Double agent: how Escherichia coli switches from commensal to pathogen in the urinary tract infection.

Escherichia coli is a biological double agent. In the crowded ecosystem of the human gut, it is a beneficial commensal. It helps to synthesize vitamins and prevent the colonization of harmful pathogens. Yet, when this same bacterium migrates to the urinary tract, it can undergo a radical transformation into a uropathogen. This causes infections that affect hundreds of millions of people annually.

For decades, clinical medicine has searched for genetic resistance. This refers to permanent mutations in bacterial DNA that render antibiotics ineffective. However, many patients suffer from recurrent urinary tract infections (UTIs) even when prescribed antibiotics are technically "appropriate." This discrepancy suggests the problem is not always a change in the bacterium's blueprint. Instead, it is a change in its behavior. This review establishes that E. coli utilizes phenotypic plasticity. This is the capacity for dynamic, non-heritable, and reversible adaptation. Such plasticity allows the bacteria to survive both the host environment and antibiotic onslaughts.

The failure of standard susceptibility testing

The central crisis in treating UTIs is the "phenotypic tolerance gap." Clinicians currently rely on standard Antibiotic Susceptibility Testing (AST) to determine treatment. These tests typically measure the Minimum Inhibitory Concentration (MIC). This is the lowest concentration of a drug that prevents visible bacterial growth. Tests are conducted under controlled, laboratory conditions.

However, the paper argues these tests are fundamentally mismatched with real infections. Standard AST evaluates "planktonic" bacteria (free-floating cells) growing rapidly in nutrient-rich media. In contrast, a real infection occurs in a hostile, nutrient-poor environment. This environment has high osmolality (the concentration of solutes in a solution) and physical shear stress from urine flow. Because the lab test ignores specialized states like intracellular bacterial communities (IBCs) or biofilms, it fails to predict real-world behavior. This mismatch explains why a strain can appear "susceptible" in a petri dish but remain untreatable in the body.

A multi-layered survival strategy

The transition from a harmless gut resident to a uropathogen is an active, signal-integrated decision. The review decomposes this adaptation into a tiered regulatory response:

1. Rapid Osmotic and Metabolic Sensing: Within minutes of entering the urine, increased osmolality triggers the EnvZ/OmpR two-component system (a signaling mechanism where one protein senses a stimulus and transfers a signal to another). This causes a "porin shift." This process remodels the bacterial outer membrane to reduce the entry of hydrophilic antibiotics like $\beta$-lactams by 60–80%. Simultaneously, leucine scarcity inactivates the regulator Lrp. This relieves repression of fimbrial genes, allowing the bacteria to stick to the bladder wall.
2. Nutrient Acquisition and Growth Slowdown: As bacteria encounter the extreme iron limitation of human urine, they activate the Fur regulon to produce siderophores (molecules that scavenge iron from host proteins). At the same time, the accumulation of (p)ppGpp (alarmone nucleotides that signal cellular stress) triggers a "stringent response." This slows DNA replication and protein synthesis. It effectively puts the bacteria into a low-metabolic "sleep" state.
3. Structural and Morphological Reprogramming: Under prolonged stress, global regulators like RpoS (a master sigma factor that redirects RNA polymerase to stress-response genes) drive physical changes. These include filamentation (the elongation of cells), surface roughening, and the formation of multicellular aggregates or biofilms.

As shown in, these stages—adhesion, colonization, toxin release, and tissue damage—are part of a dynamic cycle. This cycle is fueled by these adaptive shifts.

Quantifying the tolerance gap

The paper distinguishes between genetic resistance and phenotypic tolerance using four parameters: heritability, MIC, the Minimum Duration for Killing (MDK), and the genetic basis.

The most striking evidence lies in the MDK${99}$. This is the time required to kill 99% of a bacterial population. The authors cite studies where tolerant strains had normal MICs. These strains appeared perfectly susceptible in standard tests but required much longer to die. For instance, in ampicillin-exposed E. coli, tolerant isolates required $6.2 \pm 1.8$ hours to achieve 99.9% killing. Non-tolerant isolates required only $1.8 \pm 0.4$ hours. Similarly, high-persistence strains treated with ciprofloxacin showed an MDK$$ of 8.4 hours. Low-persistence controls were cleared in just 2.1 hours.

Furthermore, the paper highlights the survival advantage provided by these regulatory shifts. High-RpoS UPEC isolates showed a 5.7-fold increase in survival after six hours of ampicillin exposure. This was compared to low-RpoS isolates with identical MIC values. This confirms the bacteria are not ignoring the drug through mutation. Instead, they are actively managing their physiology to endure it.

Limitations of current modeling

The review notes significant hurdles in how we study these interactions. Much of our understanding relies on the 5637 human bladder epithelial cell line. However, the authors warn that 5637 cells are derived from a carcinoma (cancer) patient. This introduces three major confounding factors: * Altered Surface Chemistry: Cancer cells overexpress specific sugars like the Tn antigen. This may change how bacterial adhesins bind compared to healthy tissue. * Modified Receptor Density: The $\alpha$3$\beta$1 integrin receptor is a key docking site for E. coli. It is upregulated in these cancer cells. This may inflate observed invasion rates. * Immune Misrepresentation: These cells secrete lower levels of inflammatory cytokines like IL-6 and IL-8. Therefore, current models likely underestimate the true host inflammatory response.

Additionally, the review points out a research gap. We still lack a systematic way to categorize which specific antibiotic classes trigger which specific morphological changes. This limits our ability to design personalized treatments.

The verdict: Move beyond the MIC

The evidence is clear. Relying solely on standard MIC-based susceptibility testing may fail patients with recurrent UTIs. The "double agent" nature of E. coli means that antibiotic administration can inadvertently trigger phenotypic plasticity. This includes filamentation or biofilm formation, which ensures bacterial survival.

The path forward requires "anti-persistence" strategies. This includes using bacteriophages (viruses that infect bacteria) to disrupt biofilms. It also includes targeting metabolic pathways to prevent glucose-mediated tolerance. Finally, host-directed therapies like HDAC inhibitors may prevent the epigenetic reprogramming of the bladder. Until we treat bacterial behavior rather than just their DNA, the cycle of recurrence will remain.

Figures from the paper