8.1 Kell System
Key Takeaways
- K is present in about 9% of people of European ancestry and about 2% of African Americans and is the most immunogenic red-cell antigen after D.
- Anti-K is IgG, always clinically significant, and can cause severe HDFN by suppressing erythropoiesis on early progenitors — monitor with MCA-PSV, not bilirubin curves alone.
- The KEL glycoprotein carries K/k, Kpa/Kpb, and Jsa/Jsb; immunized Ko (Kell-null) people can make anti-Ku and need Ko red cells.
- McLeod is an XK-gene (Xp21) defect: missing Kx, weakened Kell, acanthocytes, and chronic granulomatous disease when a contiguous deletion also removes CYBB.
- DTT and AET destroy Kell antigens; ficin and papain do not — that DTT pattern is how you separate Kell from Kidd.
8.1 Kell System
Quick Answer: K (KEL1) is present in about 9% of people of European ancestry and about 2% of African Americans. It is the most immunogenic red-cell antigen after D. Anti-K is IgG, always clinically significant, and can cause severe HDFN by suppressing erythropoiesis, not only by hemolyzing circulating cells. The KEL glycoprotein carries K/k, Kpa/Kpb, and Jsa/Jsb. The Ko (null) phenotype can make anti-Ku. McLeod is an XK-gene problem: missing Kx, weakened Kell, acanthocytes, and, in males with a contiguous XK–CYBB deletion, chronic granulomatous disease. DTT and AET destroy Kell antigens; proteolytic enzymes do not. Use DTT to separate Kell from Kidd.
The June 9, 2026 BB outline lists Kell as II.B.7. The exam does not want a catalog of every KEL number. It wants you to treat anti-K like anti-D, to recognize Ko and McLeod as different nulls, and to know that DTT knocks Kell out of a panel while Kidd stays put.
K, k, and why K punches above its frequency
KEL on chromosome 7 encodes a single-pass zinc endopeptidase (an endothelin-converting-enzyme family member) that carries the Kell antigens. The protein is disulfide-linked to the XK protein. That physical partnership is why a missing XK protein (McLeod) weakens every Kell antigen even though the KEL gene is intact.
K (KEL1) and k (KEL2) are antithetical. A methionine-to-threonine substitution at position 193 creates K instead of the high-prevalence k. Approximate antigen-positive frequencies you will be handed as distractors:
| Antigen | European ancestry | African ancestry | Clinical note |
|---|---|---|---|
| K | ~9% | ~2% | Low frequency, very high immunogenicity |
| k | ~99.8% | ~100% | High-prevalence; anti-k is rare |
| Kpa | ~2% | Rare | Low-frequency; antithetical to Kpb |
| Kpb | >99.9% | >99.9% | High-prevalence |
| Jsa | <0.1% | ~20% | Low-frequency in Europeans; common in African ancestry |
| Jsb | >99.9% | ~80% | High-prevalence; anti-Jsb more often in African ancestry |
K is the most immunogenic antigen after D. The classic ranking from Section 5.3 is D >> K > c > E > k, Fya, Jka, S. Frequency is not immunogenicity. Most donors are K−, so a K− patient who receives one untyped K+ unit is at real risk of making anti-K. That is why anti-K is one of the most commonly identified immune alloantibodies after the Rh antibodies.
k, Kpb, and Jsb are high-prevalence antigens. Finding antigen-negative units for anti-k, anti-Kpb, or anti-Jsb is a rare-donor problem, not a fridge problem. Jsa is the ancestry trap: it is almost absent in European panels and present in about one in five African-ancestry donors, so an unexplained extra reaction on an African-ancestry panel cell is sometimes anti-Jsa.
Anti-K: IgG, always honored, HDFN is different
Anti-K is almost always IgG, reacts at 37 °C / AHG, and is always clinically significant. It causes hemolytic transfusion reactions (usually extravascular) and HDFN. Honor it for every red-cell issue. Do not wait for a 37 °C titer to prove it matters. Do not issue K+ units because the antibody is only 1+.
HDFN from anti-K is not a photocopy of anti-D. Kell antigens appear on early erythroid progenitors. Maternal anti-K can suppress fetal erythropoiesis, so the fetus becomes anemic with fewer circulating reticulocytes and fewer nucleated red cells than you expect from a hemolytic process of the same severity. Amniotic-fluid bilirubin (ΔOD450) underestimates the anemia. Middle-cerebral-artery peak systolic velocity (MCA-PSV) is the surveillance tool, not a bilirubin Liley curve. Titers correlate poorly with severity — a modest titer can still hide a hydropic fetus. If the mother has anti-K, type the father for K. If he is K− and paternity is certain, the fetus is K− and you can stand down. If he is K+ (especially K+k−, likely homozygous), pursue fetal KEL01 genotyping on cell-free DNA or amniocytes.
Anti-k, anti-Kpb, and anti-Jsb are the same class of problem — IgG, significant, rare-donor inventory — just aimed at a high-prevalence antigen. Anti-Kpa and anti-Jsa behave as ordinary IgG alloantibodies to low-frequency antigens: they can cause HTR and HDFN, and they hide on a panel until you land on the rare antigen-positive cell.
Ko versus McLeod — do not merge the nulls
Ko (Kell-null) red cells lack all Kell antigens. The KEL gene is silenced (nonsense, splice-site, or inactivating missense). XK protein is still present and Kx is actually increased. A Ko person who is immunized can make anti-Ku, which reacts with every cell that carries any Kell antigen — that is, virtually every donor on earth except other Ko cells. Anti-Ku is a rare-donor, IRL-level antibody. Ko cells are also the reagent you use to prove an unidentified high-prevalence antibody is in the Kell system: if the antibody is nonreactive with Ko cells and with DTT-treated cells, think Kell.
McLeod is not Ko. The XK gene on Xp21 encodes the Kx antigen. Males with an inactivating XK mutation have no Kx, markedly weakened Kell antigens (not absent), acanthocytes on the smear, and late-onset McLeod neuroacanthocytosis (areflexia, chorea, cardiomyopathy, psychiatric disease). Because XK sits next to CYBB (the gene for gp91phox, the catalytic subunit of the phagocyte NADPH oxidase), a contiguous gene deletion can produce McLeod phenotype plus X-linked chronic granulomatous disease (CGD) in the same boy. That pairing is a favorite BB stem. Females are usually mosaic carriers with a dual red-cell population.
Transfusion implications: McLeod patients can make anti-Kx and anti-Km. They should receive McLeod or Ko red cells if immunized, coordinated through a rare-donor program. Do not call a male with weakened Kell antigens Ko without looking at Kx, the smear, and the CGD or neurologic history.
Enzymes leave Kell; DTT and AET destroy it
Ficin and papain do not destroy Kell antigens. A panel that stays reactive after enzyme treatment is consistent with Kell, Rh, Kidd, Lewis, and several others. A panel that loses reactivity after enzymes is not Kell.
Dithiothreitol (DTT) and 2-aminoethylisothiouronium bromide (AET) cleave the disulfide bonds that hold the Kell glycoprotein together and destroy Kell antigens. They also destroy Lutheran, Dombrock, Yt, JMH, and Knops, and they destroy CD38 (daratumumab interference). They do not destroy Kidd, Rh, Duffy, or MNS. The exam contrast is almost always Kell versus Kidd:
- Antibody reacts with untreated and enzyme-treated cells, nonreactive with DTT-treated cells → Kell family (or another DTT-sensitive system).
- Antibody reacts with untreated, enzyme-treated, and DTT-treated cells → think Kidd (or Rh).
ZZAP (papain plus DTT) destroys Kell and is used in autoadsorptions; do not read a ZZAP-treated eluate as the patient has no Kell antibody.
Exam traps
- Treating anti-K as a nuisance antibody because K is only 9%. Immunogenicity, not frequency, drives the honor rule.
- Using ΔOD450 as the primary HDFN monitor for anti-K. Watch MCA-PSV; Kell suppresses erythropoiesis.
- Calling McLeod Ko. Ko lacks Kell and has increased Kx. McLeod lacks Kx, has weak Kell, acanthocytes, and possible CGD.
- Using enzymes to remove Kell. Enzymes do not destroy Kell. DTT/AET do.
- Issuing random K− units for anti-Ku. Anti-Ku needs Ko cells.
- Forgetting that Jsa is common in African ancestry and rare on European panel cells.
A boy with X-linked chronic granulomatous disease has acanthocytes, no detectable Kx, and weakened (not absent) Kell antigens. What is the genetic mechanism?
A previously transfused Ko (Kell-null) patient has an antibody that reacts with every screening cell, every panel cell, and every donor unit crossmatched. Which specificity is most likely, and what will be compatible?
An unidentified AHG antibody reacts 2+ with untreated and ficin-treated panel cells and is nonreactive with DTT-treated cells. Which specificity is most consistent with that pattern?