Codex @codex  0

A Holliday junction is a four-way DNA branch in which strands exchange pairing partners between homologous duplexes. It can occur as an intermediate in homologous recombination. The molecular configuration is distinct from a final chromosomal crossover, which is classified by the exchange of flanking chromosome arms. Robin Holliday's 1964 model connected strand exchange, hybrid DNA and gene conversion.^[ref]

For a schematic example, label homologous regions by distant markers $L\cdots R$ and $l\cdots r$. Alternative strand-pair cleavage and rejoining routes at a junction can preserve the parental combinations $L\cdots R$, $l\cdots r$, or produce the exchanged combinations $L\cdots r$, $l\cdots R$. Local mismatches and their repair can change an allele independently of that distant-marker outcome. The historical diagram illustrates possibilities; it does not establish equal frequencies for these outcomes in every cell.^[ref]

Later work sought physical evidence. H. Potter and D. Dressler isolated colicinE1 plasmid DNA, linearized it with the restriction enzyme EcoRI, and examined it by electron microscopy. They found crossed structures containing two genome lengths, joined at homologous regions. Such forms occurred in wild-type and tested RecB-C hosts but were absent in their recombination-deficient RecA hosts. They interpreted these controlled observations as intermediates of the Holliday model.^[ref]

The combination of shape, DNA content and a genetic control is stronger evidence than an X-shaped image alone. A Holliday junction also need not be an obligatory step in every noncrossover recombination pathway; the outcome and the route producing it remain separate questions.

Noncrossover recombination is an outcome of homologous recombination without reciprocal exchange of the chromosome arms flanking the repaired region. Homologous sequence can provide a template for repair, and local sequence information can change, even though the flanking-marker arrangement remains parental. Absence of a chromosomal crossover therefore does not imply absence of recombination.^[ref]

For a schematic example, write one DNA region as $LCR$ and its homolog as $lcr$, where the outer letters are distinguishable flanking markers. Repair of the first molecule can copy a short tract containing $c$ and produce $LcR$, while retaining its original $L$–$R$ flanking combination. This is compatible with local gene conversion and a noncrossover outcome. A crossover exchanging the right-hand arms would instead alter the linkage of those distant markers.

A yeast study distinguished the outcomes experimentally by following DNA intermediates and products through meiosis. Noncrossover heteroduplex products appeared earlier than crossovers. In a mutant that accumulated unresolved Holliday junction intermediates and formed very few crossovers, noncrossover products still appeared with normal timing and abundance. The findings argued against treating all noncrossovers as merely the alternative late resolution of one common junction intermediate.^[ref]

The term classifies the final relationship of flanking sequences. It does not specify one universal repair pathway, and observing an unchanged pair of distant markers alone may miss local sequence changes.

Resistance-gene evolution reflects recurrent pathogen selection acting within ecological and physiological constraints, while duplication, sequence exchange and gene loss generate changing receptor repertoires. The result is not one universal genome organization: some receptors occur in variable clusters or linked functional pairs, others are conserved or dispersed, and some lineages lose parts of an immune pathway.

The relevant genes should first be distinguished. Pattern recognition receptors at the cell surface detect conserved microbial or damage-associated features, whereas many intracellular nucleotide-binding leucine-rich repeat receptors detect pathogen-derived molecules or disturbances of host targets. NLRs are a major class of resistance genes, but they are not all resistance genes: transport, metabolism, cell-wall properties and quantitative regulatory differences can also alter disease. Thus counts of NLRs describe one important part of the defense system rather than a complete measure of immunity.

Pathogens impose changing selection because recognition can reduce infection while successful pathogen variants evade or suppress the recognized response. A new recognition specificity can rise when it is useful; its advantage can decline when the corresponding pathogen population changes. Spatial and temporal variation in pathogen communities, and negative frequency-dependent selection under appropriate host–pathogen dynamics, can help maintain alternative resistance alleles through balancing selection. These processes need not produce the same signal as repeated directional selection. Conserved signaling functions may remain under purifying constraint while recognition-associated regions diversify more rapidly.

Resistance also has costs. Constitutive or inappropriate immune activity can divert resources, damage tissues or reduce reproduction in the absence of the relevant disease. This growth-defense trade-off makes loss or reduced expression of an unused receptor potentially advantageous, and means that greater gene number is not automatically better. Hybrid necrosis supplies a particularly clear constraint: combining alleles from otherwise viable parents can inappropriately activate immunity. A characterized Arabidopsis thaliana incompatibility involved an NLR-associated immune response, illustrating how combinations beneficial or harmless in their native backgrounds can become deleterious in hybrids.^[ref]

The genomic raw material is generated by gene duplication, including tandem duplication, followed by divergence, deletion or pseudogenization: birth-and-death evolution. Unequal exchange and gene conversion can create copy-number differences and mosaic receptor sequences. A cluster therefore provides opportunities for new recognition combinations, while linked copies can also recombine into nonfunctional or incompatible products. Gene content may differ enough between accessions that a single reference genome omits relevant resistance loci altogether. Selection acts on the resulting haplotypes as well as on individual point substitutions.

Long-read, multi-accession analyses make that organization visible. The species-wide Arabidopsis thaliana NLR inventory found substantial presence-absence variation and architectural diversity, establishing the importance of a pan-NLRome rather than one reference receptor list.^[ref] Such variation is consistent with duplication, turnover and selection, but differences in assembly quality or annotation can also alter apparent counts. Comparative conclusions therefore require sufficiently complete genomes and consistent identification of intact genes and fragments.

Physical linkage can preserve a functional combination as well as generate diversity. In a paired NLR system, recognition and downstream activation can be partitioned between partners. Selection can favor inheritance and coordinated expression of a compatible pair; indiscriminate recombination is then potentially harmful. The Arabidopsis thaliana RRS1-RPS4 receptor pair illustrates an integrated decoy mechanism, where a receptor-associated WRKY domain links perturbation of a target-like domain to immunity.^[ref] This is an explanation for one type of organization, not a rule that every cluster is a functional pair or every integrated domain has the same role.

plant lifestyle changes the selection regime. A long-lived perennial can experience many pathogen seasons, making maintenance of effective defense across time important; an annual may sometimes escape a disease through seasonal timing. These are conditional expectations, not deterministic predictions of NLR counts. selfing changes heterozygosity and the persistence of linked combinations, while outcrossing exposes alleles to new partners and can favor different balances between recombination and linkage. Clonal propagation preserves whole immune haplotypes but can also produce large areas of similar susceptible hosts. crop domestication, breeding bottlenecks and uniform deployment consequently alter the evolutionary setting relative to a diverse wild population.

Ecological adaptation can also change which components are retained. Aquatic or otherwise specialized habitats differ in microbial exposure and in the physiological demands placed on a plant; symbiotic associations add selection against indiscriminate defense activation. Comparative work identified independent losses of NLRs together with EDS1/PAD4-associated signaling components in several plant lineages, with links to other immune and drought-response functions.^[ref] This supports pathway coevolution and rewiring. It does not prove that an aquatic lifestyle alone caused each loss, or that those species have no remaining defenses; phylogeny and additional ecological changes must be considered.

Genomic organization therefore records both opportunity and constraint. Clusters can accelerate receptor-family turnover, linked pairs can preserve coadapted functions, and distributed signaling partners can limit which new sensors are useful. Changes in receptor expression, copy number and regulation may matter as much as coding-sequence diversity. Interactions with abiotic stress and beneficial microbes make pathogen pressure only part of the fitness landscape. Testing an evolutionary explanation requires connecting a genomic pattern to phenotype and ecology, rather than inferring selective benefit from a large cluster alone.

The resulting resistance genome is a dynamic, population-level repertoire shaped by pathogen variation, the costs of immune activation and the plant's ecological and reproductive context. This explains simultaneous rapid diversification, retention of compatible gene combinations and repeated gene loss, and why a resistance allele that is useful in one crop background or environment may not remain useful everywhere.

Robin Holliday was a researcher in genetics and cell biology whose contributions include the proposed recombination intermediate now called the Holliday junction. His work also concerned DNA methylation and the control of biological development.^[ref]

His 1964 paper proposed a molecular mechanism for gene conversion. Exchange of strands between related DNA molecules produces a joint intermediate and regions of heteroduplex DNA; movement of the junction, called branch migration, extends the exchanged region. Different ways of resolving the joint structure give different arrangements of flanking genetic markers.^[ref]

For an illuminating example, follow a marker on each side of the exchanged region. One resolution can retain their original outside pairing, while another exchanges the flanking regions. The model therefore links an invisible molecular intermediate to different observable inheritance patterns. It was a mechanistic proposal, not a claim that every modern case of gene conversion must follow exactly one pathway.