PFIC 3 “MDR3 deficiency”, or “ABCB4 disease

By Alex Knisley, MD
Consultant Histopathologist
Institute of Liver Studies
King's College Hospital

Some persons will have come to this website to learn more about a disorder called “PFIC-3.” A more precise, and less confusing, name for the disorder is “MDR3 deficiency”, or “ABCB4 disease.” Just different spoonsful of alphabet soup, I can hear people saying – but let me try to set out why I think the distinctions matter.

The shorthand name for one of the clinical-laboratory studies, or “blood tests”, done to monitor children with liver disease is “GGT”. GGT stands for “gamma-glutamyl transpeptidase.” This is an enzyme, a protein that speeds up a chemical reaction. Samples of blood serum (what is left of blood after clotting has occurred and the clot is separated out) contain some GGT. The concentrations of GGT activity in serum can be measured in different ways; the values obtained are used to track the severity of cholestasis. In all but a few forms of cholestasis, GGT values rise in parallel with bilirubin values. (Bilirubin is the yellow substance that gives jaundice its color.)

“PFIC” was originally defined as the class of rare inherited childhood liver disorders in which GGT values did not rise in parallel with bilirubin values. “PFIC-1” was the first member of that class to be described, the original “Byler disease”, and the first member for which a responsible gene was identified. “PFIC-2” was the second member to be distinguished, and the second member for which a responsible gene was identified. The search continues for others.

In “PFIC-3”, however, GGT values are not low. They are high.

So confusion has been great. Doctors and nurses are puzzled – can there be a “high-GGT” form of PFIC, when the definition of PFIC requires that GGT values be low? (In my opinion: No; logically there can’t.) Many other inherited childhood liver disorders, as well as even more acquired childhood liver disorders, are characterized by high GGT values. More puzzlement – my patient or my family member, who has a small child, has “high-GGT” liver disease; does (s)he have “PFIC-3”? (Not necessarily.)

Those puzzled questions, I think, are the result of a misleading name, one that has got even medical professionals muddled. I’d rather that the terms used to describe any disorder be clear and specific, and that’s why I prefer the designations “MDR3 deficiency” and “ABCB4 disease” to “PFIC-3.”

What, then, are “MDR3” and “ABCB4”?

MDR3 is an officially approved abbreviation for an enzyme called “multidrug resistance protein 3.” (The third identified member of a set, that is, of proteins that in structure resemble “multidrug resistance protein 1”, which acts to push various substances out of liver cells.) Yes, there are committees of scientists that track, evaluate, and assign names to newly discovered substances, to ensure that everyone sings from the same page of the hymnbook. The officially approved abbreviation for the gene encoding MDR3 (providing the cell with instructions on how to make MDR3) is ABCB4, short for “ATP-binding cassette gene class B, member 4.” Defects in ABCB4 – gene designations are put in italics, if you were wondering – lead to complete or partial lack of working MDR3. This lack then causes disease of the liver, including the bile ducts.

How does that happen? To explain that, I have to describe what MDR3 does in the healthy liver.

MDR3 contributes to forming bile. Bile is a mix of water, and salts, and bilirubin pigment, and wastes or toxins that the liver has partly broken down, and fatty substances called lipids, and detergents called bile acids. Without MDR3, the bile is deficient in one of those fatty substances, a lipid called phosphatidylcholine. The tiniest branches of the biliary tract (the bile canaliculi) have walls made of parts of liver cells (hepatocytes). Those walls, or membranes, are two layers thick. MDR3 in the hepatocyte spans the inner and outer layer of the canalicular membrane and moves phosphatidylcholine from the inner layer to the outer layer.

A quick digression about “detergents.” That oil and water don’t mix is proverbial. Detergents, however, let such substances mingle with one another. Laundry detergent breaks up grease and lets it disperse in water. Bile acids inside the small intestine break up the fat in what we’ve eaten and let it disperse in the mix (mostly water) of partly digested food. Once the fat is dispersed, the cells that line the intestinal wall can take it up. Without bile acids, we can’t absorb fat, or fat-soluble substances like some vitamins, from our diet. This is why cholestatic children require vitamin supplements: Bile acids in such children aren’t making the journey from hepatocytes to small intestine.

Let’s head back to the membranes that line the biliary tract. If MDR3 is present in the canalicular wall, and if the MDR3 present is working normally, the outer layer of the canalicular membrane will contain phosphatidylcholine. Under the influence of bile acids, then, phosphatidylcholine can float out of the membrane and into the bile, like a grease spot out of a work shirt and into the wash water. In the bile the phosphatidylcholine forms a complex with the bile acids and acts as a “chaperone.”

The action of phosphatidylcholine as a chaperone – sticking close to the bile acids and keeping them from causing trouble, like teachers watching rowdy teenagers on Prom Night – is essential because bile acids are very corrosive substances. The canalicular membranes, and the membranes of the cells that line the bile ducts, contain large quantities of lipids. If bile acids in the bile are not chaperoned by phosphatidylcholine, they will attack the membranes and injure them by leaching lipids away. This sort of injury can lead to cell death or dysfunction. Cell death and dysfunction in the hepatobiliary system lead to jaundice, and itching, and inflammation, and scarring, which eventually may be fatal without a liver transplant.

And that is the sequence proceeding from mutation in ABCB4 (or ABCB4 disease) to absence or malfunction of MDR3 (or MDR3 deficiency) to unhappy, undergrown, scratching, yellowish, miserable children on the transplant waiting list.

Why are GGT values high in MDR3 deficiency? Because GGT is present in large quantities in the canalicular and bile-duct membranes (mostly canalicular). When those are damaged, GGT is released into the bile. GGT then leaks across the wall of the biliary tract into the blood and the concentrations of GGT activity in the serum rise.

Why are GGT values low in PFIC-1? Because GGT is lacking in the canalicular membranes in that disorder – if it isn’t there, it can’t be released.

Why are GGT values low in PFIC-2? Because bile acids are not normally pumped into bile in that disorder – although GGT is present in the canalicular membranes, without the detergent action of bile acids the GGT can’t be released.

MDR3 deficiency has a broad clinical spectrum. In total deficiency, the effects are quite severe, often are evident in early childhood, and can require liver transplantation for relief. In partial deficiency, the effects are milder. They may include gallstones, or off-and-on itching, and may show up only in adulthood or even middle age. When MDR3 deficiency is suspected in adults, two principal approaches to diagnosis exist. When it is suspected in young children, a third approach also can be useful.

Most definitive, of course, is analysis of ABCb4, looking for gene changes that can be predicted to disrupt gene function. This is expensive and takes a long time. In some cases, the abnormalities found are not clear-cut, and additional studies must be done. It is most useful for families who want to take advantage of prenatal diagnosis.

A reasonable tack to take is to analyze bile itself to learn if phosphatidylcholine is deficient. Such deficiency can be inferred, if other things fit the picture, to be the result of lack of MDR3, or of subnormal MDR3 function. (Sampling bile, however, is not an easy business, particularly in the very young.) Bile analysis can permit the diagnosis of functional deficiency of MDR3.

Genetic analysis and bile analysis can be used in diagnosis of both severe and mild forms of MDR3 deficiency. For severe forms, immunohistochemical studies of liver biopsy materials also can be conducted. Antibodies that tag MDR3 protein are used in this approach. If no MDR3 protein can be demonstrated in liver tissue, as shown in Figure 1, actual (rather than functional) MDR3 deficiency can be diagnosed. If the antibodies react, and if the pattern of reaction is appropriate, then MDR3 is present, as shown in Figure 2. This does not give any information, however, on how well the MDR3 actually functions. Persons with disease manifest after infancy likely have some MDR3 protein – protein that works, but that that does not work very well. Immunohistochemical studies are not conclusive in such cases. Persons with disease manifest in infancy, that is, with severe clinical signs and symptoms, are more likely to lack all MDR3 protein, and are better candidates for immunohistochemical study of liver biopsy materials.

So far, to my knowledge, no one has been found who has functional severe MDR3 deficiency, as demonstrated by analyses of bile that show phosphatidylcholine to be almost totally absent – and who has documented ABCB4 mutations – and who has no lack, on immunohistochemical study, of MDR3 protein along canaliculi in a liver biopsy specimen. But such a person will almost certainly be found, because some mutations in other genes like ABCB4 lead cells to make forms of proteins that are normally handled by the cell, going to the right places and reacting properly with antibodies, but do not perform their usual function at all. In my opinion, it is just a matter of time till a person with this sort of ABCB4 disease shows up. When immunohistochemical studies demonstrate no MDR3 protein, the test has given a clear answer: “Severe MDR3 deficiency is present.” When they demonstrate MDR3 protein, the answer can only be: “This study does not support the idea that severe MDR3 deficiency is causing this patient’s disease. However, it also does not disprove that idea.”