and mouse LD50 ($\rho = -0.47$, $P < 0.0001$).
  • Reason for discrepancy: Approximately two-thirds of isolates (including many highly virulent ExoS+ strains) caused minimal cell lysis (<5%), rendering the assay unable to discriminate virulence levels within this large group. Only ExoU+ isolates showed high cytotoxicity, but even within this group, cytotoxicity did not strongly predict mouse LD50.

B. Antibiotic Resistance and Virulence

• Both models confirmed an inverse relationship between antibiotic resistance and virulence. Isolates resistant to a higher number of antibiotics (ranked 0–6) exhibited significantly higher LD50/LT50 values (lower virulence).
• This suggests a fitness cost associated with resistance mechanisms, a trend detectable by both the mouse and G. mellonella models.

C. Hypervirulent Clones and Lineages

• High-Risk Clones (HRCs): Isolates belonging to 10 defined HRC STs (e.g., ST235, ST111) were not significantly more virulent than non-HRC isolates in either the mouse or G. mellonella models.
• Genetic Clusters: Phylogenetic clustering based on core genome similarity revealed no distinct "hypervirulent" clusters. While some variation existed, no single lineage was consistently superior in virulence across the population.
• Implication: The poor clinical outcomes associated with HRCs are likely driven by their multidrug resistance and ability to infect compromised hosts, rather than an intrinsic "hypervirulence" phenotype.

D. Role of T3SS Effectors

• In Mice: ExoU+ isolates were generally more virulent than ExoS+ isolates. ExoS-/ ExoU- isolates were the least virulent.
• In G. mellonella: While ExoS+ and ExoU+ isolates were both more virulent than ExoS-/ ExoU- isolates, there was no significant difference in virulence between ExoS+ and ExoU+ groups.
• Discordance: Some ExoS-/ ExoU- isolates (e.g., PASP398) were highly virulent in G. mellonella but avirulent in mice, suggesting G. mellonella is susceptible to virulence factors other than T3SS effectors that are less relevant in the murine bacteremia model.

5. Significance and Conclusion

• Utility of G. mellonella: The study validates G. mellonella as a scalable, cost-effective, and humane first-line model for screening large numbers of P. aeruginosa isolates to identify virulence trends and population-level relationships (e.g., resistance vs. virulence).
• Limitations of Cell Culture: The A549 cytotoxicity assay is not a reliable substitute for in vivo models when assessing general P. aeruginosa virulence, as it fails to capture the virulence of non-cytotoxic (non-ExoU) strains which often cause severe systemic disease.
• Nuanced Virulence: The findings suggest that P. aeruginosa virulence is highly context-dependent. No single "hypervirulent" clone dominates; rather, virulence is a complex interplay of T3SS effectors, antibiotic resistance fitness costs, and host-specific factors.
• Recommendation: While G. mellonella is an excellent screening tool, the authors conclude that final conclusions regarding virulence mechanisms or specific isolate potential should still be confirmed in mouse models, particularly when investigating specific molecular determinants or when the G. mellonella results show high divergence from expected mouse outcomes.