Deletions and Duplications
In 1917, Calvin Bridges noticed something odd in his Drosophila stocks. A recessive eye-color allele, Bar, sometimes seemed to behave as if it were dominant. The mystery was not a new allele. It was a missing chromosome segment. This unmasked a recessive gene that had no partner to compete with. Chromosomes are not indivisible units. They can lose pieces (deletions, also called deficiencies). They can also gain extra copies of pieces (duplications). These structural changes rewrite the "dose" of genes a cell carries. Dose matters enormously. A cell used to two copies of a gene often cannot function normally with one copy, or three. Deletions and duplications alter gene number rather than gene sequence. Because of this, they reveal biology that point mutations cannot. They expose recessive alleles hiding on the intact homolog. They map genes to physical chromosome bands with exquisite precision. They also explain a class of human congenital syndromes. These are caused not by a mutant gene but by an altered chromosomal dose. This article explains where deletions and duplications come from at the molecular level. It shows how their phenotypic consequences arise (pseudodominance and dosage sensitivity). It also shows how geneticists exploited them, before DNA sequencing existed, to build detailed chromosome maps. The same logic let Bridges and later Theodosius Dobzhansky and others pinpoint genes to visible chromosome bands. This logic underlies clinical cytogenetics used today. It helps diagnose disorders like cri-du-chat syndrome and Charcot-Marie-Tooth disease type 1A.
- A deletion removes a chromosomal segment; a duplication adds an extra copy of one, altering gene dosage without changing gene sequence.
- Both arise mainly from unequal crossing-over between misaligned repeats, or from chromosome breakage and faulty repair.
- Pseudodominance: a recessive allele on the intact homolog is expressed because the deletion removed its dominant partner — proof a gene has been physically lost.
- Heterozygous deletions cause synapsis loops in meiosis (visible as an unpaired "buckle"), a cytological signature used to map deleted genes to chromosome bands.
- Deletion mapping crosses a set of overlapping deletions against recessive alleles to localize genes without sequencing, decades before molecular tools existed.
- Duplications drive gene family evolution (e.g., globin genes) but also cause dosage-sensitive disease (e.g., CMT1A from PMP22 duplication); large deletions/duplications underlie human syndromes like cri-du-chat and DiGeorge syndrome.
Deletions and duplications change how many copies of a gene a cell has, and it is this altered dose — not any new mutant sequence — that produces their distinctive genetic and clinical signatures.
The core idea
A deletion (deficiency) is the loss of a chromosomal segment, ranging from a few base pairs to many megabases spanning dozens of genes. A duplication is the presence of an extra copy of a segment, which may sit adjacent to the original (tandem duplication), be inserted elsewhere on the same chromosome, or reside on a different chromosome entirely. Both are classified as structural chromosomal mutations, distinct from point mutations (which change sequence within a gene) and from aneuploidies (which change whole-chromosome number). Deletions and duplications change gene dosage: the number of functional copies of each gene in the affected region, without necessarily altering the DNA sequence of the genes that remain. The genetic consequences follow directly from this dosage change. An organism heterozygous for a deletion has only one copy of every gene in the deleted region — it is hemizygous there, exactly as a male is hemizygous for X-linked genes. Any recessive allele on the homologous, non-deleted chromosome is now unopposed and is expressed, a phenomenon called pseudodominance. This is the single most diagnostic genetic signature of a deletion: a cross that "should" give a recessive phenotype only in 1/4 of offspring instead gives it in half, because the deletion-bearing chromosome contributes nothing to mask the recessive allele. Duplications produce the mirror-image dosage problem — an extra copy raises the amount of gene product, which for dosage-sensitive genes (those whose phenotype depends on precise stoichiometry, often transcription factors, ion channels, or components of multiprotein complexes) is itself pathogenic. Both classes of variant are visible cytologically. Homozygous deletions and duplications are frequently lethal or severely deleterious — most of the informative genetics is done with heterozygotes, where the deleted or duplicated chromosome is paired at meiosis with a structurally normal homolog. Because the two homologs differ in length or content over the affected region, the chromosome cannot align gene-for-gene along its entire length. The homolog compensates by looping out the unpaired material, producing a visible deletion (compensation) loop or duplication loop in the synaptonemal complex — first exploited cytologically in the giant polytene chromosomes of Drosophila salivary glands, whose banding pattern is fine enough to see individual missing or duplicated bands under the light microscope.
Mechanism
How deletions and duplications arise. The dominant mechanism is non-allelic homologous recombination (NAHR), also called unequal crossing-over. Chromosomes carry many repeated sequences — transposable elements, segmental duplications, tandem gene clusters. During meiotic prophase I, homologous chromosomes (or, on rare occasions, sister chromatids) can misalign so that non-corresponding copies of a repeat pair up instead of the true allelic copies. If crossing-over then occurs within this misaligned pairing, one recombinant chromatid gains the segment between the repeats (a duplication) and the reciprocal chromatid loses it (a deletion). Because this requires only that repeats be similar enough to mispair, chromosomal regions rich in repeats (segmental duplications, Alu elements, low-copy repeats) are recombination "hotspots" for recurrent deletions and duplications — this is why particular human microdeletion/microduplication syndromes recur at the same genomic coordinates in unrelated patients. A second route is chromosome breakage. Ionizing radiation, certain chemicals, or replication stress can cause double-strand breaks; if repair machinery (non-homologous end joining or microhomology-mediated end joining) rejoins broken ends incorrectly, or if a broken fragment lacking a centromere is lost during cell division, the result is a terminal or interstitial deletion. Breakage-fusion-bridge cycles, in which a chromosome lacking a telomere fuses with a sister, forms a dicentric bridge, and breaks again at anaphase, can generate deletions and duplications repeatedly across cell generations — a process important in cancer genome instability.
How the phenotype follows. Consider a heterozygote carrying one normal chromosome and one chromosome with an interstitial deletion spanning genes A–*B–C*. At synapsis, the normal homolog cannot find a partner for A, B, C along the deleted chromosome, so it buckles outward into a single-stranded loop while the flanking regions pair normally. If the normal homolog carries a recessive loss-of-function allele of B (call it b), the organism has only the b allele functioning in that region — the deleted chromosome contributes nothing — so the recessive phenotype is expressed despite the organism being nominally "heterozygous." This pseudodominance is the operational test for a deletion.
Dosage sensitivity. Not all genes respond the same way to copy-number change. Many genes tolerate having only one functional copy (haplosufficient) with no discernible phenotype — this is why heterozygous deletions of most single genes are silent carriers. Other genes are haploinsufficient: one functional copy does not make enough product to sustain a normal phenotype, so losing one copy — even with the other copy fully normal — causes disease. Symmetrically, some genes are dosage-sensitive to gain: an extra copy from a duplication disrupts stoichiometric balance in a pathway or complex, producing disease from triplosensitivity. Genes encoding parts of large stoichiometric assemblies (ribosomes, transcription factor complexes, myelin components) are disproportionately dosage-sensitive in both directions, which is why segmental duplications and deletions of the same chromosomal region can each cause distinct clinical syndromes.
The evidence
Bridges (1917) and the discovery of pseudodominance via Bar. Calvin Bridges, working in T. H. Morgan's fly lab, studied the X-linked Bar eye mutation, which reduces the normally round Drosophila eye to a narrow bar shape. Bridges found stocks in which a recessive allele at a nearby locus began appearing in flies that should have been heterozygous — behaving, in his words, as though "uncovered." Cytological examination (aided later by salivary gland chromosome squashes, pioneered by Painter and exploited extensively by Bridges himself in the 1930s) showed that some Bar chromosomes actually carried a small physical deletion removing the neighboring gene entirely, and separately he documented that Bar itself often arose as a tandem duplication of a chromosomal band (region 16A on the X chromosome), which further duplicated (Ultrabar) to worsen the phenotype in exact proportion to the number of extra copies — an early, precise demonstration of a dosage effect: eye facet number fell in a graded, roughly stepwise manner with one, two, or three copies of the duplicated region. This was one of the first direct demonstrations that a phenotype could depend on gene dose rather than on any altered gene sequence.
Bridges' deletion mapping in Drosophila. Bridges and later Dobzhansky formalized the use of deletions to map genes. The logic: if a strain heterozygous for a deletion and a recessive allele shows the recessive phenotype (pseudodominance), the deleted region must include that gene's locus. By assembling a panel of Drosophila strains bearing different, partially overlapping X-chromosome deletions (each precisely visible as a missing band or set of bands in polytene chromosome squashes), and testing each deletion against a battery of recessive markers, researchers could localize each recessive gene to the smallest chromosomal interval common to all deletions that uncovered it — and absent from all deletions that did not. This method, requiring no molecular biology at all, placed dozens of Drosophila genes onto physical chromosome maps with band-level resolution decades before DNA sequencing existed, and its logic (deletion-mapping / deficiency-mapping) remained a standard tool in bacterial and phage genetics (e.g., Benzer's fine-structure mapping of the rII locus in bacteriophage T4) and in early human gene mapping using deletion syndromes.
McClintock and the discovery of the ring chromosome / breakage-fusion-bridge cycle. Barbara McClintock's studies of broken maize chromosomes in the 1930s–40s showed directly, cytologically, that a chromosome lacking a telomere after breakage will fuse with a sister chromatid, form a dicentric bridge at anaphase, and break again — repeatedly generating new deletions and duplications across successive cell divisions. This established chromosome breakage and faulty rejoining, not just meiotic misalignment, as a mechanistic source of structural variants, and it is the same breakage-fusion-bridge instability now recognized as a driver of copy-number changes in tumor genomes.
Human cytogenetics: Lejeune and cri-du-chat (1963). Jérôme Lejeune, who three years earlier had shown that Down syndrome is caused by trisomy 21, next showed that a distinctive human syndrome — characterized by a high-pitched, cat-like infant cry, intellectual disability, and characteristic facial features — was associated not with an extra chromosome but with a partial deletion of the short arm of chromosome 5 (5p-). This was among the first demonstrations that loss of a chromosomal segment, rather than loss or gain of a whole chromosome, produces a distinct, reproducible human clinical syndrome, directly extending Bridges' deletion logic from flies to humans and founding clinical deletion cytogenetics.
Worked examples
Example 1 — pseudodominance reveals a hidden deletion in Drosophila. A researcher crosses a wild-type female, heterozygous for a recessive autosomal allele cn (cinnabar eye color; genotype cn⁺/cn), to a male from a mutagenized stock whose visible phenotype is wild-type but whose chromosome structure is unknown. Ordinary Mendelian expectation for cn⁺/cn × cn⁺/cn is 3/4 wild-type : 1/4 cinnabar among F1. Instead, among 400 F1 offspring, the researcher observes:
| Phenotype | Observed | Mendelian 3:1 expectation |
|---|---|---|
| Wild-type | 196 | 300 |
| Cinnabar | 204 | 100 |
Roughly a 1:1 ratio, not 3:1. This is the signature of pseudodominance: half the offspring inheriting the mutant father's chromosome received a chromosome that is deleted for the cn⁺ region rather than carrying a normal cn⁺ allele. Any offspring that also inherited the mother's cn allele has no functional cn⁺ anywhere in the genome (hemizygous), so cn is expressed — explaining why cinnabar appears in about half the progeny instead of a quarter. Cytological examination of the father's polytene chromosomes would be expected to show a missing band at the cn locus, confirming the deletion directly. Note the diagnostic feature: the ratio shifts from 3:1 toward 1:1, not because dominance relationships changed, but because one "allele" is physically absent.
Example 2 — deletion mapping localizes an unknown gene to a chromosome interval. A geneticist has four Drosophila X-chromosome deletion strains (Df1–Df4) with known, overlapping cytological extents across bands 10A–10F, and tests each against three uncharacterized recessive mutations (m1, m2, m3) by crossing deletion-heterozygous females to mutant males and scoring F1 males (hemizygous for the maternal X) for pseudodominant expression.
| Deletion | Bands removed | m1 uncovered? | m2 uncovered? | m3 uncovered? |
|---|---|---|---|---|
| Df1 | 10A–10C | Yes | No | No |
| Df2 | 10B–10D | Yes | Yes | No |
| Df3 | 10D–10F | No | Yes | Yes |
| Df4 | 10C–10E | No | Yes | No |
Reasoning band by band:
- m1 is uncovered only by deletions containing band 10A–10B (Df1, Df2) and not by Df3/Df4 (which lack 10A–10B) → m1 maps to 10A–10B.
- m2 is uncovered by Df2, Df3, and Df4, all of which share only band 10D in common → m2 maps to 10D.
- m3 is uncovered only by Df3, which uniquely includes bands 10E–10F beyond what Df2 and Df4 cover → m3 maps to 10E–10F.
This is exactly the logic Bridges and Dobzhansky used: the gene must lie within the intersection of all deletions that uncover the recessive phenotype, and outside any deletion that fails to uncover it. With enough overlapping deletions, resolution approaches single-band precision — comparable, in an era before sequencing, to a modern deletion panel used in bacterial or yeast fine-structure mapping.
Common misconceptions & traps
| Misconception | Correction |
|---|---|
| "A deletion is just a bigger point mutation." | A deletion removes an entire segment (possibly many genes); a point mutation changes sequence within one gene. Their genetic signatures (pseudodominance vs. simple recessiveness) differ fundamentally. |
| "Pseudodominance means the allele became dominant." | The allele's dominance relationship never changed. It is expressed only because its partner locus is physically absent, not masked. |
| "Duplications are always harmless extra copies." | Dosage-sensitive genes can be as harmed by an extra copy (triplosensitivity) as by a missing one; duplication of PMP22 causes CMT1A disease. |
| "Homozygous deletions are just 'more of the same' as heterozygous ones." | Homozygous deletion of any gene needed for viability is typically lethal; almost all deletion genetics is done in heterozygotes, which is why deletion loops and pseudodominance matter so much. |
| "All chromosome regions tolerate copy-number change equally." | Repeat-rich, dosage-sensitive regions (segmental duplications, gene clusters encoding stoichiometric complexes) are hotspots for recurrent deletion/duplication syndromes; other regions tolerate one or three copies with no phenotype. |
The scientist behind it
Calvin Blackman Bridges (1889–1938) worked as an undergraduate assistant, then full research scientist, in Thomas Hunt Morgan's legendary "Fly Room" at Columbia University, alongside Alfred Sturtevant and Hermann Muller. Bridges was renowned for an almost obsessive precision at the microscope — he was the first to exploit the giant polytene chromosomes of Drosophila salivary glands for cytological mapping, matching visible chromosome bands to genetic loci with a rigor that transformed genetics from an abstract statistical science into one anchored in visible chromosome structure. His 1916 discovery of chromosomal nondisjunction had already proven the chromosome theory of heredity by showing exceptions to Mendelian ratios tracked exactly with visible chromosomal abnormalities; his subsequent work on Bar eye duplications and X-chromosome deletions extended that same logic — abnormal ratios correlate with abnormal chromosomes — to structural rearrangements, establishing deletion mapping as a rigorous genetic tool still conceptually identical to CNV (copy-number variant) mapping used in human genomics today.
Origins · 1 minCalvin Blackman Bridges and how this was discovered
Where this matters today
Clinical cytogenetics now diagnoses dozens of human microdeletion and microduplication syndromes, many recurring at the same genomic coordinates because they arise by NAHR between flanking segmental duplications — DiGeorge syndrome (22q11.2 deletion), Williams syndrome (7q11.23 deletion), and Smith-Magenis / Potocki-Lupski syndromes (reciprocal 17p11.2 deletion and duplication of the same interval, respectively, producing distinct but related phenotypes). Charcot-Marie-Tooth disease type 1A, the most common inherited peripheral neuropathy, is caused by a duplication of PMP22; the reciprocal deletion of the same region causes a milder neuropathy (HNPP), a clean human demonstration that dosage in both directions is pathogenic for the same gene. Diagnostic laboratories detect these with array comparative genomic hybridization (array CGH) or FISH, methods that are the direct molecular descendants of Bridges' band-by-band cytology. In agriculture and plant breeding, controlled duplications and translocation stocks are used to map genes onto specific chromosome arms in crops, echoing classical Drosophila deletion mapping. In cancer biology, tumor genomes are riddled with somatic deletions (frequently inactivating tumor-suppressor genes such as TP53 or RB1 via loss of the remaining functional copy — "second hit" logic building on dosage) and duplications/amplifications (activating oncogenes such as ERBB2/HER2, whose amplification is directly targeted by trastuzumab therapy in breast cancer). In evolutionary biology, ancient gene duplications are the raw material for new gene function — the human globin gene cluster, with its tandem paralogs expressed at different developmental stages, is a textbook product of duplication followed by divergence, a process this article's mechanism directly explains and that foreshadows how still larger structural rearrangements — Inversions and Translocations — further reshape chromosome architecture and its genetic consequences.
Ask the tutor: How does deletion mapping logic extend to mapping genes using overlapping inversions or translocation breakpoints?
How can you tell, in a cross, whether you're seeing pseudodominance versus a simple recessive allele?
Look at the ratio and confirm with cytology. A simple recessive cross (heterozygote × heterozygote or heterozygote × homozygous recessive) gives the standard Mendelian ratios (3:1 or 1:1) predicted by two functioning alleles competing. Pseudodominance instead produces ratios consistent with only one functional allele being present in some offspring — often shifting an expected 3:1 to something close to 1:1 — because the "other" chromosome is deleted for that locus rather than carrying a wild-type allele. Definitive confirmation comes from cytology: examining polytene chromosomes (in Drosophila) or using FISH/array CGH (in humans) to show a physically missing band or segment coinciding with the locus in question.
Why are heterozygous deletions often viable while homozygous deletions are usually lethal?
A heterozygote retains one intact copy of every gene in the deleted region on its normal homolog, so essential genes are still expressed, just at reduced (haploid) dose — tolerable for haplosufficient genes, though it may unmask recessive alleles (pseudodominance) or cause disease if any gene in the region is haploinsufficient. A homozygote (or hemizygote, e.g., a male with an X-linked deletion) has zero functional copies of every gene in the region. If the deletion spans even one gene essential for development or viability, there is no backup copy anywhere in the genome, and the organism typically dies during embryogenesis — this is why large homozygous deletions are rarely recovered and most deletion genetics studies heterozygotes.
What is a "compensation loop" and why does it form?
When one homolog carries a deletion and its partner is structurally normal, the two chromosomes cannot align gene-for-gene across the deleted interval, since the normal chromosome has genetic material with no counterpart to pair with. During meiotic synapsis, the normal homolog's extra material loops outward, unpaired, while the flanking regions on both chromosomes continue to pair normally. This loop is directly visible under the microscope, especially in organisms with large, banded polytene chromosomes like Drosophila salivary glands, and its presence and extent can be matched to specific missing chromosome bands — the cytological basis of deletion mapping.
How does a duplication differ mechanistically from simple gene amplification seen in some cancers?
Mechanistically they can arise the same way — unequal crossing-over, replication errors, or breakage-fusion-bridge cycles all generate extra copies of a segment. The term "duplication" is generally used for a single extra copy (or a modest, stably inherited number) present in the germline or as a heritable structural variant, while "amplification" typically describes many extra copies (sometimes dozens) arising somatically within a cell lineage, often in cancer, through repeated cycles of instability rather than a single crossover event. Both are dosage phenomena: amplified oncogenes like ERBB2/HER2 in breast cancer overproduce a growth-signaling protein for exactly the same reason a duplicated PMP22 overproduces myelin protein — extra gene copies mean extra product.
Can deletions and duplications be adaptive rather than harmful?
Yes, over evolutionary time. While large deletions and duplications are frequently deleterious in an individual (through haploinsufficiency or dosage imbalance), gene duplication is also a principal engine of evolutionary innovation. A duplicated gene copy is initially redundant, freeing one copy from the full weight of purifying selection; it can then accumulate mutations that generate a new function (neofunctionalization) or split the ancestral function between the two copies (subfunctionalization). The vertebrate globin gene family — with distinct paralogs expressed at embryonic, fetal, and adult stages — is a textbook example of ancient duplications whose descendants diverged into specialized, non-redundant genes.
Why do certain human microdeletion syndromes recur at exactly the same chromosome region in unrelated patients?
Because the deletion breakpoints are not random — they are determined by the underlying DNA sequence. Regions flanked by large, highly similar segmental duplications (low-copy repeats) are prone to non-allelic homologous recombination, in which the flanking repeats on misaligned chromosomes recombine and delete (or duplicate) everything between them. Since the flanking repeats are at fixed genomic coordinates, this mechanism reproducibly deletes or duplicates the same interval in different individuals, explaining why syndromes like DiGeorge (22q11.2) or Williams syndrome (7q11.23) recur with strikingly similar breakpoints across unrelated patients rather than at random locations.