Conjugation and Plasmids
Why can a bacterium that has never seen an antibiotic become resistant overnight? The answer often lies in a single act of cellular sex: conjugation. Two bacteria touch, and a bridge forms. DNA flows from one cell into the other. This rewrites the recipient's genetic possibilities within minutes. This is not reproduction — no offspring are created. But it is one of the most consequential events in microbial evolution. Conjugation is the founding case study for Horizontal Gene Transfer. This is the process by which genes move sideways between organisms. They move sideways rather than only vertically from parent to progeny. It matters because bacterial evolution is not solely mutation and selection acting on a fixed genome. Entire functional modules — antibiotic resistance cassettes, virulence factors, metabolic pathways — can be acquired in one step. They can then be immediately put to use. Plasmids, the small extrachromosomal DNA circles that usually carry these traits, are the vehicles. Conjugation is the engine. It is mediated by a physical pilus and a multiprotein secretion apparatus. The story here centers on the F ("fertility") plasmid of Escherichia coli. This is the system where conjugation was first discovered. It remains the best understood. We will trace exactly how a pilus finds a recipient. We will see how a relaxase enzyme nicks and unwinds the DNA. We will follow how a single strand threads through a channel into a new cytoplasm. Meanwhile, the donor simultaneously resynthesizes what it gave up. We will also revisit the classic experiments. These include Lederberg and Tatum's cross-feeding of auxotrophic mutants. They also include Bernard Davis's U-tube. Together, these proved this transfer requires actual physical contact. It is not simply DNA floating through the medium. By the end, you should be able to explain the molecular anatomy of an F plasmid. You should be able to walk through the stepwise mechanism of conjugative transfer. You should be able to cite the evidence that established cell-to-cell contact as obligatory. And you should be able to connect this single mechanism to the global spread of antibiotic resistance. This is a bridge toward thinking about Microbial Ecology and the Microbiome. It invites us to see networks of genetic exchange, not just isolated populations.
- Conjugation transfers DNA from a donor to a recipient bacterium through direct cell contact via a pilus and a protein channel; it requires no free DNA in the environment.
- Plasmids are autonomously replicating circular (usually) DNA molecules separate from the chromosome; the F plasmid is the prototypical conjugative plasmid.
- Transfer begins when relaxase nicks one strand at oriT, and a single strand is pumped 5′-end-first into the recipient through a Type IV secretion system (T4SS).
- Lederberg and Tatum (1946) showed genetic recombination between E. coli strains; Davis's U-tube (1950) proved that physical contact, not diffusible DNA, was required.
- Hfr strains, formed when F integrates into the chromosome, allow timed chromosomal transfer — the basis of interrupted-mating gene mapping.
- Conjugative plasmids are a principal route for spreading antibiotic resistance genes across species and genera, making conjugation central to clinical microbiology today.
Conjugation moves DNA one strand at a time, through a contact-dependent protein channel, from a plasmid-bearing donor into a recipient cell — turning a single genetic innovation into a rapidly spreading trait across a bacterial population.
The core idea
A plasmid is a DNA molecule, typically circular and double-stranded, that replicates independently of the bacterial chromosome. Plasmids range from about 1 kilobase to over 1 megabase in size and can carry a handful of genes or several hundred. They are not required for a bacterium's survival under ordinary conditions — cells cured of their plasmids grow perfectly well in the lab — but they often carry genes that are extremely valuable under specific conditions: resistance to antibiotics or heavy metals, degradation of unusual carbon sources, production of toxins or bacteriocins, or genes for the very apparatus needed for conjugation itself. Plasmids that encode this transfer apparatus are called conjugative plasmids; the F plasmid of E. coli (about 100 kb) is the classical example. Plasmids that lack the full transfer machinery but can still be transferred by hijacking another plasmid's apparatus are called mobilizable.
Conjugation is the process by which a plasmid, or occasionally chromosomal DNA, is transferred from a donor cell to a recipient cell through direct physical contact. It is one of three canonical routes of horizontal gene transfer in bacteria, alongside transformation (uptake of naked DNA) and transduction (transfer via bacteriophage). Conjugation is distinguished by two features: it requires a specialized surface appendage called a pilus to initiate contact, and DNA is transferred through a dedicated protein channel rather than by uptake of free DNA or packaging into a virus particle. The donor cell is defined operationally as F⁺ (carrying F) and the recipient as F⁻ (lacking it); after successful transfer, the former recipient becomes F⁺ itself, capable of acting as a donor in turn. The deeper significance is evolutionary. Conjugation lets a beneficial gene combination spread through a population far faster than mutation and vertical inheritance alone could achieve, and it can cross species and even genus boundaries because many conjugative systems have broad host ranges. This is why a resistance gene that arose in one bacterial lineage can appear, essentially fully formed, in an unrelated pathogen within a clinical setting.
Mechanism — how conjugative transfer actually happens
Anatomy of the F plasmid
The F plasmid is a circular, double-stranded DNA molecule of roughly 100 kb present at one to two copies per cell. About a third of it — the tra operon, comprising some 25 genes — encodes the entire machinery for pilus assembly, DNA processing, and transfer. Key components include:
- traA, encoding pilin, the structural subunit of the F pilus.
- traD, encoding the coupling protein that docks the DNA-processing complex to the membrane channel.
- tra genes encoding the Type IV secretion system (T4SS), a multiprotein channel spanning both membranes of the donor and connecting to the recipient.
- traY/traM and the relaxosome components, which recognize and nick DNA at a specific site.
- oriT</strong> (origin of transfer), a cis-acting DNA sequence where nicking begins — this is not a gene but a site.
F also carries IS (insertion sequence) elements — mobile genetic sequences homologous to sequences scattered around the E. coli chromosome. These homologous regions are what allow F to integrate into the chromosome by homologous recombination, producing an Hfr (high-frequency recombination) strain — a topic returned to below.
Step-by-step transfer
1. Pilus-mediated contact. The F pilus, a hollow filamentous polymer of pilin protein extending several micrometers from the donor cell surface, contacts a recipient cell lacking F (an F⁻ cell, which lacks the surface exclusion proteins that would block re-entry). The pilus tip binds the recipient's outer membrane; contact triggers pilus retraction, drawing the donor and recipient membranes into close apposition, forming a stable mating pair. This is not fusion — it is a docking junction bridged by the T4SS channel.
2. Relaxosome assembly and nicking at oriT. In the donor, a complex of proteins called the relaxosome (TraI acting as relaxase, plus TraY, TraM, and the host integration host factor, IHF) assembles at oriT. TraI's relaxase domain makes a single-strand nick at a specific phosphodiester bond within oriT, becoming covalently attached to the 5′ end of the nicked strand via a tyrosine residue — the same chemistry used by many site-specific recombinases.
3. Strand displacement and rolling-circle replication. The nicked strand is unwound from the duplex. In the donor, DNA polymerase III uses the intact (unnicked) strand as a template to synthesize a new complementary strand — replacing the strand being exported — via rolling-circle replication. The donor therefore retains a complete double-stranded copy of F throughout and after transfer.
4. Translocation through the T4SS channel. The relaxase (TraI), still covalently bound to the 5′ end, pilots the single displaced strand through the coupling protein TraD and into the T4SS channel, which spans the donor's inner and outer membranes, the periplasmic mating-pair junction, and the recipient's outer and inner membranes. DNA moves 5′-end-first — the relaxase leads the way into the recipient cytoplasm, dragging the single strand behind it like a molecular grappling line.
5. Complementary strand synthesis in the recipient. Once inside the recipient, the single strand is circularized (relaxase catalyzes recircularization, the reverse of the initial nicking reaction) and serves as a template for synthesis of its complementary strand, again by host replication machinery. The result: the recipient now possesses a complete double-stranded copy of F.
6. Phenotypic conversion. The new F⁺ cell begins expressing tra genes, assembles its own pilus, and becomes a competent donor — but only after a short delay, because F also encodes proteins (FinO/FinP) that transiently repress tra gene expression, preventing wasteful over-expression of pili in an already F⁺ population. Recipient cells that become F⁺ also acquire surface exclusion proteins (TraS, TraT) that block further entry of DNA from other F⁺ cells, preventing redundant mating. The entire process, from pilus contact to full re-synthesis of double-stranded DNA in the recipient, typically takes only a couple of minutes for the F plasmid itself, though transfer of an entire Hfr-linked chromosome takes about 100 minutes at 37°C — a timing fact so reliable it became a mapping tool (see Worked Examples).
Hfr strains and timed chromosomal transfer
Because F carries IS elements homologous to sequences in the bacterial chromosome, F can occasionally integrate into the chromosome via homologous recombination, converting an F⁺ cell into an Hfr (high-frequency recombination) cell. In an Hfr cell, the oriT is now embedded within the chromosome. Conjugation initiated from an Hfr donor nicks at this integrated oriT and begins transferring chromosomal DNA linearly, starting from a fixed point and proceeding in a fixed direction determined by the orientation of F's integration. Because mating pairs are fragile and typically break spontaneously (or can be broken deliberately by vortexing, as Wollman and Jacob exploited), only a portion of the chromosome — often just the leading genes — is transferred before contact is lost. The very end of F (containing most tra genes) is transferred last, which is why recipients in Hfr crosses only rarely become F⁺ themselves; conjugation is interrupted before the whole F sequence, including its transfer origin's distal end, completes. This asymmetry is the basis of interrupted mating gene mapping (see Worked Examples).
The evidence — proving contact-dependent transfer
Lederberg and Tatum (1946): discovery of bacterial recombination
Joshua Lederberg and Edward Tatum mixed two auxotrophic strains of E. coli K-12, each requiring multiple nutrients absent from minimal medium:
- Strain A: required methionine and biotin (met⁻ bio⁻), but was prototrophic (wild type) for threonine, leucine, and thiamine.
- Strain B: required threonine, leucine, and thiamine (thr⁻ leu⁻ thi⁻), but was prototrophic for methionine and biotin.
Neither strain alone could grow on minimal medium lacking all these supplements — that was the entire point of using multiply auxotrophic strains, so that any colony growing on minimal medium could not be explained by a single spontaneous back-mutation (the probability of independently reverting several mutations simultaneously is astronomically low, roughly the product of individual reversion rates, often less than 10⁻¹⁴). When the two strains were mixed together, plated on minimal medium, and incubated, rare prototrophic colonies (roughly 1 in 10⁷ cells) appeared, able to grow with no supplements at all. This showed that genetic material had been exchanged and recombined between the two strains, producing progeny with a new combination of alleles. It was the first demonstration of genetic recombination in bacteria and won Lederberg (with Tatum) a share of the 1958 Nobel Prize. But the experiment alone did not prove that direct contact was required — the effect was consistent with either contact-dependent transfer or transfer of a soluble, diffusible substance (as had recently been shown for pneumococcal transformation by Avery, MacLeod, and McCarty).
Bernard Davis (1950): the U-tube experiment
Bernard Davis settled the question decisively. He constructed a U-shaped glass tube in which the two arms were separated at the bottom by a fine sintered-glass filter with pores small enough to allow free passage of DNA, viruses, and dissolved medium, but too small to allow bacterial cells to cross.
- One arm was inoculated with one auxotrophic strain, the other arm with a second auxotrophic strain, exactly as in Lederberg's experiment.
- Medium was pumped back and forth across the filter to ensure that anything smaller than a bacterial cell (soluble DNA, phage particles) freely mixed between the two chambers, while the bacteria themselves remained physically confined to their own arm.
- After incubation, cells from each arm were plated separately on minimal medium.
Result: no prototrophic recombinants appeared from either arm. When the same two strains were mixed directly in a single vessel (no filter), prototrophs appeared at the expected frequency.
Conclusion: genetic exchange required actual cell-to-cell contact — it was not achieved by a filterable, diffusible agent. This ruled out transformation (free DNA) and transduction by a filter-passing phage as explanations for the Lederberg–Tatum result and established conjugation as a mechanistically distinct, contact-dependent process. Later work (Hayes, 1952) identified the donor/recipient asymmetry (F⁺/F⁻ and Hfr strains), and electron microscopy by Anderson, Wollman, and others directly visualized the F pilus bridging mating cells, providing the physical structure to match the genetic evidence.
Worked examples
Example 1: Interrupted mating and gene mapping
An Hfr strain transfers its chromosome starting from a fixed origin at a constant rate under standard lab conditions (37°C), taking about 100 minutes to transfer the entire E. coli chromosome. Suppose an Hfr strain donor is prototrophic for genes thr⁺, azi^r (azide resistance), lac⁺, and gal⁺, mated against a recipient that is thr⁻ azi^s lac⁻ gal⁻. Samples of the mating mixture are agitated in a blender at various times to disrupt conjugation, then plated to score which donor markers have appeared in recombinant recipients. Suppose the data are:
| Marker | Time of first appearance (min) |
|---|---|
| thr⁺ | 2 |
| azi^r | 9 |
| lac⁺ | 18 |
| gal⁺ | 25 |
Because transfer proceeds linearly at an approximately constant rate from a fixed origin, the order of marker entry directly gives gene order on the chromosome, and the time differences are proportional to physical map distance (roughly 2 minutes ≈ 1 map unit in classical E. coli linkage maps under these conditions). From this single interrupted-mating experiment we can conclude the gene order is:
origin — thr — azi — lac — gal — (rest of chromosome) with thr and azi about 7 minutes apart, azi and lac about 9 minutes apart, and lac and gal about 7 minutes apart. This is exactly the logic Wollman and Jacob used in the 1950s to build the first genetic map of E. coli, expressed not in centimorgans but in minutes of transfer time — a unit still used in E. coli genetics today, where the full map is calibrated to 100 minutes.
Example 2: A clinical resistance-spread scenario
A hospital ward has an outbreak in which Klebsiella pneumoniae isolates carry a plasmid, call it pKPC, bearing a carbapenemase gene that inactivates carbapenem antibiotics — a last-resort drug class. Sequencing shows the identical plasmid, with >99.9% nucleotide identity, also present in unrelated Escherichia coli isolates from the same ward, even though chromosomal typing shows the E. coli and Klebsiella strains are phylogenetically very distant (different genera).
Reasoning: if resistance had arisen by independent chromosomal mutation in each species, we would expect distinct resistance mechanisms and no shared extrachromosomal element. Instead, the near-identical plasmid sequence in two different genera is the signature of a conjugative plasmid with broad host range, transferred horizontally between species sharing the same ward environment (a phenomenon well documented for IncF and IncN plasmid families carrying bla_KPC and similar genes). The practical implication: infection control must treat resistance genes, not just resistant strains, as the transmissible unit — decolonizing or isolating one species does not prevent spread if the plasmid can move into a different resident species of the patient's normal flora. This is precisely the kind of dynamic that motivates surveillance of plasmid content, not just species identity, in modern clinical microbiology.
Common misconceptions & traps
| Misconception | Why it's wrong |
|---|---|
| "Conjugation is bacterial reproduction/sex in the eukaryotic sense" | No fusion of genomes or meiosis occurs; it is unidirectional DNA transfer, not equal genetic mixing, and does not produce new individuals by itself. |
| "Plasmids are essential for bacterial survival" | Most plasmids are dispensable under normal growth; they confer conditional advantages (e.g., resistance) but often carry a fitness cost absent selection. |
| "All plasmids can transfer themselves" | Only conjugative plasmids encode the full transfer apparatus; many plasmids are non-mobilizable or require a helper conjugative plasmid to be mobilized. |
| "DNA transferred is double-stranded, just like the donor's plasmid" | Only a single strand is transferred through the channel; the complementary strand is synthesized after entry, and the donor also resynthesizes its own lost strand. |
| "F⁺ and Hfr are unrelated strain types" | Hfr arises directly from F⁺ by integration of F into the chromosome via homologous recombination at IS element sites; they are two states of the same plasmid. |
| "Recipients that receive DNA always become full donors" | In Hfr crosses, transfer is usually interrupted before the distal tra genes complete, so recipients typically gain only chromosomal markers, not full F⁺ status. |
The scientist behind it
Joshua Lederberg came to genetics as a medical-school dropout turned graduate student, working with Edward Tatum at Yale in the mid-1940s on the metabolic genetics of E. coli — an organism most geneticists of the era considered genetically inaccessible because bacteria were not thought to undergo anything like sexual recombination. Lederberg's insight was methodological as much as biological: by using strains with multiple, independently arising auxotrophic mutations, he made any recombination event statistically undeniable, since simultaneous reversion of three or more mutations by chance was vanishingly improbable. This design turned an apparently intractable organism into a genetic workhorse. Bernard Davis, working slightly later, brought a physical rigor to the question by physically separating the two populations with a filter — a simple but decisive control that any hand-waving claim about "something in the medium" could not survive. Together, Lederberg's discovery and Davis's proof of mechanism transformed bacteria from organisms presumed to reproduce only clonally into organisms capable of a rudimentary, contact-mediated genetic exchange — reshaping how biologists thought about microbial evolution, and laying groundwork for the entire field of bacterial genetics, including the later mapping work of Hfr strains by Wollman and Jacob at the Pasteur Institute.
Origins · 1 minJoshua Lederberg and how this was discovered
Lederberg co-discovered bacterial conjugation with Edward Tatum, showing that bacteria could exchange genetic material — a foundational result for bacterial genetics and molecular biology, recognized by the 1958 Nobel Prize in Physiology or Medicine (shared with Tatum and George Beadle).
Where this matters today
Conjugation is not a historical curiosity confined to textbook genetics — it is an active, ongoing process shaping medicine, biotechnology, agriculture, and environmental microbiology right now. In clinical settings, conjugative plasmids are the principal vehicle for the horizontal spread of antibiotic resistance genes, including extended-spectrum β-lactamases (ESBLs), carbapenemases (such as bla_KPC and bla_NDM-1), and multidrug-resistance cassettes that combine several resistance genes on a single plasmid. Because conjugation can cross species and genus boundaries, resistance can spread through a patient's gut microbiota, a hospital plumbing biofilm, or an agricultural feedlot far faster than resistance could arise by mutation alone in each lineage independently — a major driver of the global antimicrobial resistance crisis tracked by the WHO and CDC. In biotechnology, engineered conjugative and mobilizable plasmids are workhorse tools for introducing DNA into bacteria that resist standard transformation protocols, including many environmentally and agriculturally important species; conjugation is also exploited to transfer engineered genes into plant-pathogenic or plant-symbiotic bacteria such as Agrobacterium tumefaciens, whose Ti plasmid transfer system (itself a relative of conjugative T4SS machinery) is the basis of much plant genetic engineering. In agriculture and food safety, surveillance of plasmid-borne resistance genes in livestock-associated bacteria informs policy on antibiotic use in animal husbandry, since resistance plasmids can move from commensal gut bacteria in livestock into human pathogens. In environmental and evolutionary microbiology, conjugative plasmids are recognized as a major currency of gene flow within microbial communities — a concept central to modern thinking about Microbial Ecology and the Microbiome, where a community's functional capacity depends not just on which species are present but on which mobile genetic elements are circulating among them. Understanding conjugation mechanistically — pilus, relaxosome, T4SS channel, rolling-circle replication — is therefore not only classic genetics but a live framework for combating resistant infections and engineering microbial genomes.
Ask the tutor: How does the F plasmid's tra operon regulation (FinO/FinP repression) prevent an entire bacterial population from wastefully expressing pili once most cells are already F⁺?
Is conjugation the same thing as bacterial reproduction?
No. Conjugation transfers genetic material from one existing cell to another; it creates no new cell and is not required for bacterial reproduction, which occurs by binary fission. Conjugation is better understood as a mechanism of horizontal gene transfer that reshuffles which genes a cell possesses, sometimes dramatically changing its phenotype (for example, conferring antibiotic resistance), without any division event occurring as part of the transfer itself. The donor cell survives transfer largely unchanged, having resynthesized the strand it exported, and continues its normal life cycle of growth and division exactly as before mating.
Do all plasmids transfer by conjugation?
No. Only conjugative plasmids encode the complete apparatus — pilus, relaxosome, and Type IV secretion system — needed to transfer themselves. Many naturally occurring plasmids are "mobilizable": they carry an oriT and sometimes a relaxase but lack the mating-pair formation genes, so they can only transfer if a conjugative plasmid happens to be present in the same cell to supply the missing machinery in trans. Still other plasmids are entirely non-mobilizable and spread only vertically, through cell division, or occasionally via transformation if released from lysed cells.
How is conjugation different from transduction?
Transduction moves DNA between bacteria using a bacteriophage as the vehicle: phage particles mistakenly package host or plasmid DNA instead of (or along with) their own genome and inject it into a new host cell during a subsequent infection cycle. Conjugation instead requires direct, living cell-to-cell contact mediated by a pilus and a dedicated secretion channel, with no viral particle involved at any stage. Both are forms of horizontal gene transfer, but transduction depends on phage biology and infection cycles, while conjugation depends on the tra gene machinery of a conjugative plasmid or, in some cases, integrative conjugative elements (ICEs) within the chromosome.
Why does the F plasmid's integration into the chromosome matter for genetics?
When F integrates to form an Hfr strain, the chromosome itself becomes conjugatively transferable starting from a fixed origin determined by F's insertion site and orientation. Because transfer proceeds at a roughly constant rate and mating pairs break stochastically over time, the order and timing of marker transfer directly reveal gene order and relative distances along the chromosome. This is the basis of interrupted mating mapping, historically used by Wollman and Jacob to build the classical E. coli genetic map, expressed in minutes of transfer time rather than recombination frequency.
Can conjugation occur between different bacterial species or even genera?
Yes, and this is one of its most consequential properties. Many conjugative plasmids, especially clinically important resistance plasmids of the IncF, IncN, and IncA/C incompatibility groups, have broad host ranges and can transfer between distantly related genera such as Escherichia, Klebsiella, Salmonella, and beyond. This cross-genus mobility is a major reason antibiotic resistance genes spread so rapidly through hospital environments and animal husbandry settings, since a single conjugative plasmid can move a resistance gene into whatever bacterial species happen to share the same niche.
What stops a cell from being re-infected repeatedly by the same plasmid?
F⁺ and Hfr cells express surface exclusion proteins, notably TraS and TraT, embedded in the inner and outer membranes respectively. These proteins interfere with productive mating-pair formation or DNA transfer when another F⁺ cell attempts to conjugate with an already F⁺ recipient, substantially reducing (though not completely abolishing) redundant transfer. This is analogous in logic, though not mechanism, to superinfection exclusion systems seen in some bacteriophages, and it helps a donor population avoid wasting resources on unproductive matings within an already F⁺ population.
Why do plasmid-encoded traits sometimes disappear from a population without antibiotic pressure?
Carrying a plasmid, and especially expressing pili and the tra machinery, imposes a metabolic and sometimes structural cost (pili can be targets for certain phages and can slow growth slightly). In the absence of selection favoring the plasmid's cargo genes (for example, without antibiotic exposure to favor a resistance plasmid), cells that lose the plasmid, or that acquire compensatory mutations reducing its cost, can be outcompeted less, and plasmid-free segregants may gradually rise in frequency. This is why resistance genes can decline, though often slowly and incompletely, when selective antibiotic pressure is removed from a population.