Wednesday, March 05, 2008

Familial Hypercholesterolaemia and porcelain

What IS an LDL particle? There is a very neat picture here. Perhaps the most striking thing about it is that there is only one single protein molecule embedded in the particle surface, the apo-B100 protein molecule. It wraps neatly around the sphere of lipid to give support and identity. Identity is very important.

Of course the LDL particle has to come from somewhere. It is derived from IDL (Intermediate Density Lipoprotein) and before that from VLDL (Very Low Density Lipoprotein), which was secreted by the liver. All of these precursors also have the apo-B100 protein molecule, plus a whole load of other proteins too, apo-C1, apo-C2, apo-C3, and apo E. VLDLs are big (relatively), they need to be to fit all those proteins on to their surface! Most of these proteins are there to either turn on or turn of lipoprotein lipase, in order to get fatty acids from the lipid particle in to cells. That's how lipids get bulk transported. Apo-E gets the particle back in to the liver, should the liver want it back.

Quite where all of these proteins disappear to as LDL is formed is a bit of a mystery to me. Certainly it looks like LDL, derived from IDL/VLDL, has only apo-B100 and none of the other surface proteins of VLDL/IDL particles. Probably they hop on to HDL particles. Some of them came from there in the first place, rather than the liver directly, so that would be appropriate.

So there is this bag of lipid with just one identification tag on its surface, looking for a home. There is a receptor which exactly fits the apo-B100 id tag. The LDL cholesterol receptor. This is the archetypal, best studied endocytosis receptor of all time. LDL attaches to this receptor and the whole lipoprotein is engulfed by the cell sporting the receptor. The particle gets stripped of the engulfing apparatus, eventually fused with a lysosome and the cell is in clover with a supply of both lipid and cholesterol. It recycles the LDL receptor to the cell surface and does it all again. Cells appear to need both lipid and cholesterol.

But here's the interesting bit. The LDL receptor is a statin as in:

"Consistent with this hypothesis, they found that addition of LDL to the culture medium of normal human fibroblasts inhibits the activity of 3-hydroxy-3-methylglutaryl coenzyme A reductase (HMG-CoA reductase), the rate-limiting enzyme in the cholesterol biosynthetic pathway

EDIT: When you go to the original authors it becomes obvious that it is the cholesterol itself which inhibits HMG-CoA, a classic negative feed back mechanism. Have to do some more thinking about xanthomata. End edit.

The LDL receptor inhibits the intracellular synthesis of cholesterol, at the HMG-CoA step, when activated by an LDL particle. There is impeccable logic to this. Cholesterol is a complex molecule, synthesised in a stupendously intricate and energy demanding process. Usually the liver does this for cells of the body, shipping it out in lipoprotein particles. After all the liver is a biosynthetic powerhouse with lots of energy to play with. It's not some poor smooth muscle cell or fibroblast struggling to make metabolic ends meet in the arterial media. If, as a cell, you have just swallowed a big ball of cholesterol you can afford to switch off the production apparatus. It would be wasteful not to.

Cells need cholesterol. Homozygous familial hypercholesterolaemia means that a patient's cells have zero functional LDL receptors. If they can't get cholesterol by swallowing whole LDL particles they really do have to make it themselves. But the other feature of a non functional LDL receptor (remember, it's a statin) is that it doesn't give the ability to limit local cholesterol synthesis, so there might be quite a lot of cholesterol made. Enough to generate tendon xanthomas. These xanthomas should be composed of cholesterol synthesised from acetate in local cells, not from sticky LDL particles. I don't suppose anyone has checked this.

The synthesis of one molecule of cholesterol requires 14 NADPH molecules and 18 ATPs. It yields 6 phosphate (Pi) molecules and 4 pyrophosphate (PPi) molecules. While PPi appears to be an inhibitor of soft tissue calcification Pi is a promoter. Alkaline phosphatase, an enzyme whose production is upregulated by hyperglycaemia, converts PPi (calcification inhibitor) to Pi (calcification promoter). Hyperglycaemia is a feature of refined carbohydrate diets. Did I say low fat?

So there is the potential for an awful lot of inorganic phosphate ions, possibly 14 per molecule, being produced as a spin off from cholesterol synthesis on low fat diets. One LDL particle provides about 2000 molecules of cholesterol.

We know that insulin converts smooth muscle cells to bone secreting cells. Local cholesterol synthesis provides a supply of inorganic phosphate. A low fat (high sugar) diet raises insulin.

Perhaps it should come as no surprise that there is the phenomenon of "porcelain aorta", seen in patients with homozygous familial hyprecholesterolaemia. Aortic cells manufacturing cholesterol, which they can't get from LDL, produce inorganic phosphate and pyrophosphate as a spin off from the synthesis. A modern low fat diet, based on sugar, elevates blood glucose which increases alkaline phosphatase production and so the conversion of pyrophosphate to inorganic phosphate. A modern low fat diet elevates insulin too, which makes smooth muscle cells secrete bone matrix proteins.

Add calcium to phosphate in soft tissues and you have bone.

How good is the evidence base for current dietary recommendations in Familial Hypercholesterolaemia, particularly in children, who don't seem to get put on a pharmaceutical statin (yet)?

The Cochrane Review seems to think they are completely unsubstantiated.

Should all homozygous FH people be on a pharmaceutical statin? Another post there.


NB I didn't see any calcium or phosphate in that picture of the LDL particle. Soddy: "Rutherford, this is transmutation!" Rutherford snapped back, "For Christ's sake, Soddy, don't call it transmutation. They'll have our heads off as alchemists." Or cardiologists.


. said...

Thanks for the explanation. Nice picture you found too. I guess my mental image of little fried eggs and tiny strips of bacon floating around was a bit wide of the mark.

Currently wading into Good Calories, Bad Calories or rather The Diet Delusion as it's titled here (in NZ). It would have been a lot more shocking if I hadn't already learned all I have in the last month, but it's pretty shocking anyway.

. said...

Sorry to post more than one comment, but I'm all excited reading Good Calories, Bad Calories.

I was lead to believe by some reviewers that it would be impenetrably dense prose, but it's in fact very readable. Understated, true, but that suits the importance of the subject matter.

Anyway, I really want to learn how to critically read studies and research results, so I'm going to make a list of elements to watch out for.

Heres' my starting list:
- The groups should be equivalent or randomised (e.g. no use having all fat people in the control group and all thin people in the intervention group)
- Only one element should be different between the groups - the element being tested.
- Either there should be a placebo for the control group or they shouldn't know what the intervention group are doing (so they can't take make their own "improvements" and skew the result).
- The numbers in each group should be roughly the same.
- There should be adequate measurements at the start in order to have a "baseline".
- Some context should be given in the summary - e.g. if there's been some startling event during the study that may have resulted in behaviour modification (terrorist attack, hurricane, etc.)
- There should be an indication of how behaviour compliance was enforced (if at all) - e.g. self-reporting is probably not very accurate.
- The terminology should be precise - e.g. not "fats" but the kind and source.

Any other suggestions?

Stan Bleszynski said...

This is fascinating material, thanks for posting it. One thing that strikes me is that it proves the fact that serum lipoproteins are secondary markers and could not
possibly be contributing to arteriosclerosis, which was known long ago! Since FH homozygotous cases have no receptors for LDL then their tissue cholesterol and
cholesterol in their cardiovascular plaque must be produced by the cells themselves. Using glucose as the basic material! Since they lack the self-regulatory feedback from the faulty or non-existing receptors, thus their cells naturally tend to overproduce cholesterol. It seems so obvious. I hope it is accurate.

Last but not least, I noticed that in case of a high fat hich cholesterol diet (in non-FH)oversupply of intercellullar cholesterol is simply not possible because of the two feedback mechanisms: (1) the receptors on the cellullar membranes get saturated and clip the transport, and (2) freed receptors for LDL B100 migrate from the surface inwards and acts as a "statin" to suppress internal production of cholesterol by the cell. Simple and effective. It makes me feel better.

Stan (Heretic)

Peter said...

Hi Stan,

I think it's pretty unarguable that all tissues in homozygous FH patients are being asked to produce large amounts of cholesterol. If this was normal there would be non need for LDL cholesterol at all, clearly not the case. The LDL receptor was a red herring based on the book quote, and I suspect that the underlying problem with xanthomata is more complex. Obviously xanthoma formation is not simply related to LDL cholesterol level or they would be automatic in anyone with a raised cholesterol level of any origin, which they're not. I still can't see why people ignore the basic defect in FH, which is that peripheral cells cannot get bulk cholesterol. The build up in the circulation is clearly secondary. Procedures such as plasma apheresis and liver transplants also do far more than lower LDL cholesterol, much as the statins do. It's a real problem looking for answers to questions which should have been asked but so far don't seem to have been!


Peter said...

Hi chainey,

These things are quite difficult! One of my pharmacology tutors explained it this way. The problem is that most people read the abstract. If that really grabs them they MIGHT read the introduction and discussion/conclusion. If these agree with their own ideas that's it. It's only when they don't ring true that folks really settle down to read the results. These often contradict the conclusions in the discussion. Only when you get angry do you read the methods.

To be reasonable it is phenomenally difficult to only change one variable in diet trials. But when you add in relaxation, exercise, smoking cessation and group support you can't draw any conclusions re the diet aspect alone.

PS did you see the edit, tendon xanthomata are on hold, the rest of the post still looks OK to me


So FL Gal said...

peter, i have a question for you. I've upped my feine's fat and protein levels by adding shrimp and coconut oil to her diet. my question to you is, the coconut oil, is it safe for the cat to have everyday, or twice a week? it makes her coat like chinchilla and she bounces around like a kitten.

she ok with this regimen?

Peter said...

Hi so fl gal,

There is a link to a web site mentioned in the comments here

posted by Anna. There is so much to cat nutrition that it needs more than I can do here. I have to say that I've heard anecdotal reports of hepatic lipidosis in cats on coconut oil when it has been used as an attempted antiviral for FIV infection. I think that lots of reading around the rawfeeding sites is well worth doing and trying to keep as close to prey species is well worth it too.


. said...

so fl gal

If I may jump in here. I don't think Peter's link will work (at least not for copy/paste), so, to cut to the chase, the discussion he refers to is here
... and the recipe referred to is here: Food making at

Interestingly, I happened upon that very page myself not long ago. I need to get cracking and make up some chow for my Tiger.

Peter said...

Thanks chainey, that's what comes of using the office computer in coffee breaks at work!


Sue said...

So, if someone has familial hypercholesterolaemia would it be okay for them to eat a high fat, low carb diet?

Are you saying that in someone with F.H. the body cannot regulate the production of cholesterol so will merrily make too much. In the presence of a high sugar diet the production of cholesterol is increased. If people with F.H. ate a low carb diet the body will make less cholesterol?

Peter said...

For heterozygotes this is probably true. For Homozygotes, with two copies of the gene and no ability to uptake LDL through its normal receptor the situation is more difficult to understand. Bear in mind that there are many many many defects in the LDL receptor gene, all giving elevated LDL-C, present as heterozygotes in about 1:500 of any given population world wide. It is the commonest inherited "disease". It is only associated with heart disease in those heterozygoyes who have heart disease. In heterozygotes without heart disease it's not. The 1:1,00000 people with homozygous FH are in a slightly different situation. They have no choice but to make their cellular cholesterol in situ unless something like a receptor for Lp(a) steps in to replace the LDL receptor. Hard to say how well this might work. Currently there is no research being done other than lowering LDL-C, which is the usual end point reported/treated in the swathe of novel LDL-C receptor defects being reported worldwide.

How effective this is is best summed up here.

Note the "finally" and the "if" in the last sentence. No one knows if lipid lowering works for homozygotes. All of the lipid lowering techniques do more than lower LDL-C anyway.

"Multicenter analysis of such persons would finally determine if the early lipid-lowering procedures can significantly reduce the risk of premature cardiovascular disease in homozygous FH"


Anonymous said...

Chainey, those long links will work fine if you just triple click them, and copy. If you just highlight the text on the screen, they won't work sometimes. It's a pain. I with blog software would automatically change those URLs into hyperlinks.