Horizontal Gene Transfer: How Bacteria Share Resistance

A Single Bacterium Can Spread Antibiotic Resistance to 10 Billion Neighbors Within Hours

Bacteria reproduce asexually — a single cell divides into two identical daughters, passing genes vertically to the next generation. But bacteria also transfer genes horizontally to unrelated bacteria, even across species boundaries, without reproducing. Horizontal gene transfer (HGT) explains a phenomenon that initially baffled 1950s microbiologists: patients hospitalized for Shigella dysentery in Japan began recovering from multiple antibiotics simultaneously. The resistance genes had jumped — not evolved from scratch in each cell, but transferred en masse from other bacterial species. The WHO's 2023 Antimicrobial Resistance (AMR) report attributes HGT as the primary mechanism by which resistance genes spread globally across bacterial populations.

Transformation: Taking Up Naked DNA

Transformation is the uptake of free DNA fragments from the environment — DNA released when other bacteria die and lyse. Frederick Griffith demonstrated the phenomenon in 1928 using Streptococcus pneumoniae before DNA was known to be the genetic material. He showed that heat-killed virulent (smooth) strains could "transform" live avirulent (rough) strains into virulent forms — the dead cells had released their DNA, which the live cells absorbed and integrated.

Not all bacteria are naturally competent — capable of taking up exogenous DNA. Natural competence requires expression of a DNA uptake system, typically triggered by stress conditions: starvation, high cell density, or DNA damage. Competent species include Streptococcus pneumoniae, Haemophilus influenzae, Bacillus subtilis, and Neisseria gonorrhoeae. Some competent species show sequence preferences — Haemophilus preferentially takes up DNA containing the 9-base-pair sequence AAGTGCGGT, which appears at high frequency in Haemophilus genomes but not in other species, providing a form of "DNA self-recognition."

Once inside the cell, incoming double-stranded DNA is degraded to single strands; one strand is integrated into the chromosome by RecA-mediated homologous recombination at regions of sufficient sequence similarity (typically >70–80% identity). Short insertions of resistance genes flanked by conserved regions can integrate even with limited flanking homology.

Transduction: Viral Delivery

Transduction uses bacteriophages (viruses that infect bacteria) as gene delivery vehicles. During phage infection, the phage sometimes accidentally packages bacterial DNA instead of phage DNA into its capsid. When that phage infects another bacterium, it delivers the bacterial DNA rather than its own genome. Two forms exist:

• Generalized transduction — the phage packages random fragments of host DNA. Any gene in the donor bacterium can potentially be transferred, including chromosomal antibiotic resistance genes. Phage P1 (infecting E. coli) and P22 (infecting Salmonella) are classic generalized transducing phages.
• Specialized transduction — a lysogenic phage integrated into the bacterial chromosome sometimes excises imprecisely, taking adjacent chromosomal genes with it. Only genes flanking the phage integration site can be transferred. Lambda phage integrates between the gal and bio operons in E. coli; imprecise excision creates phages carrying gal or bio genes.

Conjugation: Direct Cell-to-Cell Transfer

Conjugation is the most clinically important HGT mechanism for antibiotic resistance spread. A donor bacterium harboring a conjugative plasmid extends a protein tube called a pilus that makes physical contact with a recipient cell. The pilus retracts, bringing the cells into proximity. A channel (mating bridge) forms between the cells, and a copy of the plasmid is transferred to the recipient. The entire process takes 1–5 minutes. A single donor in a dense bacterial culture can conjugate with hundreds of recipients per hour.

Plasmids and Resistance Gene Cargo

Conjugative plasmids are circular DNA molecules with three functional modules: an origin of replication (allows independent maintenance in the host cell), a transfer region (encodes pilus and DNA transfer machinery), and cargo genes. Resistance genes are often carried on plasmids as passengers. R plasmids (resistance plasmids) frequently carry multiple resistance genes simultaneously — a single plasmid may confer resistance to ampicillin, tetracycline, streptomycin, and chloramphenicol.

The New Delhi metallo-beta-lactamase (NDM-1) gene, which confers resistance to virtually all beta-lactam antibiotics including carbapenems (last-resort antibiotics), was first identified in a Swedish patient treated in India in 2008. It spreads on conjugative plasmids and has since been detected in over 100 countries. The blaNDM-1 gene is often co-located on plasmids with genes for resistance to fluoroquinolones, aminoglycosides, and trimethoprim — creating extensively drug-resistant (XDR) organisms with virtually no antibiotic treatment options.

Genomic Islands

Genomic islands (GIs) are large DNA segments (10–200 kb) present in some strains of a species but absent in others, with sequence composition (GC content, codon usage) atypical for the host chromosome — hallmarks of foreign acquisition via HGT. Types include:

• Pathogenicity islands (PAIs) — carry virulence genes. The 40 kb PAI of uropathogenic E. coli encodes hemolysin, fimbriae adhesins, and iron acquisition systems that convert a commensal gut bacterium into a pathogen capable of causing urinary tract infections and sepsis.
• Resistance islands — carry clustered antibiotic resistance genes. The AbaR resistance islands in Acinetobacter baumannii can carry 45+ resistance genes in a single 86 kb island.
• Metabolic islands — carry biosynthetic pathway genes for unusual nutrients; enable new niche colonization.

Implications for Antibiotic Development

HGT has profound implications for how new antibiotics are designed and deployed. Because resistance genes spread horizontally across species boundaries, a resistance mechanism that evolves in soil bacteria encountering natural antibiotics can reach human pathogens rapidly once clinical antibiotic use selects for resistant cells. The 2019 WHO priority pathogens list (ESKAPE organisms: Enterococcus faecium, Staphylococcus aureus, Klebsiella pneumoniae, Acinetobacter baumannii, Pseudomonas aeruginosa, Enterobacter spp.) are all associated with plasmid-mediated HGT of resistance.

Strategies targeting HGT itself include:

• Pilus assembly inhibitors (pilicides) that block conjugation — experimental compounds that reduce conjugation efficiency 10,000-fold in vitro
• CRISPR-Cas systems engineered into bacteria to target and destroy resistance plasmids upon conjugal entry
• Phage therapy targeting resistant strains — bypasses the antibiotic resistance problem entirely
• Restricting antibiotic use in agriculture (the EU banned prophylactic antibiotic use in livestock in 2022) to reduce selection pressure that drives resistance accumulation and spread

The AMR Review commissioned by the UK government (O'Neill Report, 2016) projected that if antibiotic resistance continues at current trends, drug-resistant infections will kill 10 million people annually by 2050 — more than cancer. HGT is not a background biological process. It is the primary engine of one of the greatest public health threats of the 21st century.