Minor Blood Group Systems: MNS, Kell, Duffy, and Others

The MNS system was first described by Landsteiner and Levine in 1927, making it one of the earliest blood group systems discovered after ABO. It is controlled by two adjacent genes on chromosome 4: GYPA, encoding glycophorin A, and GYPB, encoding glycophorin B. Both glycophorins are sialoglycoproteins that project from the red cell surface. The amino acid sequence at the N-terminal end of glycophorin A determines whether a person types as M, N, or MN. The sequence of glycophorin B determines the S and s antigens.

Because the two genes are closely linked, GYPA and GYPB alleles are usually inherited together as haplotypes. The four common haplotypes are MS, Ms, NS, and Ns. A person inherits one haplotype from each parent, giving ten possible genotype combinations, though the phenotypic distinction between homozygous and heterozygous states (for example MM versus MN) is detectable only by testing with both anti-M and anti-N antisera. Anti-M and anti-N are often naturally occurring IgM antibodies found in humans who have never been transfused, which made them relatively accessible reagents. Anti-S and anti-s, by contrast, are mostly immune antibodies requiring immunised donor sources, which limited their availability in some laboratories.

In forensic casework, MNS typing contributed meaningfully to discrimination because the MN heterozygote class is common, so a homozygous MM or NN result on a stain excludes roughly 70 to 72 percent of donors. When S/s data were also available, the combined MNS phenotype narrowed the contributor pool further. The main limitation was antigen degradation: the sialoglycoprotein antigens are sensitive to bacterial contamination and heat, and older or wet-stored bloodstains sometimes gave no readable result, particularly for S and s.

The Kell system is encoded by a single gene, KEL, on chromosome 7. It contains more than 30 recognised antigens, but the two most forensically relevant are K (antigen 001, also called Kell) and k (antigen 002, also called Cellano). These are antithetical: a person carries K, k, or both. The K antigen is present in approximately 9 percent of Europeans, making the genotype KK rare (under 0.2%) and Kk the dominant K-positive phenotype (about 8.8%). The large majority of the population, roughly 91 percent of Europeans, is kk.

The K antigen is highly immunogenic: a single transfusion or pregnancy can stimulate anti-K production. For forensic purposes, this immunogenicity was irrelevant to the typing procedure but clinically important if a laboratory worker handling donor antisera was regularly exposed. Anti-K antisera used in blood typing were obtained from transfusion recipients or multiparous women who had made the antibody naturally. The antibody is IgG and works best in an indirect antiglobulin (Coombs) test, which added a step to the typing protocol but was manageable in a reference laboratory.

Forensically, the rarity of K antigen was the key statistic. If a bloodstain typed as K-positive, the contributor pool was immediately restricted to approximately 9 percent of the population. When combined with ABO and Rh, a K-positive O-negative sample could narrow the contributor fraction substantially. Conversely, a K-negative result was uninformative on its own because 91 percent of people are K-negative, but it remained part of the cumulative probability calculation across all systems.

The Kidd system is controlled by the SLC14A1 gene on chromosome 18, which encodes a urea transporter protein on the red cell membrane. The two main antigens are Jk(a) and Jk(b), antithetical alleles. Three common phenotypes exist: Jk(a+b-), Jk(a+b+), and Jk(a-b+). A rare null phenotype, Jk(a-b-), exists but is seen almost exclusively in Polynesian and Finnish populations and was rarely encountered in UK or US casework.

The Kidd antigens present a particular challenge for forensic serology. Anti-Jk(a) and anti-Jk(b) are typically IgG antibodies that fix complement, and the antigens are known for dosage effect: cells from a homozygous Jk(a+b-) donor react more strongly with anti-Jk(a) than cells from a heterozygous Jk(a+b+) donor. This dosage effect is generally a clinical concern in transfusion medicine, but it also means that weak reactions in absorption-elution tests on bloodstains needed careful interpretation. A faint result could reflect a genuine heterozygote rather than a failed test or a negative.

The Kidd antigens are also notoriously labile. They degrade rapidly under conditions of heat, storage, and bacterial contamination. Forensic laboratories routinely reported that Kidd typing was the first result to be lost in aged or compromised stains. Where it was obtainable, the Jk(a-b+) phenotype (approximately 23 percent of Europeans) or the Jk(a+b-) phenotype (approximately 26 percent) provided useful discrimination; the most common phenotype, Jk(a+b+), at roughly 50 percent of Europeans, contributed less to exclusion on its own.

The Duffy system is encoded by the ACKR1 gene (formerly DARC) on chromosome 1. The gene product is the Duffy antigen receptor for chemokines, a multi-pass transmembrane protein. The two main antigens are Fy(a) and Fy(b), antithetical alleles. In Europeans, the three common phenotypes are Fy(a+b-), Fy(a+b+), and Fy(a-b+), with approximate frequencies of 17, 49, and 34 percent respectively.

The Duffy system has a distinctive population genetics dimension. In West African and African-American populations, a promoter-region variant silences ACKR1 expression in red cells, producing the Fy(a-b-) phenotype. This null phenotype reaches frequencies of 65 to 99 percent in sub-Saharan African populations. The biological significance is that Fy(a-b-) red cells resist invasion by Plasmodium vivax malaria, giving the null phenotype a strong selective advantage in endemic regions. For forensic serologists, this population frequency difference meant that a Fy(a-b-) result on a bloodstain was informative about the likely ancestry of the contributor in a statistically crude but practically useful way, particularly in investigations predating DNA profiling.

Antisera for Fy(a) and Fy(b) were immune IgG antibodies from transfused or pregnant individuals. Anti-Fy(a) was more readily available commercially because anti-Fy(a) producers are more common, whereas anti-Fy(b) was harder to obtain and more variable in quality. Both antigens survive reasonably well in dried bloodstains compared to Kidd, making Duffy typing somewhat more reliable on older samples.

The central principle of multi-system blood group typing is that unlinked gene loci segregate independently, so phenotype frequencies multiply. If a stain types as blood group O (frequency 44% in a European population), Rh-negative (frequency 15%), MN (50%), K-negative kk (91%), Jk(a+b+) (50%), and Fy(a+b+) (49%), the combined frequency is approximately 0.44 x 0.15 x 0.50 x 0.91 x 0.50 x 0.49, which is roughly 0.74 percent. This means only about 1 in 135 people in a European population could be the source of that stain based on blood type alone.

Serologists expressed this in court as either a probability of exclusion (in this example, approximately 99.26 percent of donors in the reference population would be excluded) or as a likelihood ratio comparing the probability of the evidence if the suspect is the source versus if a random person is the source. The terminology varied by jurisdiction and era, but the underlying calculation was the same. UK courts from the 1970s onward heard blood group evidence expressed in these terms, as did US courts, and the conceptual infrastructure was then adopted wholesale for DNA profiling when STR typing arrived in the late 1980s.

The main constraints on the statistical strength of blood group panels were antigen degradation, antiserum availability, and mixed stains. A mixed bloodstain, from two contributors, produced a combined phenotype that could not be separated into individual profiles by serological methods alone. The analyst could detect that more antigens were present than a single person could carry (for example, both M and N, S and s, K and k all present), but could not determine the proportions contributed by each donor. This limitation was fundamental to the method and was fully superseded only by DNA STR profiling with its ability to detect and at times separate mixed DNA contributors.

Fresh whole blood can be typed by direct agglutination: add antiserum to a suspension of red cells and read the clumping reaction. Dried bloodstains on fabric or other surfaces cannot be tested this way because the cells are lysed, aggregated, and often degraded. Three main techniques were adapted for stain work.

Absorption-elution is the most widely applied method for minor blood group antigens on stains. A cutting from the stained fabric is incubated with a specific antiserum at 4 degrees Celsius; if the target antigen is present, antibodies bind to it. The cutting is then washed thoroughly to remove unbound antibody. The temperature is raised to 56 degrees Celsius to elute the bound antibody from the antigen. The resulting eluate is mixed with fresh typed red cells; agglutination of those cells indicates that the eluate contains the specific antibody and therefore that the original stain expressed the antigen. The method is sensitive but requires careful temperature control and a negative control stain to confirm specificity.

Mixed agglutination exploits a different principle. Antiserum is absorbed onto the stain debris directly, then fresh indicator red cells expressing the antigen are added. If antibody is bound to the stain material, the indicator cells agglutinate in a characteristic rosette pattern around the stain particles, visible under a microscope. Mixed agglutination was particularly useful for ABO typing of epithelial cells and was adapted for some minor antigen work, though it was less standardised than absorption-elution for the Kell and Duffy systems.

The transition from blood group serology to DNA profiling in casework was gradual, not abrupt. Many laboratories ran blood group panels and DNA typing in parallel through the early 1990s. Blood group results appeared in court alongside early DNA evidence in cases such as R v Adams (UK) and in multiple US appellate decisions of the same period. The shift was driven by the far higher discriminating power of STR DNA profiles, not by any deficiency in the serological methods on their own terms. Serological evidence, where it was obtained correctly and interpreted with population-appropriate frequency tables, remained admissible and was never systematically overturned by subsequent DNA testing. The forensic serology record in this era was, on balance, reliable within its stated statistical limits.