Integrative and Conjugative Element

An integrative and conjugative element (ICE) (historically termed a conjugative transposon) is a self-transmissible, modular chromosomal mobile genetic element ranging from ~18 kilobases (kb) to >600 kb that drives horizontal gene transfer (HGT) across bacterial species, families, and phyla. Unlike extrachromosomal conjugative plasmids, which replicate autonomously in the cytoplasm and risk segregational loss during cell division, an ICE resides integrated within the host bacterial chromosome and is passively replicated alongside chromosomal DNA with near-zero metabolic burden. Under specific physiological or environmental triggers—such as the RecA/LexA SOS DNA-damage response (stimulated by fluoroquinolones or mitomycin C), sub-inhibitory tetracycline or β-lactam exposure, stationary-phase quorum sensing, or oxidative stress—the ICE excises from the chromosome as a covalently closed circular double-stranded DNA intermediate, nicks its origin of transfer (oriT), and transfers a single-stranded DNA (ssDNA) copy through a self-encoded Type IV Secretion System (T4SS) mating channel into a recipient bacterium, while the donor copy undergoes rolling-circle synthesis and re-integrates into the donor chromosome.

Comprehensive comparative genomics by Eduardo Rocha, Guglielmini, and colleagues (2011) revealed that chromosomally integrated ICEs outnumber autonomous conjugative plasmids across sequenced bacterial genomes by nearly 2:1, making ICEs the most abundant conjugative apparatus in prokaryotes. In clinical infectious disease, ICEs serve as the primary chromosomal vehicles disseminating multidrug antimicrobial resistance (AMR) cassettes—including metallo-β-lactamases (blaNDM-1, blaVIM-2), serine carbapenemases (blaKPC-2, blaOXA-48), vancomycin resistance (Tn1549/vanB), and macrolide/tetracycline efflux operons—alongside heavy-metal tolerance (mer, czc, ars), symbiosis islands (the 502-kb Mesorhizobium loti ICEMlSymR7A), and pathogenicity islands across ESKAPE nosocomial pathogens (Pseudomonas aeruginosa, Klebsiella pneumoniae, Acinetobacter baumannii, Enterobacter spp., and Enterococcus faecium) and Vibrio cholerae (SXT/R391 family).

The Three Core Functional Modules Plus Variable Cargo Islands

Every canonical ICE is organized as a mosaic syntenic chromosome segment bounded by flanking direct repeats (attL and attR, typically 8 to 60 bp in length) and partitioned into three conserved core backbone modules interspersed with variable adaptation/cargo hotspots (HS1–HS5):

1. Integration and Excision Module (int and xis): Encodes a site-specific integrase (Int)—most commonly a phage λ-like tyrosine recombinase (utilizing a catalytic Arg–Lys–His–Arg–His/Trp pentad and a nucleophilic active-site Tyr to form a transient 3'-phosphotyrosyl covalent bond and Holliday junction intermediate) or, in a minority of elements (such as Tn916/Tn5397 in Firmicutes), a large serine recombinase or DDE transposase. Excision in the reverse direction (attL × attR → attB + attP) requires a small, basic recombination directionality factor (RDF or excisionase, Xis) that bends the att DNA arms to favor excisive over integrative synapsis.
2. Conjugative DNA Processing / Relaxosome Module (oriT, mob/traI, traJ): Contains the cis-acting origin of transfer (oriT) bearing an inverted repeat and the nic cleavage site, paired with a cognate relaxase (MOB family: MOB_H, MOB_P, MOB_F, or MOB_C)—a histidine-triad (HUH) or tyrosine transesterase that cleaves a specific phosphodiester bond on the transferred (T) strand and remains covalently linked to the 5' phosphate of the single-stranded DNA throughout intercellular transport.
3. Mating-Pair Formation (MPF / T4SS) and Coupling Protein (T4CP): A 12- to 25-gene operon encoding the transmembrane Type IV Secretion System (T4SS) translocation channel (VirB1–VirB11 / Tra homologs), the extracellular conjugative pilus, and the hexameric inner-membrane Type IV Coupling Protein (T4CP, VirD4/TraD/TraG), an ATP-hydrolyzing translocase that docks the covalent Relaxase–5'-ssDNA nucleoprotein complex onto the T4SS pore.

Why ICEs Integrate Specifically at the 3′ Ends of Chromosomal tRNA Genes

Over 70% of characterized γ-proteobacterial and β-proteobacterial ICEs (including ICEclc and PAPI-1 in Pseudomonas, ICEKp in Klebsiella pneumoniae, and SXT/R391 at prfC or tRNA loci) integrate specifically into the 3′ distal end of chromosomal tRNA genes (most frequently tRNA-Gly, tRNA-Leu, tRNA-Ser, tRNA-Phe, and tRNA-Pro) or transfer-messenger RNA (tmRNA / ssrA).

This evolutionary targeting strategy solves three fundamental problems for a mobile element:

1. Universal Cross-Species Conservation: Because the cloverleaf secondary structure (TΨC stem-loop and acceptor stem) of tRNA genes is under intense purifying selection across all bacteria, targeting a 15–50 bp sequence at the 3' end of a tRNA gene provides a conserved chromosomal attachment site (attB) across distantly related genera and families, enabling broad-host-range conjugation.
2. Non-Disruptive Duplication of the Host tRNA (attL / attR): The circular ICE's attachment site (attP) carries an exact duplicate of the terminal 15–50 nucleotides of the target host tRNA gene (including its 3'-CCA tail and Rho-independent terminator). When the tyrosine integrase recombines attP into attB, it restores an intact, fully functional host tRNA gene at one boundary (attL or attR) while leaving the truncated duplicate at the opposite boundary—preventing any loss of host translational fitness.
3. Transcriptional Insulation and Structural Stability: Chromosomal tRNA loci are flanked by strong intrinsic terminators and nucleoid-associated protein (H-NS / Fis) binding sites that insulate the surrounding host chromosome from read-through transcription of incoming foreign cargo operons.

Because integrated ICEs are replicated seamlessly within a linear bacterial chromosome assembly (contig), they are invisible to standard plasmid-only finders (PlasmidFinder) and 16S rRNA or 7-locus housekeeping multilocus sequence typing (MLST). Bioinformatic genome-mining pipelines identify recently acquired ICE and genomic island insertions without relying solely on database sequence homology by scanning chromosomes for compositional oligonucleotide and codon usage anomalies relative to the native host chromosome backbone:

1. 256-Dimensional Tetranucleotide Kullback–Leibler Divergence (D_KL): Every bacterial species maintains a characteristic genome-wide frequency signature across all 4⁴ = 256 four-nucleotide words (k = 4, Q(w)) shaped by its native DNA replication machinery, restriction-modification systems, and DNA methylation enzymes. For a sliding chromosomal window W of length L_win (e.g., 5–10 kb with 1-kb step) exhibiting empirical tetranucleotide frequencies P_W(w), the Kullback–Leibler divergence from the core host chromosome background Q_host(w) is:

D_KL(P_W ‖ Q_host) = ∑_{w ∈ {A,C,G,T}⁴} P_W(w) log₂ [ P_W(w) / Q_host(w) ] (in bits) While core housekeeping operons maintain low compositional divergence (D_KL ≈ 0.02–0.05 bits, Z < 1.5 SD), recently integrated ICEs carrying foreign AMR cargo (blaNDM-1, blaKPC-2, armA 16S rRNA methyltransferase) exhibit sharp, table-top spikes (5- to 8-fold elevation in D_KL, typically Z = +4.5 to +7.4 SD) accompanied by local %GC skew transitions (ΔGC = 6%–14%) bounded precisely by the tRNA attL and attR direct repeats.

1. Relative Synonymous Codon Usage (RSCU) Shift: For an amino acid with g_a degenerate synonymous codons, RSCU_c = X_c / [ (1/g_a) ∑ X_k ]. Because foreign ICE genes were adapted to the tRNA pool of a previous donor species, their RSCU vector diverges significantly from native host ribosomal/housekeeping genes.

A major mechanism discovered in Pseudomonas ICEclc (Jan Roelof van der Meer and colleagues) and Vibrio SXT elements is epigenetic bistability ("division of labor"): because expressing the 20-gene T4SS mating pilus and excising the chromosome transiently arrests cell division and risks lethal chromosome breakage if the cell divides before re-integration, the ICE remains transcriptionally repressed in 95% to 99% of cells via H-NS-like xenogeneic silencing proteins or SetR/InrR repressors. Only a specialized 1% to 5% subpopulation of "transfer-competent" donor cells activates the excisionase (xis), circularizes the ICE, and assembles the T4SS mating apparatus.

Alarmingly, common clinical empirical therapies directly trigger ICE excision and transfer in patients: administering fluoroquinolones (ciprofloxacin) or trimethoprim–sulfamethoxazole stalls bacterial DNA replication forks, activating the RecA* coprotease filament. RecA* stimulates autocatalytic cleavage of the SXT/R391 repressor SetR (a phage λ cI homolog), lifting repression of the transcriptional activators SetC and SetD and increasing ICE excision and horizontal transfer of blaNDM-1 and sul2/dfrA resistance across hospital enterobacteria by 100- to 1,000-fold.

How does an Integrative and Conjugative Element (ICE) differ from a conjugative plasmid and a bacteriophage?

An ICE is an evolutionary hybrid of a bacteriophage and a conjugative plasmid. Like a lysogenic prophage, an ICE integrates site-specifically into the host bacterial chromosome (using a tyrosine or serine recombinase at attB/attP sites, usually a tRNA gene) and replicates passively with the host chromosome rather than autonomously in the cytoplasm. However, unlike a phage—which packages DNA into icosahedral protein capsids and lyses the host cell—an ICE excises as a circular DNA molecule and transfers single-stranded DNA directly into a living recipient cell through a plasmid-like Type IV Secretion System (T4SS) conjugative pilus.

Why do over 70% of proteobacterial ICEs integrate specifically into the 3′ ends of tRNA genes?

Targeting the 3' end of a chromosomal tRNA gene (tRNA-Gly, tRNA-Leu, tRNA-Ser, tRNA-Asn) provides two major evolutionary advantages: first, tRNA nucleotide sequences are highly conserved across distantly related bacterial families and orders, guaranteeing that an ICE transferred into a new host species will find an identical chromosomal attachment site (attB). Second, because the ICE's own attP site carries an exact copy of the 3' terminal 15–50 bp of that tRNA gene, site-specific integration reconstructs a complete, functional tRNA gene at the integration boundary without knocking out an essential host gene.

Why do ciprofloxacin and tetracycline trigger horizontal transfer of ICE-borne antibiotic resistance genes in hospital patients?

Many ICEs have evolved regulatory sensors that couple their excision and conjugation directly to the presence of clinical antibiotics. In the SXT/R391 family (Vibrio and Enterobacterales), ciprofloxacin and trimethoprim cause DNA replication fork stalling and activate the host RecA SOS response, which cleaves the ICE master repressor SetR and boosts conjugative transfer up to 1,000-fold. Similarly, in Tn916 (Enterococcus/Staphylococcus) and CTnDOT (Bacteroides), sub-inhibitory tetracycline stalls ribosomes on a short leader peptide or activates the TetQ–RteA–RteB sensor cascade, inducing xis/int transcription and driving horizontal spread of resistance across the patient's microbiome.

1. Wozniak, R. A. F., & Waldor, M. K. (2010). "Integrative and conjugative elements: mosaic mobile genetic elements enabling dynamic lateral gene flow." Nature Reviews Microbiology, 8(8), 552–563. https://doi.org/10.1038/nrmicro2382
2. Guglielmini, J., Quintais, L., Garcillán-Barcia, M. P., de la Cruz, F., & Rocha, E. P. C. (2011). "The repertoire of ICE in prokaryotes underscores the unity, diversity, and ubiquity of conjugation." PLOS Genetics, 7(8), e1002222. https://doi.org/10.1371/journal.pgen.1002222
3. Johnson, C. M., & Grossman, A. D. (2015). "Integrative and conjugative elements (ICEs): what they do and how they work." Annual Review of Genetics, 49, 577–601. https://doi.org/10.1146/annurev-genet-112414-055018
4. Beaber, J. W., Hochhut, B., & Waldor, M. K. (2004). "SOS response promotes horizontal dissemination of antibiotic resistance genes." Nature, 427(6969), 72–74. https://doi.org/10.1038/nature02241