Multi-species plasmids and K. pneumoniae clonal spread driving blaNDM outbreak across seven UK healthcare sites

This study characterizes a multi-site UK outbreak of blaNDM-positive Enterobacterales driven by the clonal expansion of *Klebsiella pneumoniae* ST101 and the horizontal spread of IncHI2/IncHI2A plasmids across multiple bacterial species, underscoring the critical need for linked surveillance systems to track infections across fragmented healthcare settings.

The Gist

Imagine a bustling city (Liverpool, UK) where several hospitals are like different neighborhoods. In 2023, the city's "health police" (the microbiology lab) noticed something scary: a super-bug called NDM was showing up way more often than usual. This bug is a master thief; it carries a special tool (an enzyme) that can break open almost any antibiotic lock we try to use to stop it.

Here is the story of how the researchers solved the mystery of where this thief was coming from and how it was spreading, using a mix of detective work and high-tech DNA scanning.

The Mystery: A City-Wide Heist

Usually, when a super-bug outbreak happens, it's like a single gang of thieves (one specific type of bacteria) moving from house to house. But this time, the "thieves" were wearing different masks. The researchers found the NDM gene hiding inside six different species of bacteria (like Klebsiella, E. coli, and Enterobacter). It was as if the same stolen blueprint was being used by a gang of burglars, a group of pickpockets, and a team of hackers all at once.

The Two Ways the Bug Spread

The researchers discovered the outbreak was being driven by two distinct "delivery trucks":

1. The "Clone Truck" (The Bacterial Gang)
One major delivery truck was a specific family of bacteria called Klebsiella pneumoniae ST101.

• The Analogy: Imagine a specific family of burglars who are so good at their job that they keep cloning themselves. They moved from one hospital ward to another, taking over rooms and spreading the infection.
• The Evidence: The DNA of these bacteria was almost identical, like photocopies of the same document. They were mostly found in the Intensive Care Unit (ICU) of one hospital, moving between patients like a contagious cold.

2. The "Plasmid Truck" (The Mobile Toolbox)
The second, and perhaps more dangerous, delivery truck was a plasmid.

• The Analogy: Think of a plasmid not as a bacteria, but as a USB drive or a swappable toolbelt. The NDM gene was stuck on this USB drive.
• The Magic: Bacteria can swap these USB drives with each other, even if they are different species. A Klebsiella could hand its "super-bug USB" to an E. coli, and suddenly the E. coli becomes a super-bug too.
• The Discovery: The researchers found that a specific type of USB drive (called IncHI2/IncHI2A) was the main culprit. It was so efficient that it carried the NDM gene into 82% of the cases. It was like a universal adapter that allowed the resistance gene to jump from one bacterial species to another, spreading the "super-power" across the entire city's hospital network.

The Human Factor: The "Patient Commute"

The study also highlighted a modern problem: Patient Movement.

• The Analogy: Imagine patients as commuters. In the old days, a patient might stay in one hospital. But in modern healthcare, patients are like commuters who take the bus to Hospital A for a check-up, the train to Hospital B for surgery, and a taxi to Hospital C for rehab.
• The Problem: Because these patients moved between different hospitals (and different administrative "neighborhoods"), the bacteria traveled with them. The bacteria didn't care about hospital boundaries or different management teams; they just hopped on the patient and moved to the next ward. This made it very hard to trace the outbreak because the "crime scene" kept changing locations.

The Solution: Putting the Puzzle Together

The researchers used two main tools to solve this:

1. Short-Read Sequencing: Like reading a book page by page to get the general story. This helped them identify the different bacteria and map out the patient movements.
2. Long-Read Sequencing: Like getting a high-definition, 3D map of the book's binding. This allowed them to see the exact structure of the "USB drives" (plasmids) and prove that the same one was jumping between different bacteria.

They even did a lab experiment where they successfully transferred this "USB drive" from one bacteria to another in a petri dish, proving that the bacteria could indeed swap these resistance tools on their own.

The Takeaway

This study teaches us a vital lesson about the future of medicine:

• Don't just look at the bacteria: We can't just track the "burglars" (the bacteria species). We have to track the "tools" (the plasmids) they are carrying.
• Think big: Outbreaks don't stop at hospital doors. Because patients move between different hospitals and systems, we need to treat the whole healthcare network as one connected system. If we don't share data and coordinate across all these "neighborhoods," the super-bugs will keep slipping through the cracks.

In short, the NDM outbreak was a city-wide event driven by a masterful bacterial clone and a very efficient, mobile "USB drive" that jumped between species, all while hitching rides on patients moving through a complex, interconnected healthcare system.

Technical Summary

1. Problem Statement

In 2023, a clinical microbiology laboratory in Merseyside, UK, observed a significant year-on-year increase in bla_NDM (New Delhi metallo-β-lactamase) positive isolates. bla_NDM confers resistance to nearly all beta-lactam antibiotics, including carbapenems, posing a critical threat to patient outcomes.

• The Challenge: Outbreak investigations traditionally focus on clonal spread within a single bacterial species. However, the rise of mobile genetic elements (MGEs) like plasmids allows resistance genes to jump between different species (Enterobacterales), complicating transmission tracking.
• The Context: The healthcare system in Merseyside involves multiple NHS trusts and hospitals with complex patient transfer pathways. The study aimed to determine whether the outbreak was driven by the expansion of a specific bacterial clone, the horizontal transfer of plasmids across species, or a combination of both, and to map these transmission routes across administrative boundaries.

2. Methodology

The study employed a retrospective genomic-epidemiologic approach combining high-resolution sequencing with clinical metadata.

• Study Population: 68 bla_NDM-positive isolates identified between January and December 2023. 63 isolates (from 61 participants and 1 environmental sample) passed quality control.
• Sequencing Strategy:
  • Short-read (Illumina): Performed on all 63 isolates for core genome analysis, species identification, and resistance gene detection.
  • Long-read (Oxford Nanopore): Performed on a subset of 24 isolates to resolve complex plasmid structures, gene cassettes, and synteny.
• Bioinformatics:
  • Phylogenetics: SNP-based phylogenetic trees were constructed to identify clonal lineages (specifically K. pneumoniae ST101).
  • Plasmid Analysis: Hybrid assembly was used to reconstruct full plasmid sequences. Plasmid similarity was assessed to identify shared backbones.
  • Epidemiologic Linking: Patient metadata (ward moves, admission dates) was linked to isolate data. "Strong links" were defined as patients on the same ward on the same day; "weak links" as patients on the same ward at different times.
• Experimental Validation: Conjugation assays were performed to test the transferability of identified plasmids between E. coli donor and recipient strains.

3. Key Contributions

• Dual-Drive Mechanism: The study provides evidence that the outbreak was driven by two distinct mechanisms operating simultaneously:
  1. Clonal Expansion: The spread of a specific high-risk clone, Klebsiella pneumoniae Sequence Type 101 (ST101).
  2. Plasmid-Mediated Spread: The horizontal transfer of a specific IncHI2/IncHI2A plasmid carrying bla_NDM-1 across six different bacterial species (E. coli, K. pneumoniae, Enterobacter hormaechei, Enterobacter kobei, Citrobacter freundii, Citrobacter brakii).
• Multi-Site Tracking: Successfully traced transmission across seven healthcare sites managed by different NHS trusts, highlighting the necessity of cross-trust data sharing.
• Methodological Integration: Demonstrated the value of combining long-read sequencing (for plasmid resolution) with routine clinical metadata (for patient movement tracking) to solve complex, multi-species outbreaks.

4. Results

Bacterial Diversity and Clonal Spread:

• Species: The outbreak involved six species and 28 distinct Sequence Types (STs).
• Dominant Clone: K. pneumoniae ST101 was the only major clonal expansion, accounting for 18 of the 21 K. pneumoniae isolates (86%).
• Transmission Chains: SNP analysis of K. pneumoniae ST101 revealed two sublineages with <15 SNPs difference, consistent with person-to-person transmission.
  • Sublineage 1: Confined to Hospital 1 (Critical Care) over 5 months.
  • Sublineage 2: Spread across Hospitals 1, 3, and 5, linked by patient transfers and shared ward admissions (specifically the acute medical unit H1_RX30 and critical care H1_KH45).

Plasmid Dynamics:

• Dominant Plasmid: 82% of bla_NDM-1 isolates (14/17) harbored the gene on a conserved IncHI2/IncHI2A plasmid (~377 kbp).
• Multi-Species Transmission: This plasmid was found in E. coli, K. pneumoniae, and E. hormaechei.
• Epidemiologic Evidence: A strong epidemiologic link was identified between a K. pneumoniae ST101 isolate and an E. hormaechei isolate in the same critical care ward (H1_KH45) at the same time, both carrying the identical IncHI2/IncHI2A plasmid. This strongly suggests cross-species plasmid transfer within the patient environment.
• Conjugation: The IncHI2/IncHI2A plasmid was experimentally confirmed to be conjugative, transferring from E. coli to E. coli recipients, conferring meropenem resistance.
• Other Variants: bla_NDM-5 was found on diverse plasmids (IncF, IncR) with a conserved transposable unit (IS26-bla_NDM-5-ble-trpF-dsbD-IS91), but lacked the strong epidemiologic clustering seen with bla_NDM-1.

Clinical and Systemic Findings:

• Patient Complexity: 62% of patients had been hospitalized in the 6 months prior to isolation. Patients frequently moved between wards and hospitals (up to 3 different trusts).
• Data Gaps: Transmission signals were lost when patients moved between hospitals with different data systems (e.g., Hospital 4, 6, and 8 lacked detailed ward movement data), highlighting a systemic surveillance blind spot.

5. Significance and Implications

• Paradigm Shift in Outbreak Investigation: The study argues that outbreak investigations must move beyond single-species typing. In modern healthcare, resistance determinants (plasmids) can spread independently of the bacterial host, requiring surveillance that tracks both the clonal lineage and the mobile genetic element.
• Healthcare System Integration: The outbreak spanned multiple NHS trusts with fragmented data systems. The study underscores that effective AMR control requires a "One Health" approach at the health-system level, where data sharing across administrative boundaries is critical to trace transmission chains that would otherwise appear as isolated cases.
• Role of Critical Care: The critical care unit (H1_KH45) acted as a major hub for transmission, likely due to high patient acuity, frequent transfers, and the presence of environmental reservoirs (e.g., sinks).
• Future Directions: The authors call for the development of standardized methods to compare plasmid similarity and integrate plasmid tracking into routine Infection Prevention and Control (IPC) protocols. The falling cost of long-read sequencing makes this feasible.

In conclusion, this paper illustrates a "perfect storm" of AMR spread: a successful bacterial clone (K. pneumoniae ST101) combined with a highly promiscuous plasmid (IncHI2/IncHI2A) moving through a complex, multi-trust healthcare network, driven by patient mobility and facilitated by gaps in cross-system data integration.