Incomplete Dominance

In classical Mendelian dominance, the heterozygote (Aa) is phenotypically indistinguishable from the homozygous dominant (AA); one functional allele produces enough gene product to reach a physiological threshold, so the recessive allele's absence goes unnoticed. Incomplete dominance breaks that assumption. Here the heterozygote is phenotypically distinct from both homozygotes and lies between them on a continuous or graded scale — red RR, pink Rr, white rr flowers being the textbook case. Geneticists denote the alleles without dominant/recessive capitalization conventions (often using superscripts like $C^{R}$ and $C^{W}$) precisely because neither allele masks the other; each contributes proportionally to the phenotype. Crucially, incomplete dominance is a statement about phenotype, not genotype. The underlying Mendelian mechanics — segregation of two discrete alleles at meiosis, independent assortment, random fertilization — proceed exactly as Mendel described. What changes is the mapping from genotype to phenotype. A monohybrid cross of two heterozygotes still produces offspring in the genotypic ratio 1 AA : 2 Aa : 1 aa. In ordinary dominance, that maps to a 3:1 phenotypic ratio because AA and Aa look alike. In incomplete dominance, AA, Aa, and aa are all phenotypically different, so the 1:2:1 genotypic ratio surfaces directly as a 1:2:1 phenotypic ratio. It is essential to distinguish incomplete dominance from two things it is often confused with. First, it is not Codominance, where the heterozygote does not blend but instead simultaneously expresses both parental phenotypes distinctly and separately (like AB blood type showing both A and B antigens, or roan cattle hair showing individual red and white hairs side by side rather than pink). Second, and more importantly historically, it is not "blending inheritance," an older, incorrect hypothesis (associated loosely with pre-Mendelian and Darwinian-era thinking) proposing that hereditary material itself physically mixes like fluids, so that traits would dilute irreversibly each generation and variation would eventually disappear. Incomplete dominance produces a blended appearance while the alleles themselves remain whole, discrete, and unaltered — they segregate cleanly again in the next generation, which is exactly why an F2 cross of pink x pink snapdragons regenerates red and white flowers, something true blending inheritance could never explain.

The molecular basis of incomplete dominance is almost always a matter of gene dosage. Most genes encode enzymes, structural proteins, transporters, or receptors whose functional abundance determines how strongly a phenotype is expressed. A diploid AA genotype produces two working copies' worth of gene product; aa (a loss-of-function or null allele) produces none or a nonfunctional variant; Aa produces roughly half the normal amount. For many traits, half the normal dose is still sufficient to cross whatever physiological threshold produces the full phenotype — this is Mendelian dominance, and the underlying phenomenon is called haplosufficiency. But for traits where the phenotype scales continuously with the amount of gene product — where there is no such buffering threshold — the heterozygote's roughly 50% dose produces a roughly 50% phenotype. This is haploinsufficiency in its broad, quantitative sense: one functional allele is insufficient to produce the full wild-type phenotype, yielding an intermediate result rather than a threshold effect.

Step by step, using flower color as the model: 1. The $C^R$ allele encodes a functional enzyme in the anthocyanin biosynthesis pathway (in real snapdragon genetics this pathway converts colorless precursors into red anthocyanin pigment). 2. The $C^W$ allele is a loss-of-function variant — a mutation disrupting the enzyme's active site or its promoter, so no functional enzyme, or much-reduced enzyme, is produced from that allele. 3. In genotype $C^RC^R$, both alleles transcribe and translate functional enzyme; cells accumulate a full dose of pigment-synthesizing capacity, driving anthocyanin concentration up to a level the eye reads as deep red. 4. In genotype $C^WC^W$, no functional enzyme is made at all; the biosynthetic pathway stalls at the colorless precursor stage; petals appear white (reflecting all wavelengths, absorbing none). 5. In genotype $C^RC^W$, only one allele transcribes functional enzyme. Enzyme concentration in petal cells is roughly halved relative to $C^RC^R$. Because pigment concentration in this system scales close to linearly with enzyme dose over the relevant range, pigment output is roughly halved too — enough anthocyanin to tint the petal pink, but not enough to reach full red saturation. 6. The visual "blend" therefore is a direct, quantitative readout of enzyme dosage — it is chemistry, not some mysterious averaging of hereditary fluids. This dosage logic generalizes far beyond flower pigments. In human familial hypercholesterolemia, the LDLR gene encodes the LDL receptor that clears cholesterol-carrying LDL particles from blood. Individuals with two functional alleles clear LDL efficiently; heterozygotes with one functional and one loss-of-function allele express roughly half the normal receptor density on cell surfaces, leading to intermediate (roughly double normal) LDL cholesterol levels and moderately elevated cardiovascular risk; homozygous loss-of-function individuals — extremely rare — have essentially no functional receptors, extremely high LDL cholesterol, and severe childhood-onset atherosclerosis. The clinical phenotype scales with receptor dose across all three genotypes: a textbook incomplete dominance in a life-and-death context.

The second key mechanistic point is that incomplete dominance is fully compatible with, and indeed predicted by, Segregation and Independent Assortment. Nothing about meiosis changes. Each parent still passes exactly one allele per gamete; alleles do not physically fuse, dilute, or degrade; and a cross of two pink (Rr) heterozygotes still generates gametes in a 1:1 ratio of R and r, which combine at fertilization according to the standard Punnett square logic to give a 1 RR : 2 Rr : 1 rr genotypic ratio. The only thing that has changed relative to a classic dominant/recessive trait is the phenotype-mapping function: instead of a step function (any R present → red), the phenotype-mapping function is closer to linear in allele dose (0 R alleles → white, 1 R allele → pink, 2 R alleles → red).

The discovery of incomplete dominance is inseparable from the rediscovery of Mendel's own work in 1900. Three botanists — Hugo de Vries, Carl Correns, and Erich von Tschermak — independently rediscovered Mendel's 1866 paper and confirmed his ratios in their own plants. Carl Correns, working with the four-o'clock plant Mirabilis jalapa and later described comparably in snapdragons (Antirrhinum majus), performed the cross that became the textbook demonstration of blended dominance.

The experiment. Correns crossed a true-breeding red-flowered Mirabilis jalapa line with a true-breeding white-flowered line. Under a naive "traits are always cleanly dominant/recessive" expectation extrapolated wholesale from Mendel's pea work, one would predict all F1 offspring to resemble one parent (say, all red, if red were dominant). Instead, every single F1 plant bore uniformly pink flowers — a phenotype found in neither parent. This alone was striking: it ruled out simple dominance of either color allele. The decisive test came in the F2 generation. Correns self-pollinated the pink F1 plants and scored the flower color of the resulting offspring. The F2 population segregated in almost exactly the following proportions:

This 1:2:1 phenotypic ratio was the crucial evidence, for two reasons. First, it exactly matched the underlying 1:2:1 genotypic ratio predicted by segregation of a single gene with two alleles — confirming that a blended trait still obeys Mendel's law of segregation, rather than requiring some entirely different hereditary mechanism. Second, and just as important, the reappearance of pure red and pure white flowers in the F2 generation directly refuted classical blending inheritance. If hereditary material genuinely blended like fluids — as some pre-Mendelian models proposed — then once red and white were mixed into pink, no amount of further crossing should ever regenerate pure red or pure white; the "blend" would be permanent and would simply dilute further each generation. Instead, red and white reappeared crisply, in exactly Mendelian proportions, proving that the red allele and the white allele had each remained physically intact and discrete inside the pink F1 heterozygotes, ready to segregate apart again at meiosis. The blend was purely phenotypic; the genotype had never blended at all. Later molecular work (well into the twentieth century, after the discovery of DNA structure and gene expression) supplied the mechanistic confirmation: crosses of this kind typically involve one allele encoding an active enzyme in a pigment pathway and the other allele being a null or hypomorphic mutation, exactly the gene-dosage picture described above. The classical breeding data told geneticists that alleles segregate intact even when phenotypes blend; molecular biology later explained why the blend occurs — a dose-dependent biosynthetic pathway rather than any mixing of hereditary substance. A complementary, quantitative line of evidence comes from human familial hypercholesterolemia, studied extensively by Michael Brown and Joseph Goldstein beginning in the 1970s. They showed by direct receptor-binding assays on cultured skin fibroblasts that cells from heterozygous patients bound and internalized LDL particles at roughly half the rate of cells from unaffected homozygotes, while cells from rare homozygous patients showed almost no LDL receptor activity at all. Plasma LDL cholesterol levels tracked this same three-tier pattern (normal, intermediate, extreme), giving a direct biochemical demonstration — receptor number counted per cell, not just phenotype scored by eye — that heterozygote intermediacy reflects literal gene dosage. This work, which earned Brown and Goldstein the 1985 Nobel Prize in Physiology or Medicine, extended the logic of Correns's flower-color evidence into human molecular medicine.

Example 1 — a snapdragon dihybrid-style cross with unequal starting stocks. Suppose a breeder crosses a true-breeding red, tall snapdragon ($C^RC^R\,TT$) with a true-breeding white, dwarf snapdragon ($C^WC^W\,tt$), where flower color shows incomplete dominance ($C^R$/$C^W$) and stem height is completely dominant (tall $T$ over dwarf $t$), and the two genes assort independently on different chromosomes. F1 genotype: $C^RC^W\,Tt$ — phenotype: pink, tall (100% of F1, since both parents are homozygous at both loci). F1 × F1 self-cross. Each F1 parent produces gametes in a 1:1:1:1 ratio: $C^RT$, $C^Rt$, $C^WT$, $C^Wt$ (independent assortment of the two genes). A 4×4 Punnett square gives 16 equally likely genotype combinations. Collapsing to phenotype classes using multiplicative independence:

• Flower color segregates 1 red : 2 pink : 1 white (from the $C$ locus alone).
• Height segregates 3 tall : 1 dwarf (ordinary dominance at the $T$ locus).

Combined phenotypic ratio (multiply the two independent ratios): $(1:2:1) \times (3:1)$ over 16 total parts:

So F2 phenotypic ratio is 3 red-tall : 1 red-dwarf : 6 pink-tall : 2 pink-dwarf : 3 white-tall : 1 white-dwarf (total 16 parts) — the familiar 9:3:3:1 dihybrid pattern is replaced by a 3:1:6:2:3:1 pattern purely because the color locus now generates three phenotype classes instead of two. This demonstrates that incomplete dominance at one locus simply changes how many phenotypic categories that locus contributes; independent assortment of the second, ordinary-dominance locus proceeds unaffected.

Example 2 — a clinical pedigree reasoning problem (familial hypercholesterolemia). A woman is heterozygous for a loss-of-function LDLR allele (genotype $Rr$, where $R$ = normal receptor allele, $r$ = null allele); she has moderately elevated LDL cholesterol and early-onset coronary disease risk, consistent with roughly half-normal LDL receptor density. Her partner is an unaffected homozygote ($RR$) with no family history of hypercholesterolemia. Cross: $Rr \times RR$. Gametes from the affected parent: $R$ or $r$, each with probability 1/2. Gametes from the unaffected parent: $R$ only (probability 1). Offspring genotypes: 1/2 $RR$ (unaffected, normal LDL clearance), 1/2 $Rr$ (heterozygous, intermediate LDL, moderately elevated cardiovascular risk). Expected phenotype ratio among children: 1 unaffected : 1 heterozygous-affected, i.e., each child independently has a 50% chance of inheriting the condition — a straightforward autosomal pattern once you recognize the trait's dosage-sensitive, incompletely dominant nature (importantly, this pedigree pattern — roughly half of offspring affected, both sexes equally, affected individuals in every generation — is the classic signature geneticists use clinically to diagnose autosomal incomplete dominance/dosage-sensitive disorders, distinguishing them from recessive conditions that can skip generations entirely). If, hypothetically, two heterozygous carriers had children together (both parents with familial hypercholesterolemia, $Rr \times Rr$), the genotypic ratio would return to the classic 1 $RR$ : 2 $Rr$ : 1 $rr$ — and because the rare $rr$ homozygotes suffer the severe, often fatal childhood form of the disease with LDL cholesterol several-fold higher than either parent, this cross illustrates starkly why incomplete dominance is not a cosmetic curiosity but a matter of measurable, life-threatening biochemical dose.

Carl Erich Correns (1864–1933) was a German botanist who trained in Munich under Karl von Nägeli — ironically the same mentor whose skepticism and poor advice may have discouraged Gregor Mendel decades earlier from pursuing wider recognition of his own pea-hybridization results. Correns spent the 1890s conducting his own hybridization experiments on maize and other plants, independently arriving at ratios and segregation patterns strikingly similar to Mendel's, before he discovered, while preparing his own results for publication in 1900, that Mendel had already described the same laws in an obscure 1866 paper. Rather than claim sole priority, Correns explicitly credited Mendel and helped bring Mendel's work to the attention of the scientific community, alongside the near-simultaneous independent rediscoveries by Hugo de Vries in the Netherlands and Erich von Tschermak in Austria. Correns went on to direct research at the newly founded Kaiser Wilhelm Institute for Biology in Berlin and made further foundational contributions to genetics, including some of the earliest clear evidence for cytoplasmic (non-Mendelian, maternally transmitted) inheritance in plants such as four-o'clocks with variegated leaves — a phenomenon now understood to arise from organellar (chloroplast/mitochondrial) DNA rather than nuclear genes. His snapdragon and four-o'clock flower-color crosses, however, remain his most widely taught legacy, providing the clean, reproducible 1:2:1 data that made incomplete dominance a standard part of the genetics curriculum within a few years of Mendel's principles becoming widely known.

Clinical medicine and dose-sensitive disease. Familial hypercholesterolemia remains the paradigm human disease taught alongside snapdragons precisely because it is quantitatively dosage-sensitive: heterozygotes (roughly 1 in 200–250 people in most populations) have intermediate LDL receptor activity and intermediate but clinically significant cardiovascular risk, while the rare homozygotes have drastically higher LDL and early, severe atherosclerosis. Recognizing incomplete dominance in a pedigree — an intermediate phenotype segregating in every generation rather than skipping generations — helps clinicians distinguish autosomal dosage-sensitive disorders from classic recessive conditions and informs genetic counseling about the roughly 50% transmission risk from an affected heterozygous parent.

Plant and animal breeding. Ornamental horticulture exploits incomplete dominance directly: pink snapdragons, pink four-o'clocks, and roan-like intermediate colors in various ornamental cultivars are marketed varieties deliberately maintained as heterozygotes, since selfing or crossing them regenerates the full red/pink/white spectrum useful for breeding programs. Livestock breeders similarly track intermediate coat-color and other dose-sensitive traits (distinguished carefully from true codominance, such as roan cattle coat) when planning matings for predictable offspring phenotype ratios.

Pharmacogenomics and dosage sensitivity more broadly. The core lesson — that heterozygous gene dosage can produce a measurable intermediate phenotype rather than being silently masked — underlies modern understanding of haploinsufficiency in developmental disorders (where losing even one functional copy of certain dosage-sensitive genes causes disease, an extreme case bordering on incomplete dominance) and informs how clinical geneticists interpret single-copy loss-of-function mutations detected by genomic sequencing, rather than assuming all recessive-acting mutations are harmless in heterozygous carriers.

Conservation and quantitative trait breeding. Conservation geneticists modeling trait variation in small populations must account for dosage-sensitive loci when predicting phenotype distributions after inbreeding or outcrossing, since incompletely dominant loci contribute additively (rather than in a masked, threshold fashion) to quantitative variation — a bridge from single-gene Mendelian logic toward the polygenic, additive models used in quantitative and population genetics.

Ask the tutor: Can you show me how incomplete dominance and codominance would each play out differently in the same three-genotype Punnett square?