Gain and loss of plasmid-borne antibiotic resistance genes are associated with chromosomal resistance presence in Enterobacteriaceae

This study analyzes 6,895 *Enterobacteriaceae* genomes to reveal that plasmid-borne antibiotic resistance gene dynamics are primarily species-dependent and that the presence of chromosomal resistance genes significantly predicts a lineage's capacity for acquiring and retaining plasmid-borne resistance.

The Gist

The Big Picture: Bacteria's "Backpacks" of Resistance

Imagine bacteria as tiny travelers. To survive in a world full of antibiotics (which are like poison traps), they carry plasmids. Think of a plasmid as a backpack or a USB drive that bacteria can swap with each other.

Inside these backpacks are Antibiotic Resistance Genes (ARGs). These are the "cheat codes" that let the bacteria ignore the poison.

This study looked at nearly 7,000 different bacteria from the Enterobacteriaceae family (which includes famous troublemakers like E. coli and Salmonella) to understand how these cheat codes move, multiply, and disappear over time.

---

Key Discovery 1: The "Copy-Paste" vs. "Delete" Game

The researchers wanted to know: Do bacteria treat resistance genes differently than other genes in their backpacks?

• Getting New Genes (Gain): It turns out, getting a new type of resistance gene is just as easy for bacteria as getting any other random gene. It's like picking up a new tool from a toolbox; the process is the same whether the tool is a hammer or a cheat code.
• Multiplying Genes (Expansion): However, once they have a resistance gene, they go crazy with it. They make copies of it. Imagine if you found a cheat code for a video game and immediately made 100 copies of it to sell to your friends. Bacteria do this with resistance genes much more often than with normal genes.
• Throwing Genes Away (Loss): Conversely, when they decide to get rid of a resistance gene, they also do it faster than with normal genes.

The Analogy: Think of a normal gene as a pair of shoes. You buy one pair, wear them, maybe lose them eventually. But a resistance gene is like gold bars. If you find a gold bar, you immediately try to find more (expansion) to make you richer. But if the gold becomes a burden, you might try to get rid of it quickly too. The bacteria are constantly copying and deleting these "gold bars" (resistance genes) at a frantic pace, depending on which specific species of bacteria they are.

Key Discovery 2: It's All About the "Family" (Species)

The study asked: Does the type of antibiotic matter more, or does the type of bacteria matter more?

• The Verdict: The bacteria species matters way more.
• The Analogy: Imagine a school.
  • Drug Class: This is like the subject (Math, History, Art).
  • Species: This is like the student's personality.
  • The study found that some students (bacteria species) are naturally "copy-paste" machines. They will copy and delete resistance genes wildly, regardless of whether the subject is Math or History. Other students are very stable and don't change much.
  • The specific antibiotic (the subject) barely changes the behavior; it's the bacterial "personality" that drives the chaos.

Key Discovery 3: The "Chromosomal Anchor" Effect

This is the most surprising part. The researchers looked at bacteria that already had resistance genes stuck permanently in their main DNA (the chromosome). Let's call these "Anchored Bacteria."

They compared "Anchored Bacteria" to their "Sister Bacteria" (cousins that are almost identical but don't have the permanent resistance).

• The Finding: The "Anchored Bacteria" were much better at grabbing new resistance backpacks (plasmids) and much worse at losing them.
• The Analogy: Imagine two houses.
  • House A (No Anchor): The owners are casual. They might buy a security system (plasmid) today, but if it gets heavy, they might throw it out tomorrow. They are fickle.
  • House B (With Anchor): The owners already have a permanent, built-in safe (chromosomal resistance) in their basement. Because they are already "committed" to being secure, they are obsessed with security. They buy more security systems, keep them longer, and never throw them away.
  • Why? Having that permanent "safe" seems to make the bacteria more open to accepting new "backpacks" of resistance. It's a domino effect: once you have one permanent defense, you become a magnet for more.

Key Discovery 4: Specific Backpacks for Specific Jobs

The study also found that certain types of plasmid "backpacks" are very picky about what they carry and who they hang out with.

• Example: One specific type of backpack (called IncQ2) was found only in one specific species of bacteria (Leclercia adecarboxylata) and always carried a specific gene (qnrS2).
• The Analogy: It's like a specific brand of delivery truck that only ever delivers pizza and only drives in one specific neighborhood. It's not a general-purpose truck; it's a specialized vehicle for a specific job.

Why Does This Matter? (The "So What?")

1. Predicting the Future: If we see a bacterial lineage that has "anchored" resistance genes in its main DNA, we can predict that this family is a super-spreaders of new resistance. They are the ones most likely to pick up the next super-drug resistance.
2. Monitoring: Instead of just looking for the resistance gene itself, doctors and scientists should look for these "Anchored" bacterial families. If they find them, they know to be extra careful, because these bacteria are likely to grab more resistance soon.
3. Evolutionary Strategy: Bacteria aren't just passive victims; they are active strategists. They use a "copy-paste-delete" strategy to keep their resistance levels just right—enough to survive the antibiotics, but not so much that it slows them down.

Summary in One Sentence

Bacteria don't just randomly pick up resistance; specific bacterial families are naturally prone to hoarding and copying resistance genes, and if they already have a permanent version of the gene, they become even more aggressive at collecting new ones, making them the most dangerous players in the antibiotic resistance game.

Technical Summary

1. Problem Statement

Antibiotic resistance genes (ARGs) are primarily disseminated via plasmids, which are highly mobile and evolvable genetic elements. While previous studies have documented specific outbreaks or the epidemiology of individual ARGs, there is a lack of systematic understanding regarding:

• The evolutionary dynamics of plasmid-borne ARGs (pARGs) compared to other plasmid genes.
• Whether pARG turnover is driven by species-specific factors or antibiotic class.
• The relationship between chromosomal ARGs (cARGs) and the acquisition/retention of plasmid-borne resistance.
• How pARGs are maintained and amplified across bacterial lineages in the Enterobacteriaceae family.

2. Methodology

The study employed a large-scale phylogenomic approach using 6,895 high-quality Enterobacteriaceae genomes (5,335 containing plasmids) from the Genome Taxonomy Database (GTDB).

• Data Processing:
  • Genomes were filtered for quality (completeness >99%, contamination <1%).
  • Plasmids and genes were identified using geNomad and Platon.
  • ARGs were annotated using RGI against the CARD database.
  • Pan-genomes were constructed using OrthoFinder.
• Evolutionary Modeling:
  • A phylogeny-aware birth-death model (implemented in Count v10.04) was applied to quantify four evolutionary processes:
    1. Gain: Acquisition of a new gene family (0 to 1 copy).
    2. Loss: Complete disappearance of a gene family (n to 0 copies).
    3. Expansion: Increase in copy number within an existing family (n to m, m > n).
    4. Reduction: Decrease in copy number while retaining at least one copy (n to m, n > m ≥ 1).
• Statistical Analysis:
  • Comparative Rates: Paired Wilcoxon signed-rank tests compared pARGs vs. non-ARG plasmid genes.
  • Variance Partitioning: Ordinary Least Squares (OLS) models determined the contribution of Species vs. Drug Class to evolutionary rates.
  • Sister Clade Analysis: A comparative phylogenetic approach identified 464 pairs of sister clades where one lineage harbored ≥80% cARGs and the other <20%. Linear Mixed-Effects Models (LMM) were used to compare pARG gain/loss rates between these clades.
  • Plasmid Network Analysis: MOB-recon and Mash distances were used to reconstruct plasmid units and visualize gene-replicon associations.

3. Key Contributions

• Quantification of Turnover Dynamics: The study is the first to systematically quantify the four distinct evolutionary processes (gain, loss, expansion, reduction) for pARGs across nearly 7,000 genomes.
• Decoupling of Drivers: It demonstrates that while pARG acquisition (gain) follows general plasmid evolutionary rules, pARG copy-number variation (expansion/reduction) is heavily driven by species identity rather than antibiotic class.
• cARG as a Genomic Marker: It establishes a strong, quantitative link between the presence of chromosomal ARGs and an increased capacity for acquiring and retaining plasmid-borne resistance.
• Specific Replicon-ARG Associations: It identifies unique associations between specific plasmid backbones (e.g., IncQ2, Col(VCM04)) and specific resistance genes within distinct genera.

4. Key Results

A. Evolutionary Dynamics of pARGs

• Gain vs. Copy Number: pARGs exhibit similar gain rates to non-ARG plasmid genes, suggesting that the initial acquisition of resistance families follows common horizontal gene transfer pathways.
• Expansion and Reduction: pARGs show significantly higher rates of expansion and reduction compared to non-ARGs. This indicates that once acquired, pARGs undergo frequent copy-number fluctuations (e.g., via transposons or plasmid copy number mutations).
• Species Dependence: Evolutionary rates are strongly species-dependent (explaining ~66% of variance for gain and ~63% for loss). In contrast, the specific antibiotic class explains very little variance (<7%).
• Coupling: Expansion and reduction rates are tightly coupled (high expansion correlates with high reduction), whereas gain and loss are largely independent.

B. Association with Chromosomal ARGs (cARGs)

• Enhanced Acquisition: Bacterial clades harboring cARGs acquire pARGs at 2.47-fold higher rates than their sister clades lacking cARGs.
• Reduced Loss: These same clades exhibit significantly lower loss rates for pARGs (loss rate fold-change = 0.138), indicating a strong tendency to retain resistance genes.
• Plasmid Burden: cARG-containing clades carry a higher median plasmid burden (2.64 unique plasmid clusters) compared to sister clades (1.66).
• Independence: The pARG-associated replicons in cARG clades are rarely shared with sister clades (0% median sharing), suggesting these lineages actively acquire new plasmids rather than inheriting them from a common ancestor.

C. Specific Gene-Replicon Associations

• IncQ2: Exclusively associated with qnrS2 and found only in Leclercia adecarboxylata.
• Col(VCM04): Predominantly (71.4%) carries mprF and is found in the Citrobacter genus.
• ColRNAI_rep_cluster_1987: Associated with qacG (disinfectant resistance), primarily in Serratia nevei.
• Multidrug Resistance: Unlike the specific associations above, multidrug resistance plasmids show heterogeneous and complex ARG compositions.

D. Validation

The study validated its model by identifying specific evolutionary events that matched known outbreaks, such as the acquisition of blaCTX-M in Enterobacter quasihormaechei (ST873, France) and mcr in Escherichia fergusonii (China).

5. Significance and Implications

• Predictive Surveillance: The presence of cARGs serves as a robust genomic marker for identifying bacterial lineages with a high evolutionary potential to accumulate multidrug resistance via plasmids.
• Evolutionary Strategy: The "expansion-reduction" cycle suggests that bacteria use copy-number plasticity as a flexible resistance strategy, allowing rapid adaptation to selective pressures without permanent genomic commitment until stabilization occurs.
• One Health Monitoring: The findings highlight that resistance evolution is driven more by the host species' genomic background than by the specific antibiotic used, suggesting that monitoring specific bacterial lineages is more effective than monitoring specific drug classes for predicting resistance emergence.
• Mechanistic Insight: The results support the hypothesis that chromosomal integration may act as a "stabilization" step that facilitates the subsequent uptake of diverse plasmid repertoires, or conversely, that high plasmid flux drives chromosomal integration.

In conclusion, this paper shifts the focus from individual gene epidemiology to lineage-level evolutionary dynamics, revealing that the genomic context (specifically the presence of cARGs) is the primary determinant of a bacterium's ability to acquire and maintain plasmid-mediated antibiotic resistance.