Blood compatibility
A Socratic walk-through of blood compatibility — reasoned out one step at a time, not lectured.
The question we started with
THE QUESTION #Why can blood from one healthy person be lethal to another healthy person?
Two people are healthy. Their blood does the same job, carries the same haemoglobin, and looks identical under a microscope. Transfer a bag from one to the other and the recipient may be dead within the hour — not from infection or a toxin, but from their own immune system dismantling the cells that arrived. What can two healthy people differ in that makes one's blood a poison to the other, and why does the attack begin immediately, when immunity normally needs a first exposure to learn anything?
Reasoning it through
REASONING #Start with the surface. A red cell is a bag of haemoglobin with a coat of sugars and proteins on the outside, and the immune system judges cells entirely by that coat. The ABO system concerns one small piece of it: a short sugar chain, present on everyone, to which an enzyme may add one further sugar. Add one kind and the chain reads as A; add the other and it reads as B. Inherit the enzyme for each and you carry both. Inherit a non-functional version and nothing is added — that is group O, defined by an absence.
So far this is trivia. The lethal part is what floats in the plasma. You carry antibodies against precisely the versions you do not have: group A carries anti-B, group B carries anti-A, group O carries both, group AB neither. That is the whole rule, and everything downstream is bookkeeping.
But pause on how strange it is. Antibodies are meant to be the product of experience. A newborn has never met a foreign red cell, yet these appear within the first months of life anyway. The best-supported answer is that they are not really about blood at all: the gut fills with bacteria whose surface carbohydrates closely resemble the A and B structures, the immune system responds to those, and the resulting antibodies happen to cross-react with human red cells. You are not primed against other people's blood; you are primed against bacteria, and other people's blood is collateral.
That explains the speed. A first exposure normally takes days to answer; here the antibodies are already circulating in quantity, and they are of the IgM class, which is very good at clumping cells and at activating complement. Red cells rupture inside the vessels, the debris and free haemoglobin trigger clotting throughout the circulation and injure the kidneys, and blood pressure collapses. A modest volume of the wrong group can do this.
Now the bookkeeping. Only one question matters: does the recipient carry an antibody against anything on the donated cells? Group O cells carry neither A nor B, so nobody's antibodies find a target — hence the universal red cell donor. Group AB recipients carry no anti-A or anti-B, so nothing they own can attack what arrives — hence the universal recipient. Neither is a special property of the blood; both fall straight out of the rule.
Rhesus works on the opposite principle, and the contrast is instructive. The D antigen is a protein, not a sugar, and no bacterium resembles it — so an RhD-negative person has no natural anti-D. It must be learned, from a transfusion or a pregnancy, and there the consequences fall on someone else: if an RhD-negative woman carries an RhD-positive fetus, cells crossing into her circulation, mostly at delivery, can sensitise her. The antibody she makes is IgG, which unlike IgM crosses the placenta — so in a later RhD-positive pregnancy her antibodies enter the fetus and destroy its red cells. Haemolytic disease of the newborn is characteristically a second-child problem, and it is now rare where anti-D immunoglobulin is given to the mother, clearing the fetal cells before her immune system can learn from them.
The analogy
THE ANALOGY #Think of two crates of identical goods, distinguished only by a maker's mark stamped on the outside. A warehouse has standing orders to destroy anything bearing a mark that is not the house's own — and, unusually, those orders were written before any such crate ever arrived, because the clerks had seen that mark on signage all over town and assumed it meant trouble. The first foreign shipment is destroyed at the gate, with no learning period at all.
A maker's mark is chosen to signal something, whereas these are ordinary sugar chains with no known signalling purpose; their relevance is an accident of sitting where antibodies can reach them. And Rhesus reverses the arrangement — no standing orders, and destruction only becomes possible after a first shipment has been seen.
Clarifying the model
THE MODEL #Two simplifications above deserve correcting, both in the direction of "it is more complicated than the schoolbook rule".
First, "universal donor" is loose. It is true of red cells, which carry no A or B. It is false of plasma, which is where the antibodies live: group O plasma is full of anti-A and anti-B, so for plasma the universal donor is AB, exactly reversing the table. Whole blood and platelet products sit in between, and services screen O donors whose antibody levels are unusually high.
Second, ABO and RhD are two systems out of dozens. There are hundreds of recognised red cell antigens — Kell, Kidd, Duffy, MNS and many more — and while none attract naturally occurring antibodies of the ABO kind, a patient transfused repeatedly or previously pregnant can be immunised against any of them, and then react to blood that is a perfect ABO and RhD match. This is why laboratories screen a recipient's plasma for unexpected antibodies and cross-match the actual unit rather than relying on group labels. "O negative" is the right answer when someone is bleeding out and there is no time to type them; it is not a claim that the blood is inert.
A picture of it
THE PICTURE #How to readEach box is a blood group, with what its cells carry on the first line and what its plasma carries on the second. An arrow means the donor's red cells can be given to the recipient, and you can derive every arrow by checking one thing: nothing on the donor's cells appears in the recipient's plasma. O has arrows leaving to everyone and none arriving — that is what "universal red cell donor" means; AB has arrows arriving from everyone and none leaving. Read the second lines instead and the plasma rules run the other way entirely.
What became clearer
WHAT CLEARED #Blood groups are not different kinds of blood but different labels on the outside of otherwise identical cells — and the danger comes from antibodies you already carry, made against gut bacteria that happen to wear the same sugar. That is why the reaction needs no first exposure, and why the whole donor-recipient table is derivable from one sentence rather than memorised. Rhesus, which does require a first exposure, shows the ordinary immune timetable by contrast — and why its damage falls on a later pregnancy rather than on the person transfused.
Where to go next
ONWARD #- Why some maternal-fetal ABO mismatches cause mild jaundice while Rhesus mismatch can be devastating.
- How enzymatic conversion of A and B red cells into universal O cells is being attempted.
Key terms
TERMS #| Term | What it means |
|---|---|
| Antigen | a structure on a cell surface that an antibody can bind, here a sugar chain (ABO) or a protein (RhD). |
| Isohaemagglutinin | the naturally occurring anti-A or anti-B antibody present without prior transfusion. |
| Haemolysis | destruction of red cells, releasing free haemoglobin and injuring the kidneys. |
| Cross-match | testing a recipient's plasma directly against the specific unit to be given. |
| Haemolytic disease of the newborn | destruction of fetal red cells by maternal IgG crossing the placenta, classically anti-D. |
Every term the collection defines is gathered in the glossary.