Codominance

In classical Mendelian genetics, a dominant allele's phenotype is the only one seen in the heterozygote — the recessive allele's product exists but is functionally overridden (e.g., because the dominant allele's product is sufficient on its own, or acts as the only version detected by whatever assay defines the phenotype). Codominance describes a different molecular situation: both alleles are transcribed, both are translated into protein, and both protein products carry out their own distinct, independently identifiable function or structure in the same cell or organism. The phenotype of the heterozygote is therefore the union — not the average — of the two homozygous phenotypes. If allele A gives red pigment molecules and allele B gives blue pigment molecules, a codominant heterozygote does not turn purple (that would be incomplete dominance's blending); instead you would see distinct red patches and distinct blue patches, or a surface studded with both red and blue molecules at once. The clean distinguishing test is at the level of the gene product, not the whole-organism color. In incomplete dominance (e.g., red × white snapdragons giving pink F1), a single pigment pathway is running at roughly half throughput in the heterozygote because only one of the two alleles produces functional enzyme, and haploinsufficiency causes an intermediate amount of the same pigment — there is only one type of molecule present, just less of it. In codominance, there are two chemically distinct final products, each made by its own allele independently of the other, and each detectable by an assay that can tell them apart (antibody binding, electrophoretic mobility, distinct staining). Codominance therefore requires — and reveals — that the two alleles are not competing for the same limited catalytic step; they are each running their own complete synthetic or structural program in parallel. A useful terminological caution: dominance is not a fixed label glued to an allele forever. The same pair of alleles can look "codominant" when you assay at the molecular level (e.g., both hemoglobin variants are separable on an electrophoresis gel) but look "dominant/recessive" when you assay at a coarser physiological level (e.g., sickle-cell heterozygotes are clinically nearly normal under everyday oxygen tension, so HbA appears "dominant" for the disease phenotype, even though HbA and HbS are truly codominant at the protein level). This is why geneticists always specify the phenotype level being examined — cellular, biochemical, or whole-organism — before calling an allele pair codominant, incompletely dominant, or dominant/recessive.

Step 1 — Two functional alleles, two functional gene products. At a diploid locus, a heterozygote carries one allele from each parent, say I^A and I^B at the ABO locus. Each allele encodes a distinct, catalytically active version of the same enzyme class: a glycosyltransferase that attaches a specific sugar residue onto a shared precursor molecule (the H antigen) displayed on the red cell surface.

Step 2 — Independent transcription and translation. Both alleles are transcribed from their own promoters on their own chromosome copies; there is no silencing mechanism (no promoter competition, no dominant-negative protein interaction) that shuts one allele down. Both mRNAs are made, both are translated, and both enzyme molecules fold correctly and reach the Golgi where glycosylation occurs.

Step 3 — Both enzymes act on the shared substrate pool. The I^A-encoded transferase adds N-acetylgalactosamine to H antigen, creating the A antigen. The I^B-encoded transferase adds galactose to H antigen, creating the B antigen. In an I^A I^B heterozygote, both enzymes are present in the same Golgi compartment and both act on the shared pool of H-antigen substrate on the same red blood cell membrane — some H antigen molecules get converted to A, others to B, on the very same cell.

Step 4 — Both final products are independently detectable. An anti-A antibody binds A antigen; an anti-B antibody binds B antigen. On an AB heterozygote's red cells, both antibodies bind, at levels comparable to what is seen on A-only or B-only cells respectively. This is the operational definition of codominance: an assay exists that unambiguously detects each allele's product independently, and in the heterozygote, both signals appear at full intensity, not diminished, not blended.

Step 5 — Population and pedigree consequence. Because the heterozygote's phenotype is a distinct third class (not confusable with either homozygote and not intermediate between them on a single continuous scale), codominant loci give three cleanly distinguishable phenotypic classes for a two-allele system, in exact 1:2:1 genotype ratio in an F1 x F1 or backcross-appropriate scheme — mirroring the underlying Mendelian genotype ratio precisely, unlike ordinary dominant/recessive crosses which collapse two genotype classes into one phenotype (giving 3:1).

Step 6 — Why the sugar-transferase mechanism must be codominant, not blended. Crucially, the I^A and I^B enzymes are not competing to modify the same molecule of substrate to different degrees (which would produce a blended, intermediate antigen). They are acting on different individual H-antigen molecules scattered across the same membrane — each single antigen molecule becomes either A-type or B-type, an all-or-none local event, and the cell surface ends up as a mosaic containing genuinely both finished products in roughly equal number. This mosaic, all-or-none character at the molecular level is what guarantees the population-level phenotype is "both present," not "averaged."

Karl Landsteiner and the discovery of blood groups (1900–1901). Landsteiner mixed serum and red blood cells from different colleagues in his laboratory and observed that some combinations caused agglutination (clumping) while others did not. He classified people into groups (later named O, A, B, and — by his students Alfred von Decastello and Adriano Sturli in 1902 — the rarer AB) based on which sera agglutinated which cells. Crucially, individuals of type AB had red cells agglutinated by both anti-A and anti-B sera drawn from type-B and type-A individuals respectively — the direct serological signature of both antigens being present simultaneously on the same cells. This was not an intermediate agglutination reaction; it was full-strength agglutination by both reagents, exactly the operational signature of codominance. Landsteiner received the Nobel Prize in Physiology or Medicine in 1930 for this work, which also made safe blood transfusion possible for the first time.

Bernstein's genetic analysis (1924). Felix Bernstein used population and family blood-typing data to work out mathematically that the ABO locus is controlled by three alleles at a single gene (I^A, I^B, i), rather than by two independently segregating genes as earlier (incorrect) two-locus models had proposed. Bernstein's statistical analysis of large pedigrees and population frequency data showed that the single-locus, three-allele model correctly predicted the observed proportions of each blood type in families and populations, while two-locus models did not. This established the genetic architecture within which codominance of I^A and I^B (and their shared recessiveness relative to i) could be precisely defined — setting the direct foundation for Multiple Alleles and Blood Groups.

Biochemical confirmation (1950s onward). Walter Morgan and Winifred Watkins, along with other chemists, worked out the actual chemical structures of the A, B, H, and O antigens as specific oligosaccharide chains attached to glycoproteins and glycolipids on the red cell surface, confirming directly (by chemical isolation and structural analysis) that A-type and B-type sugar chains are chemically distinct molecules — not different concentrations of the same molecule — and that AB cells carry both sugar structures physically present on the membrane. This chemical evidence closed the loop: serology showed both antigens are detected; biochemistry showed both molecular structures actually exist on the cell.

The MN blood group — an even cleaner codominance demonstration (Landsteiner and Levine, 1927). Landsteiner and Philip Levine discovered a second red cell antigen system, MN, using antisera raised in rabbits. Unlike ABO, the MN system has no dominance complications from a third recessive allele: there are just two alleles, L^M and L^N, each encoding a distinct glycophorin-A surface variant differing by two amino acids. Homozygotes are cleanly type M or type N; heterozygotes are unambiguously type MN, reacting fully with both anti-M and anti-N sera. Because MN has no "recessive" allele complicating the picture (unlike ABO's i allele), it is often used as the pedagogically cleanest possible codominance example, and it obeys simple 1:2:1 Hardy–Weinberg-type genotype ratios in population surveys, exactly as predicted for a two-allele codominant locus.

Protein electrophoresis (mid-20th century). With the rise of starch-gel and later polyacrylamide electrophoresis, geneticists (e.g., studies on human hemoglobin variants by Vernon Ingram and others building on Linus Pauling's 1949 identification of sickle hemoglobin as a distinct molecule) could directly visualize both allelic protein products of a heterozygote as two separate bands on a gel — for example, heterozygous carriers of the sickle-cell allele (HbA/HbS) show two distinct hemoglobin bands on electrophoresis, at roughly equal intensity, providing a direct physical/biochemical demonstration of codominance at the protein level, independent of any antibody-based serology. This became (and remains) the standard laboratory technique for scoring codominant molecular markers — including modern isozyme and DNA-based markers used in mapping and forensics.

Example 1 — A three-generation ABO/MN pedigree cross. A woman of blood type AB, MN genotype, marries a man of blood type O, M genotype. Predict the blood type and MN distribution among their children, and compute the exact probability of a child being simultaneously type A and type N. ABO locus: Mother is I^A I^B (type AB). Father is type O, so he is ii (both alleles recessive, no A or B antigen made). Cross: I^A I^B × ii. Gametes from mother: ½ I^A, ½ I^B. Gametes from father: all i. Offspring genotypes: ½ I^A i (type A), ½ I^B i (type B). Note: no child can be type AB or type O from this particular cross, since the father contributes only i. MN locus: Mother is L^M L^N (type MN, heterozygous, codominant). Father is L^M L^M (type M, homozygous). Cross: L^M L^N × L^M L^M → gametes from mother ½ L^M, ½ L^N; gametes from father all L^M. Offspring genotypes: ½ L^M L^M (type M), ½ L^M L^N (type MN). No child can be type N. Combine the two independent loci (different chromosomes, so independent assortment applies):

The probability of a child being type A and type N is zero — not because of any dominance subtlety, but simply because neither parent can contribute an L^N-only combination that yields pure N (father has no L^N allele at all, so type N, which requires L^N L^N, is genetically impossible in this family). This is a useful trap: students sometimes assume codominant loci can produce "any" combination; in reality the cross is still governed by ordinary segregation, and only the genotypes actually present in the parents can appear in the offspring.

Example 2 — Paternity exclusion using a codominant marker panel. A child is blood type AB, MN. The mother is blood type A, type N. Two men are candidates for paternity: Man 1 is type B, type M; Man 2 is type O, type MN. Which man is excluded, and why, using codominance logic? Step 1 — ABO logic. The child is AB, so the child must have received one I^A allele and one I^B allele (one from each parent, since AB requires exactly one of each and neither parent can be homozygous ii and still contribute the needed allele). The mother is type A: she could be I^A I^A or I^A i. Either way she can supply the I^A allele the child needs. The child's I^B allele must therefore have come from the father. Man 1 (type B) could be I^B I^B or I^B i — either genotype can supply an I^B allele. Man 2 (type O) is ii and cannot supply an I^B allele under any circumstance, since type O individuals carry no I^B allele at all. Man 2 is excluded by the ABO locus alone. Step 2 — MN logic, done independently as a cross-check. The child is MN (heterozygous, so carries one L^M and one L^N). The mother is type N, genotype L^N L^N, so she can only contribute L^N — consistent with the child's L^N allele. The child's L^M allele must have come from the father. Man 1 is type M, genotype L^M L^M — he can supply L^M. This is consistent with Man 1 being the father. (Man 2, already excluded by ABO, is type MN and could also have supplied an L^M allele — but this is now moot, since ABO evidence alone is a complete exclusion, and exclusion at even one codominant, well-characterized locus is sufficient to rule out paternity.) Conclusion: Man 2 is excluded; Man 1 remains consistent with (though not proven as, on blood groups alone) the biological father. This example shows precisely why codominant markers are forensically powerful: because heterozygotes are unambiguous and distinct from either homozygote, every genotype in the pedigree can be read directly off the phenotype with no ambiguity about hidden recessive alleles — unlike ordinary dominant/recessive traits, where a "type A"-like dominant phenotype could conceal either a homozygous or heterozygous genotype.

Karl Landsteiner (1868–1943) was an Austrian physician and immunologist trained in Vienna, working at a time when transfusion medicine was plagued by mysterious, sometimes fatal reactions that no one could predict. Rather than accepting blood as a uniform tissue, Landsteiner suspected individual biochemical variation among people's red cells and sera. He tested this directly and methodically: drawing blood from himself and coworkers in his laboratory, separating serum from cells, and cross-mixing every possible combination, watching for clumping under the microscope. This simple, rigorous, combinatorial experiment revealed the ABO groups and, decades later, MN and Rh groups — work that transformed transfusion from a dangerous gamble into a safe, routine procedure, and along the way gave genetics one of its cleanest natural demonstrations of codominant gene action.

Transfusion medicine and organ transplantation. ABO and Rh typing before every transfusion and transplant directly depends on the codominant expression of A and B antigens; mismatched transfusions trigger dangerous antibody-mediated hemolysis, so accurate genotype-to-phenotype reasoning (exactly the logic worked out above) remains a bedside life-or-death calculation performed millions of times per year worldwide.

Forensic genetics and paternity testing. Modern forensic panels have moved from blood-group serology to codominant DNA markers — short tandem repeats (STRs) — but the underlying logic is identical to the ABO/MN paternity example above: because both alleles at a codominant locus are independently visible (as two distinct-sized DNA fragments on a gel or capillary trace), every genotype in a pedigree can be read unambiguously, making codominant markers the gold standard for identity and kinship testing, disaster victim identification, and criminal investigation.

Genetic mapping and marker-assisted breeding. Codominant markers (isozymes, RFLPs, microsatellites, SNPs scored by direct sequencing) are prized in linkage mapping and agricultural marker-assisted selection precisely because heterozygotes are distinguishable from both homozygous classes, giving three (not two) scorable genotype classes per marker and therefore much higher mapping resolution and breeding precision than dominant markers, which conflate heterozygotes with one homozygote.

Hematology and disease diagnosis. Codominant expression of hemoglobin variants (e.g., HbA/HbS carriers, HbA/HbC carriers) is directly visualized by hemoglobin electrophoresis or HPLC in newborn screening programs, allowing clinicians to distinguish silent carriers from patients with clinically significant compound genotypes, guiding genetic counseling and clinical management.

Immunohematology research and antigen engineering. Understanding that A and B antigens arise from two independently acting glycosyltransferases (rather than competing enzymes) has informed enzymatic "blood type conversion" research, in which bacterial or fungal enzymes strip A or B sugars from donor red cells to create universal-donor type O cells for transfusion — a direct biotechnological application of the codominance mechanism described above.

Ask the tutor: How does the ABO system's mixed dominance (I^A, I^B codominant, but both dominant over i) set up the general rules for reasoning about Multiple Alleles and Blood Groups with three or more alleles at one locus?