Species-specific prophage induction by ciprofloxacin in human gut metagenomes

This metagenomic study of human gut microbiomes provides in vivo evidence that while ciprofloxacin induces prophages in specific bacterial species, it does not trigger a broad, microbiome-wide induction response.

The Gist

The Hidden Viral Time Bombs: What Happens When We Take Antibiotics?

Imagine your gut is a bustling, crowded city. Inside this city live trillions of bacteria, which are the hardworking citizens keeping your body running smoothly. But hidden inside many of these bacterial citizens are prophages.

Think of a prophage as a sleeping time bomb or a dormant spy living inside a bacterium. Usually, these spies are quiet. They hide their DNA inside the host's DNA, copying themselves along with the host whenever the host divides, but they don't cause any trouble. This is called the "sleeping" or lysogenic state.

However, if the spy gets a specific signal—like a distress call or a threat—it can wake up. This is called induction. When it wakes up, it turns into a lytic virus: it starts building thousands of copies of itself, hijacks the factory, and then blows up the host cell to release a swarm of new viruses. This is the "awake" or lytic state.

Scientists have long known in the lab that a common antibiotic called ciprofloxacin acts like a loud alarm bell, waking up these sleeping spies in test tubes. But the big question was: Does this actually happen inside the messy, complex city of the human gut when real people take antibiotics?

The Big Investigation

To find out, researchers acted like detectives. They didn't just look at bacteria in a petri dish; they looked at the "trash" (DNA) left behind in stool samples from two groups of people:

1. Patients in Bangladesh who had diarrhea and had varying levels of antibiotics in their system.
2. Healthy volunteers in the US who took a controlled course of ciprofloxacin.

They used a clever trick to measure the "awakening" of the viruses. They counted how many "virus parts" (prophage DNA) existed compared to "bacteria parts" (host DNA).

• Normal Ratio (1:1): The spy is sleeping inside the host.
• High Ratio (>1): The spy has woken up, made copies of itself, and is about to (or has just) blown up the host. This is a spike in the "Prophage-to-Host" (P:H) ratio.

The Surprising Findings

1. The City-Wide Alarm Didn't Ring
The researchers first asked: "Did the antibiotic wake up everyone in the gut city?"
The answer was No.
When they looked at the gut as a whole, the average number of "sleeping spies" didn't suddenly spike. The antibiotic didn't cause a massive, city-wide explosion of viruses. This suggests that the gut is more resilient and complex than a simple test tube.

2. The Specific Neighborhoods Were Hit
However, when the detectives zoomed in on specific bacterial species, they found something fascinating. The antibiotic ciprofloxacin did wake up the spies, but only in very specific neighborhoods (specific bacterial species).

• In the Bangladesh group, ciprofloxacin woke up the spies inside Salmonella and Morganella bacteria.
• In the US group, it woke up spies in 13 different bacterial species.

It's like if you rang a fire alarm in a city, and instead of every building catching fire, only the bakery and the library started burning. The rest of the city remained unaffected.

3. The Timing Was Weird
The "explosions" didn't happen all at once. In the healthy volunteers, the biggest spike in viral activity happened on Day 3 of the antibiotic course. For some bacteria, the viruses kept waking up even after the antibiotics stopped; for others, they went back to sleep. It was a chaotic, unpredictable mix of reactions.

Why Does This Matter?

This study is a big deal because it proves that what happens in a test tube (in vitro) doesn't always happen exactly the same way in a real human body (in vivo).

• The "Low-Level" Explosion: The "explosions" in the gut were much smaller than in the lab. It's possible that the antibiotic concentration in the gut is lower than in a test tube, or that not every bacterium has a spy to begin with.
• The "Ghost" Effect: The researchers also noted that sometimes the ratio goes up not because the viruses are waking up, but because the antibiotic kills the bacteria without the viruses, leaving the viruses floating around as "ghosts."
• The Takeaway: Antibiotics aren't a universal "wake-up call" for all gut viruses. Instead, they are like a targeted trigger. They might wake up specific viral time bombs inside specific bacteria, potentially changing the balance of the gut ecosystem in subtle, species-specific ways.

The Bottom Line

Taking ciprofloxacin doesn't turn your entire gut into a viral war zone. However, it does act as a specific trigger that wakes up hidden viruses inside certain types of bacteria. This subtle shift could have long-term effects on how our gut bacteria interact and share genes, reminding us that antibiotics are powerful tools that affect the microscopic world in complex, species-by-species ways.

Technical Summary

1. Problem Statement

While it is well-established in in vitro laboratory settings that antibiotics (specifically fluoroquinolones like ciprofloxacin) trigger prophage induction—the switch of a dormant viral genome (prophage) from the lysogenic cycle to the lytic cycle, resulting in host cell lysis—there is a significant knowledge gap regarding whether this phenomenon occurs within the complex ecosystem of the human gut microbiome in vivo.

• Current Limitations: Existing research relies heavily on culturing specific bacterial isolates, which fails to capture community-level dynamics or species-specific interactions within a natural microbiome.
• Knowledge Gap: It remains unclear if antibiotic exposure in humans causes broad-scale prophage induction across the microbiome or if the effect is restricted to specific bacterial taxa.

2. Methodology

The authors employed a metagenomic approach to infer prophage induction without the need for culturing. They analyzed data from two independent human cohorts with distinct antibiotic exposure profiles:

• Cohort 1 (Bangladesh): 344 stool samples from cholera patients.
  • Exposure: Natural exposure quantified via mass spectrometry (measuring concentrations of ciprofloxacin, doxycycline, and azithromycin in stool).
  • Data: Short-read metagenomes assembled using MEGAHIT.
• Cohort 2 (USA): 60 healthy adult volunteers in a longitudinal study.
  • Exposure: Controlled administration of ciprofloxacin (500mg twice daily for 5 days).
  • Sampling: Timepoints included pre-exposure (baseline), during exposure (days 3, 5), and post-exposure (5 days later).
  • Data: Long-read metagenomes assembled using metaFlye.

Analytical Pipeline:

1. Assembly & Prediction: Metagenomic reads were assembled into contigs. Prophages were identified using geNomad and quality-checked with CheckV.
2. Host Assignment: Probable bacterial hosts for prophages were predicted using iPHoP.
3. Induction Metric (P:H Ratio): The core metric was the Prophage-to-Host (P:H) read depth ratio.
   • Logic: In the lysogenic state, prophage and host DNA copy numbers are roughly equal (Ratio $\approx$ 1). Upon induction, the prophage replicates and releases virions, increasing the relative abundance of viral sequences compared to the host genome (Ratio > 1).
   • Tool: Calculated using PropagAtE.
4. Statistical Analysis: Wilcoxon rank-sum tests (with Bonferroni correction for multiple comparisons) were used to compare P:H ratios between exposed and unexposed groups, or across timepoints.

3. Key Contributions

• First In Vivo Metagenomic Evidence: Provides the first direct evidence of prophage induction in the human gut using metagenomic data, moving beyond in vitro models.
• Refutation of Broad-Scale Induction: Demonstrates that antibiotics do not cause a uniform, community-wide induction of prophages, challenging the assumption that antibiotic exposure universally triggers the "lytic switch" across the microbiome.
• Species-Specific Resolution: Establishes that induction is a taxon-specific phenomenon, occurring only in particular bacterial species rather than the community as a whole.
• Methodological Framework: Validates the use of P:H ratios derived from metagenomic read depth as a proxy for measuring induction dynamics in complex environments.

4. Key Results

A. No Community-Wide Induction

• Analysis of thousands of prophages across both cohorts revealed no significant increase in community-wide P:H ratios associated with antibiotic exposure.
• In the Bangladesh cohort, ciprofloxacin exposure did not elevate the median P:H ratio. Azithromycin exposure was even associated with a slight decrease in P:H ratios, likely due to the selection of bacterial strains lacking prophages.
• In the US cohort, no significant difference in P:H ratios was observed between baseline and any post-antibiotic timepoint when aggregating all prophages.

B. Species-Specific Induction by Ciprofloxacin
When analyzing specific bacterial taxa, ciprofloxacin exposure was strongly correlated with increased P:H ratios in specific species:

• Bangladesh Cohort: Significant increases in P:H ratios were observed in Salmonella enterica (0.94 to 1.13) and Morganella morganii (1.02 to 1.09) in patients with detectable ciprofloxacin.
• US Cohort: 13 bacterial species showed significant increases in P:H ratios at one or more post-exposure timepoints.
  • Peak Timing: The most common peak of induction occurred on Day 3 of the 5-day treatment course.
  • Persistence: Induction signals persisted in six species even after the antibiotic course ended.

C. Impact on Host Abundance

• Despite the induction (lysis) of specific hosts, the study found no significant decline in the relative abundance of these induced species in subsequent timepoints compared to non-induced species.
• This suggests that while induction occurs, the resulting host cell death is not severe enough to drastically alter the population dynamics of these specific bacteria within the timeframe of the study, or that compensatory growth mechanisms are active.

D. Magnitude of Induction

• The observed P:H ratios in the gut were relatively low (peaking around 1.13) compared to in vitro burst sizes which can yield ratios of 100 or more.
• The authors attribute this to lower antibiotic concentrations in the gut, the presence of non-lysogenized bacteria in the population, or the persistence of free viral particles.

5. Significance

• Clinical Implications: The findings suggest that antibiotic therapy does not indiscriminately trigger a "viral storm" across the entire gut microbiome. Instead, the risk of prophage induction is species-specific. This is crucial for understanding antibiotic side effects, such as the potential release of virulence factors or antibiotic resistance genes via horizontal gene transfer, which is mediated by prophages.
• Ecological Insight: The study highlights the resilience of the gut microbiome; even when specific hosts undergo lysis, the overall community structure and the abundance of the affected species remain relatively stable.
• Future Directions: The results call for a more nuanced view of antibiotic-phage interactions, focusing on specific taxa rather than broad community shifts. It also underscores the need to investigate how these subtle, species-specific induction events might impact microbiome function and pathogen evolution over longer timescales.