biology2u
Depth
Concept

Horizontal gene transfer

T-043Home BU-203Threads systems · evolution
Statement

Sharing genes between unrelated lineages.

Why it matters

Every result in the Genetics unit up to this point describes purely vertical inheritance: genes pass from parent to offspring within a single lineage, and a population's genome changes only by mutation and by reshuffling existing variation each generation. Horizontal gene transfer breaks that assumption outright. A bacterium can acquire a functioning gene not from a parent cell but from an unrelated, sometimes distantly related, organism sharing its environment, and that gene can become a stable, heritable part of its genome within a single generation. This is the mechanism that makes antibiotic-resistance spread across species boundaries far faster than mutation and selection alone could achieve, and it is why bacterial genomes are frequently mosaic, built from segments with different evolutionary histories.

bacterial-growth-curve already establishes how quickly an exponentially dividing population can reach very large cell numbers; horizontal gene transfer is the process that lets a single successful transfer event, however rare per cell, be multiplied across such a population and fixed in it within days.

Hypotheses
A transferred gene must be stably incorporated into the recipient's genetic material — by recombination into the chromosome or as a self-replicating plasmid — for the event to count as horizontal gene transfer.Transient uptake of DNA that is degraded before replication, or DNA that enters a cell but is never propagated to daughter cells, has no evolutionary consequence and is not itself gene transfer in the sense used here; only heritable acquisition matters. Donor and recipient need not be the same species, or even the same domain of life.This is what distinguishes horizontal transfer from the vertical transmission covered elsewhere in this unit: the gene's history crosses a lineage boundary rather than following a single line of descent from parent to offspring. Recipient cells possess barriers — restriction-modification systems that degrade unfamiliar DNA, and a general requirement for at least local sequence homology for chromosomal integration by recombination — that reduce but do not eliminate the frequency of successful transfer; these barriers are why most transfer attempts fail, not why transfer never occurs.
Proof
1
\[ \text{Transformation: naturally competent cells take up free extracellular DNA released by lysed cells from the environment.} \]
Some bacterial species express uptake machinery that binds and internalises fragments of DNA present in their surroundings; once inside, a fragment can pair with homologous chromosomal sequence and be incorporated by recombination. A
2
\[ \text{Transduction: a bacteriophage mispackages host DNA instead of, or alongside, its own genome, and injects it into the next cell it infects.} \]
viral-replication-cycles already establishes that phages replicate by packaging viral genomes into new particles; transduction is simply the case where the packaging machinery errs and includes a segment of the previous host's chromosome, which is then delivered into an entirely new bacterial cell during a subsequent infection. A
3
\[ \text{Conjugation: direct cell-to-cell contact, typically via a plasmid-encoded pilus, transfers a copy of a plasmid from donor to recipient.} \]
A donor cell carrying a conjugative plasmid (such as the F plasmid) extends a pilus, draws a recipient cell into contact, and transfers a single strand of plasmid DNA across the connection; both cells then synthesise a complementary strand, leaving each with a complete copy. A
4
\[ \text{Acquired DNA becomes heritable either by homologous recombination into the chromosome or as an autonomously replicating plasmid.} \]
In either case the first Hypothesis is satisfied: the gene is now copied along with the recipient's own genetic material at every subsequent division, and is passed to all of that cell's descendants. A
5
\[ \text{Even a low per-cell transfer probability produces many successful events across a large population.} \]
bacterial-growth-curve's exponential phase can carry a population to \(10^9\) cells or more per millilitre; multiplying even a small per-cell-per-generation transfer probability by a population of that size gives a substantial absolute number of newly transformed, transduced, or conjugated cells, each now a founder for further vertical spread of the acquired gene. A
Result
\[ \text{Transformation} \ \cup\ \text{Transduction} \ \cup\ \text{Conjugation} \ \Rightarrow\ \text{genes cross lineage boundaries independent of reproduction} \]

Reading. Three distinct physical mechanisms — free DNA uptake, phage-mediated delivery, and direct pilus-mediated contact — all achieve the same evolutionary outcome: a gene moves into a genome without having been inherited from a parent cell.

Scope. All three mechanisms are documented and common among bacteria and archaea; the phenomenon is comparatively rare, though not entirely absent, in eukaryotes, where nuclear and cell-membrane barriers, and the separation of germline from soma in most animals, make heritable acquisition far less likely.

Corollaries & converses
  • Bacterial genomes are commonly mosaic: different genes within the same genome can have different evolutionary histories, so a phylogenetic tree built from one gene can genuinely conflict with a tree built from another gene in the same organisms — a direct, observable signature of horizontal transfer rather than an analytical error.
  • antibiotic-resistance genes are frequently carried on the same mobile plasmids and integrons that move by conjugation, which is why resistance can appear in a species that was never itself exposed to selection by a given drug, and can spread across a bacterial community far faster than new resistance mutations could arise and spread by vertical descent alone.
  • Converse: a set of genes that all yield mutually consistent phylogenetic trees, agreeing with the organisms' vertically-inherited ribosomal RNA phylogeny, is evidence against recent horizontal transfer of those particular genes, even though it cannot rule out transfer at some point in the more distant past.
Fails without
  • No competence machinery for uptake (transformation): free extracellular DNA is rapidly degraded by environmental and cell-surface nucleases; without a dedicated system to bind and internalise it before that happens, transformation cannot occur, and any gene present only in dead, lysed donor cells is simply lost.
  • No conjugative pilus or transfer apparatus: donor and recipient cells never form the stable mating-pair contact conjugation requires, so a plasmid remains confined to its own lineage; resistance or other traits it carries would then spread only by that lineage's own vertical reproduction and by independent mutation arising separately in every other lineage — precisely the slower situation horizontal transfer is invoked to explain the absence of.
Common errors
  • Treating transformation and transduction as interchangeable terms; transformation is the direct uptake of free environmental DNA, while transduction specifically requires a bacteriophage as an intermediate carrier.
  • Assuming conjugation requires the fusion of two cells into one; it requires only a transient pilus-mediated connection, after which donor and recipient remain two distinct, separate cells.
  • Assuming that DNA acquired by any of these three routes is always integrated into the recipient's chromosome; plasmids very commonly persist and replicate as independent extrachromosomal elements indefinitely, without ever recombining into the chromosome.
  • Treating horizontal gene transfer as exclusive to prokaryotes; it is markedly rarer in eukaryotes but well documented in some lineages, so "rare" is the accurate description, not "absent."
Discussion

Frederick Griffith's 1928 experiments with pneumococcus were the first clear demonstration that a hereditary trait could pass between bacterial strains without conventional reproduction, when a heat-killed virulent strain transformed a live avirulent strain into a virulent one. Oswald Avery, Colin MacLeod, and Maclyn McCarty followed up in 1944, identifying DNA itself, rather than protein, as Griffith's "transforming principle" — a result pivotal in establishing DNA as the hereditary material well before Watson and Crick's 1953 structure. Joshua Lederberg and Edward Tatum discovered bacterial conjugation in 1946, showing that some bacteria could exchange genetic material through direct contact.

Because horizontal transfer is so pervasive among prokaryotes, some evolutionary biologists argue that a single branching tree is not the most accurate representation of prokaryotic evolutionary history at all, and prefer a "web" or reticulated network in which lineages exchange genes as well as splitting from common ancestors. This does not overturn vertical descent as the dominant mode of inheritance even in bacteria, but it means the tree-of-life metaphor, built primarily from eukaryotic and vertical-inheritance intuitions, needs qualification once applied to the microbial world.

Common misconception: that horizontal gene transfer is a rare, exotic exception to normal inheritance. Among bacteria and archaea it is a routine, ongoing process, and it is one of the principal reasons microbial evolution — most visibly the spread of antibiotic resistance — can proceed far faster than mutation-and-selection acting on a single, vertically-inherited lineage.

Worked examples
1
\[ \text{A conjugative plasmid carrying a } \beta\text{-lactamase resistance gene enters a fully sensitive population.} \]
Once a single donor cell successfully transfers the plasmid to one recipient by conjugation, that recipient is now itself a donor: it carries the same conjugative machinery and can transfer the plasmid onward to further sensitive cells it contacts, while every resistant cell also continues to divide vertically. A
2
\[ \text{Resistance therefore spreads by two simultaneous routes: vertical division of already-resistant cells, and horizontal conjugative transfer to still-sensitive cells.} \]
This compounding of vertical growth (bacterial-growth-curve) with horizontal spread is why, once a resistance plasmid enters a bacterial population exposed to the corresponding antibiotic, the resistant fraction of that population can come to dominate far more quickly than waiting for the same resistance mutation to arise independently, by chance, in enough separate lineages. A
\[ \text{resistant fraction grows by division} + \text{conjugative transfer, not division alone} \]

Reading. Horizontal transfer adds a second growth channel for a trait on top of ordinary reproduction, which is exactly why traits spread by conjugation can outrun traits that depend on new mutation.

Scope. The same two-channel logic applies to transformation and transduction, with uptake or infection frequency in place of conjugation frequency as the horizontal term.

Further example. A laboratory culture of E. coli is initially 100% sensitive to tetracycline. A single cell carrying a conjugative plasmid with a tetracycline resistance gene is introduced. The bacterial generation time is 20 minutes. The conjugation frequency is such that each resistant cell transfers the plasmid to one sensitive cell every 40 minutes on average. Calculate the total number of resistant cells after 2 hours (120 minutes), assuming no cell death and that the plasmid is stable.

1
\[ t = 120 \text{ min}, \quad t_{\text{gen}} = 20 \text{ min}, \quad t_{\text{conj}} = 40 \text{ min} \]
We define the time horizon and the two characteristic timescales: vertical division (20 min) and horizontal transfer (40 min). A
2
\[ N_{\text{vert}}(t) = N_0 \cdot 2^{t/t_{\text{gen}}} \]
If only vertical division occurred, the number of resistant cells would double every 20 minutes. This is the baseline growth of the resistant lineage. A
3
\[ \text{Horizontal transfer adds new resistant cells.} \]
Each existing resistant cell also acts as a donor. The rate of new resistant cells generated by conjugation is proportional to the current number of resistant cells. Specifically, each resistant cell produces 1 new resistant cell every 40 minutes via conjugation. A
4
\[ \text{Total growth rate} = \text{Vertical rate} + \text{Horizontal rate} \]
The total increase in resistant cells per unit time is the sum of cells produced by division and cells produced by conjugation. Let \( R(t) \) be the number of resistant cells. The differential equation is \( \frac{dR}{dt} = R \left( \frac{\ln 2}{t_{\text{gen}}} + \frac{1}{t_{\text{conj}}} \right) \). A
5
\[ R(t) = R_0 \cdot \exp\left[ t \left( \frac{\ln 2}{20} + \frac{1}{40} \right) \right] \]
Solving the differential equation with \( R_0 = 1 \) gives the total number of resistant cells. The term in the exponent is the combined growth constant. A
6
\[ R(120) = 1 \cdot \exp\left[ 120 \left( \frac{0.6931}{20} + \frac{1}{40} \right) \right] = \exp\left[ 120 (0.034655 + 0.025) \right] = \exp[120 \cdot 0.059655] \]
Substituting the values: \( \frac{\ln 2}{20} \approx 0.034655 \) min\(^{-1}\) and \( \frac{1}{40} = 0.025 \) min\(^{-1}\). Sum = 0.059655 min\(^{-1}\). Multiply by 120 min = 7.1586. \( e^{7.1586} \approx 1285 \). A
\[ R(120) \approx 1285 \text{ resistant cells} \]

Reading. Horizontal gene transfer significantly accelerates the spread of resistance. Without conjugation, only \( 2^{120/20} = 2^6 = 64 \) resistant cells would exist. Conjugation increases this to ~1285, a 20-fold enhancement.

Scope. This model assumes a large excess of sensitive recipients and that conjugation efficiency is constant. In reality, as resistant cells become abundant, the pool of sensitive recipients shrinks, slowing the horizontal term.

Problems
  1. A microbiologist finds that a gene tree built from one housekeeping gene disagrees with the tree built from the surrounding organisms' ribosomal RNA sequences. What does this suggest, and why?
    SolutionIt suggests that particular gene has a different evolutionary history from the rest of the genome — most plausibly because it was acquired by horizontal transfer at some point rather than inherited purely vertically alongside the ribosomal RNA genes, which are transferred horizontally only rarely and so are commonly used as a more reliable marker of vertical descent (Corollaries).
  2. Explain why bacteriophage-mediated gene transfer (transduction) requires no free-living step for the donor DNA, unlike transformation.
    SolutionIn transduction the donor DNA is packaged directly inside a phage particle during that phage's replication in the previous host cell, and is delivered by injection when the phage subsequently infects a new host; it is never released as naked DNA into the environment, so it is not exposed to extracellular nuclease degradation the way transformation's free DNA is (Fails without, first bullet).
  3. Why does horizontal gene transfer make the spread of antibiotic resistance a population-wide problem rather than one confined to the original mutant lineage?
    SolutionBecause resistance genes carried on conjugative plasmids or other mobile elements can move directly into unrelated, previously sensitive lineages without requiring those lineages to independently acquire the same mutation; combined with the very large population sizes bacteria reach during exponential growth (bacterial-growth-curve), this lets resistance spread across a diverse bacterial community far faster than mutation and vertical selection acting on any one lineage alone (Worked examples).
  4. Bacteria can acquire genes by transformation (uptake of free DNA), transduction (DNA carried by a bacteriophage) and conjugation (direct transfer through a pilus). Design a simple test, using DNase and a filter that lets small particles but not cells pass, to tell which mechanism moves a resistance gene from a donor to a recipient strain.
    SolutionPlace donor and recipient in the two arms of a U-tube separated by a filter that blocks cells (Davis U-tube). If transfer still occurs across the filter, cell contact is not needed, so conjugation is excluded. Then add DNase: if transfer stops, naked DNA was being taken up (transformation); if transfer continues despite DNase, the DNA was protected inside phage capsids (transduction). If no transfer occurs across the filter at all but it does occur when the cells are mixed, the mechanism requires contact (conjugation), which is also insensitive to DNase.
  5. Quantitative: A population of 10,000 sensitive S. aureus cells is exposed to a phage that transduces a methicillin resistance gene. The phage has a burst size of 100 particles per infected cell, and 10% of the phage particles successfully infect a new sensitive cell and transfer the resistance gene. If the initial infection rate is 1 cell per minute, calculate the number of resistant cells after 10 minutes, assuming no vertical division of resistant cells (only horizontal transfer via transduction).
    Solution Step 1: Determine the number of new phage particles produced per minute. Initial infection rate = 1 cell/min. Burst size = 100 particles/cell. Phage production rate = \( 1 \times 100 = 100 \) particles/min. Step 2: Determine the number of successful transductions per minute. Efficiency = 10% = 0.1. Transductions/min = \( 100 \times 0.1 = 10 \) resistant cells/min. Step 3: Calculate total resistant cells after 10 minutes. Since we assume no vertical division of resistant cells, the number of resistant cells is simply the cumulative number of transductions. Total resistant cells = \( 10 \text{ cells/min} \times 10 \text{ min} = 100 \) cells. Note: This is a simplified model. In reality, resistant cells would also divide, and the phage population would grow exponentially. However, the question specifies "assuming no vertical division," so we only count the horizontally acquired cells.