CME Slides Forum - W.C. O'Neill

A joint Congress by ERA-EDTA and ISN

PHOSPHATE, PYROPHOSPHATE AND BISPHOSPHONATES IN UREMIC VASCULAR CALCIFICATION

W. Charles O'Neill, Atlanta, USA

Chair: Natale Gaspare De Santo, Naples, Italy

Carsten Wagner, Zurich, Switzerland

O'Neill

Prof. W. Charles O'Neill
Renal Division
Emory University
Atlanta, GA, USA

I’d also like to thank the organisers for the invitation to speak here. I’m not going to talk about phosphate transport I’m going to talk about the consequences of phosphate transport. I need to start by saying that my institution requires me to identify my complex of interest. I received support from Genzyme, Amgen and Baxter and I have a pattern related to material that I’ll be presenting here.

Now one of the reasons we have the transport of phosphate, maintained phosphate is to form bone basically composed of a mineral called hydroxyapatite and of course as you all know when this system goes rye we can end up with hydroxyapatite in places where we don’t want it and particularly in the blood vessels. I’m going to start out by mentioning that there are two different forms of vascular calcification, there is a new intimal form that occurs in atherosclerosis I’m not going to talk about that form of vascular calcification, that occurs in anyone who develops atherosclerosis whether or not you have renal failure.

I’m going to focus on medial calcification which occurs in the medial layer. It’s totally independent of atherosclerosis. The two can co-exist but the pathophysiologies are quite different. This is the form that is exacerbated in renal failure also seen with ageing and in diabetes in some rare genetic disorders that I’ll touch on.

Now, to understand this we need to go back to some inorganic chemistry so bear with me. This reaction which is commonly thought to occur but does not occur certainly does not occur in an aqueous environment, it doesn’t occur in an organism, it doesn’t occur in a test tube. The reason for this is that hydroxyapatite has a very complex structure shown here. Here’s the chemical formula.

So the chances of this kind of reaction, of this being formed spontaneously by interactions of calcium and phosphate are very unlikely. So this reaction doesn’t occur. Instead the formation of hydroxyapatite is at least a two-step process and probably more where initially we form calcium phosphate or brushite and brushite can spontaneously dehydrate into hydroxyapatite. Now what’s interesting about this is that if we look at the solubility products here the hydroxyapatite under physiologic pH is virtually insoluble you can see the apparent solubility product is 10 to the minus 50 I mean it’s very, very, very low.

So once you have hydroxyapatite it’s there for a long time. However the first step in this process, the formation of brushite has a solubility product that is 2-3 fold higher than the normal calcium and phosphate product in the plasma. Now this is important when you consider this because of two reasons, first this explains why our bones don’t dissolve because once we’ve formed hydroxyapatite it’s relatively insoluble but why don’t we all just turn to stone? But of course you also might ask the question well how can we ever form any bone if we can’t even form this compound at concentrations that are normally present in bodily fluids? The answer to that question really boils down to the microenvironment.

So that we create microenvironments in the body that then lower the solubility product to below the normal calcium phosphate product present in body fluids. To take an example in bone and again this is grossly over simplified but the collagen 1 in bone along with other proteins will bind calcium in such a way that in a non-ionic mechanism.

so that the positive charge of the calcium is still available and then this will then attract phosphate.

So on this scaffolding here this then will greatly increase the likelihood of formation of brushite and then hydroxyapatite and then we will form the hydroxyapatite.

Now the same thing can happen in the vascular wall but the protein primarily involved in this is elastin and the initial calcification site are the elastin fibres between the smooth muscle cells.

So again elastin has been well known for many years to be able to bind calcium in a non-ionic mechanism that again leads to positive charges available for the phosphate.

However as you all know our vessels don’t all turn to pipes and the reason for that is that smooth muscle cells elicit a number of inhibitors that prevent all hydroxyapatite formation.

These include pyrophosphate which I’ll discuss in a minute, matrix Gla-protein, osteopontin, there’s likely to be other inhibitors as well. I should also mention there’s magnesium that is also an inhibitor of hydroxyapatite formation.

Now to study this process in more detail we developed a model of cultured aortas so we actually take the whole aorta from rats and put it in a culture medium. It’s actually an amazingly simple procedure. These vessels will remain alive and apparently normal responsive to hormones and normal histology for weeks we’ve gone out as long as 3 weeks in culture. When you do this and you put them in a very high medium with very high calcium and phosphate levels, these vessels do not calcify as shown here. We’ve gone out as long as three weeks and seen no calcification.

However, we can induce calcification in these vessels if we injure them by rubbing a traumatic type of injury or if we add alkaline phosphatase to the culture medium and this shows here the here is the calcification in the medial layer, you can see these vessels even after nine days in culture we have a normal single layer of endothelium here. These are the non-calcified vessels.

Now if we think about what role phosphate has here of course we’re looking at a simple chemical reaction that involves phosphate so phosphate can enhance calcification clearly on a mass action basis and in our culture systems not surprisingly as we increase the phosphate concentration we get an increase in calcification. We get no calcification at a normal physiologic phosphate concentration.

So phosphate certainly can promote calcification simply by increasing the rate of this reaction here.
That may not be the whole story with phosphate because the work of Doctor Giachelli has pointed to a potential role of intracellular phosphate in this process whereby phosphate enters the cell through phosphate transporters as we’ve heard about and that the elevated phosphate concentration inside the cells will turn on a number of genes related to calcification. The exact mechanism of how that’s happening is not entirely clear.

The evidence for that comes from a series of experiments that Doctor Giachelli and her colleagues performed in cultured smooth muscle cells and again here we’ve seen this compound phosphonoformic acid and they showed that when they added this compound, they could reduce the expression of various calcification-related genes and reduce calcification in the smooth muscle cell cultures. However, as we’ve heard here today PFA is really not an inhibitor of the PiT-1 transporter that’s being proposed in these studies. So the mechanism here is unclear and the mechanism of inhibition of calcification may be due to the fact that phosphonoformic acid is also a bisphosphonate compound, it’s a non-hydrolysable analog of pyrophosphate.

So these investigators went back and repeated the experiment using silencing RNA of 2PiT-1 and were able to show or get similar results of showing changes in these calcification related downregulation of these calcium-related genes. Indicating that phosphate transport into the cell is involved in inducing a calcification- related or osteoblastic phenotypic transformation in the cells. We’ve tried to do this with the intact cultured aorta system and we have not been able to identify any effect of phosphate related to a gene transcription. So whether this is just a phenomenon that occurs in cultured cells and not in intact smooth muscle is not clear at this stage.

I want to show you the structure of pyrophosphate, it’s basically two phosphates linked together. This is a compound that’s synthesised primarily as a by-product within cells of a variety of enzymes related primarily to nucleotide and glycogen metabolism. There does not appear to be an enzyme within cells that’s specifically designed to make pyrophosphate. I misspelt this, it is pyrophosphate and not pyrophosphonate sorry about that. Pyrophosphate has been known for decades to be a very potent inhibitor of hydroxyapatite formation in vitro. In fact if you maintain a concentration of 2 µM pyrophosphate in a test tube when you see hydroxyapatite crystals in there as long as you maintain the pyrophosphate concentration at 2 µM you will get no hydroxyapatite formation.

Now the normal pyrophosphate level in human plasma is 3-4 µM, so we normally have pyrophosphate concentrations in plasma that would completely prevent hydroxyapatite formation.

The mechanism by which pyrophosphate does this is not entirely clear but it appears that it binds to nascent crystals, very small crystals prevent this step here but it may also prevent the initial formation of brushite.

The evidence that pyrophosphate is an important inhibitor of vascular calcification comes initially from our studies in cultured aortas where if we add pyrophosphate to the injured aortas we can completely inhibit calcification. If we add pyrophosphate to the cultures not alkaline phosphatase but pyrophosphatase we can induce vascular calcification. Inorganic pyrophosphates are highly specific enzymes opposed to alkaline phosphatase and only hydrolyze pyrophosphate.

Now where do we get extracellular pyrophosphate from? Because this process of vascular calcification is an extracellular process. So the pyrophosphate has got to come from somewhere. As I mentioned pyrophosphate is the product of metabolism in the cells so it can exit the cells this putative transporter ANK is responsible for pyrophosphate leaving cells. An additional mechanism for extracellular pyrophosphate is hydrolysis of ATP that is released from the cells by this ectoenzyme PC1 or ectonucleotype pyrophosphorylase otherwise known as ENPP1 producing extracellular pyrophosphate. We actually have some genetic data to support the role of these compounds. If ANK is knocked out in mice they develop ectopic calcification but not calcification in the vessels, they develop it primarily in the joints suggesting that ANK is perhaps not important in vascular smooth muscle.We have the human knockout of PC1 or ENPP1 and results in a disease known as infantile arterial calcification. This is a very striking disease where kids develop severe vascular calcification, die of congestive heart failure usually by the age of 3 or 4 if they’re not treated so indicating that this enzyme is probably quite important in vascular calcification again proving evidence that pyrophosphate is an important endogenous inhibitor of calcification.

An additional very important enzyme in this process is another ectoenzyme alkaline phosphatase and in this case tissue-non specific alkaline phosphatase present on every cell. This is the same alkaline phosphatase present in bone. This enzyme hydrolyses pyrophosphate to inorganic phosphate. The evidence that this enzyme is important in pyrophosphate metabolism is another human genetic disease hypophosphatasia where individuals have reduced levels or absent TNAP

and those patients have elevated levels of pyrophosphate in their plasma. I’m going to get back to alkaline phosphatase because it plays a key role in vascular calcification. Now you might ask well if all this is happening, if pyrophosphate is going around and is at concentrations that prevent hydroxyapatite formation how do we ever make bone? Well again in the parallel comparison of bone in artery wall the reason that we don’t get calcification in the blood vessels and we do get calcification in bone is due to the high level of alkaline phosphatase from bone, bone has one of the highest levels of any tissue in the body of TNAP. So the function of TNAP in bone appears to be to get rid of pyrophosphate allowing bone to form. Other tissues that have low levels of alkaline phosphatase then pyrophosphate levels are high enough to prevent calcification.

So what happens in renal failure? Well we did some renal studies looking at dialysis patients and found that dialysis patients had lower levels of plasma pyrophosphate than control patients without renal disease. This is primarily due to a subpopulation of patients at very low levels. You can see the upper range was the same in both groups. We have subsequently teamed up with Chris McIntyre in England who had a nice cohort of patients where he was looking at vascular calcification over time and both the baseline calcification score and the change in calcification score as you can see here decreased with increasing quartiles of pyrophosphate suggesting, these are in patients with either stage IV CKD, hemodialysis or peritoneal dialysis patients suggesting that pyrophosphate levels may be important. But I should caution that the levels of pyrophosphate in the plasma may not really reflect the levels of pyrophosphate right in the vascular wall where the calcification is occurring.

So we went on to look to see what the mechanism for this might be. We looked at aortas from two different models of renal failure the adenine model and 5/6 nephrectomy model and both cases the alkaline phosphatase activity in the vessels was double than the control animals and we looked at pyrophosphate hydrolysis it was of course -- increased. If we looked at the amount of protein in Western blot you can see here that there’s more alkaline phosphatase protein in uremic aortas than in control aortas.

The question is, is this enough to explain calcification? We went back to our cultured aorta model. We were able to transfect alkaline phosphatase gene into these vessels using an adenoviral construct and when we did that this is 40 hours and 90 hours after transfection and the yellow bars are vehicle, the red bars are with levamisole inhibitor of alkaline phosphatase and you can see here we get an upregulation of alkaline phosphatase activity and if we look at calcification in these cultured aortas you can see that the vessels that were transfected with alkaline phosphatase showed extensive calcification.

So upregulation of alkaline phosphatase at least in vitro in a cultured system is sufficient to produce calcification. So our working hypothesis is that uraemia there’s an upregulation of alkaline phosphatase in the blood vessels which then reduces the level of pyrophosphate and then -- the inhibition of hydroxyapatite formation.

Now to try to get around that we thought well maybe if we treat the animals with exogenous pyrophosphate maybe we can overcome this and inhibit calcification.

I don’t have time to go through these experiments in detail but when we did initial experiments we were able to get significant inhibition of calcification we have subsequently just recently overfined this procedure. If we give enough pyrophosphate, we can almost completely prevent calcification with pyrophosphate.

I want to move on in the interest of time to these compounds which are bisphosphonates which you’ll recognise the similarities in their structure to hydroxyapatite. We now have a carbon here instead of an oxygen and very different side groups here that assign additional properties to the bisphosphonates.

These compounds were originally designed and developed to be non-hydrolysable analogs of pyrophosphate in order to treat ectopic calcification that was our initial purpose, it was subsequently found that they had effects on bone reabsorption that are now responsible for their principal clinical utility.

We reasoned that if in uraemia the problem was an upregulation of alkaline phosphatase which was hydrolysing the pyrophosphate we could bypass this by giving a pyrophosphate analog that can’t be hydrolysed. First we tested this in the cultured aorta system and you can see here that etidronate at very low concentrations will completely inhibit a calcification.

So this is the idea here of how bisphosphonates might be useful in treating vascular calcification. We went onto the in vivo model this is in the adenine treated rats and you can see that either pamidronate or etidronate, pamidronate was more potent would completely prevent vascular calcification but required quite high doses much higher than the doses normally used in humans to suppress bone reabsorption.

Now, one of the concerns, of course, and I get back again to this parallel between the bone and the artery the concern of course is that if I’m using bisphosphonates to bypass the hydrolysis of pyrophosphate here we might have effects on bone as well because again there’s a lot of alkaline phosphatase in bone that’s preventing pyrophosphate from inhibiting hydroxyapatite formation.

So we looked at the bone in these animals in collaboration with Doctor Maluca in Kentucky and looked at bone formation and you can see that basically the inhibition of vascular calcification really paralleled inhibition of bone formation. So at doses of bisphosphonates either etidronate or pamidronate that were required to inhibit vascular calcification also inhibited bone formation. We looked at bone reabsorption as well and didn’t see any big differences there and I can’t show you all the data in the interest of time. This has all been recently published.

So to summarise the role of these compounds in vascular calcification, phosphate by being a component of hydroxyapatite can certainly when elevated drive this reaction. Certainly in an in vitro system and in in vivo systems we don’t see medial vascular calcification in the uremic model or in the cultured model unless the phosphate is elevated, unless they’re on a high phosphate diet. So clearly whether there’s an additional mechanism that requires effects of phosphate on the cell is unclear it has been demonstrated in cultured smooth muscle cells but we have not been able to demonstrate that in intact vessels.
In uraemia there appears to be an upregulation of alkaline phosphatase in the smooth muscle which then diverts pyrophosphate away form its inhibitory role inhibiting calcification resulting in vascular calcification. Now I don’t want to imply that this is all that’s going on, vascular calcification is clearly a very complex process there are other players here, other inhibitors, other processes that may promote calcification.

So I want to close being that we’re in Italy and I had the privilege of seeing this painting when we were in Florence a couple of days ago, the body has intricate systems that really prevent us going from this to that. I also had the privilege of seeing this in Florence and I just want to close by acknowledging all the people at Emory who have helped with this work and my collaborators and institutions and of course none of this would have been possible without the generous financial support of these entities here. Thank you.

Chairman: Thank you very much. We have time for a few questions. Maybe I can start. Extracellular pH has an important role in bone formation and absorption, is there any parallel role in the smooth muscle cells in the aorta?

Prof. O'Neill: Yes obviously I couldn’t talk too much about bone formation so I didn’t mention pH. pH plays an important role, in bone it primarily plays a role in absorption by decreasing pH. Hydroxyapatite solubility is extremely pH dependent. So small changes in pH can make a big difference. We actually published this several years ago in the cultured aorta system simply bumping the pH from 7.4 to 7.5 produced a sizable increase in calcification. There’s also been a paper published showing in vivo that the induction of the metabolic acidosis in vivo can inhibit vascular calcification. So pH is an important consideration especially when you think of what we’re doing to our patients every time we dialyse them where we’re increasing the pH from about 7.2 up to about 7.5 so we’re really increasing the pH by 0.3 units for a period of time and then of course the pH will stay elevated for a number of hours after dialysis. We actually went back and did it in cultured aortas where we actually stepped the pH up from 7.4 to 7.5 for about 5 hours every day basically mimicking hemodialysis on a daily basis and we were able to see significant increases in calcification. So I’m not aware that in the vascular smooth muscle there’s a mechanism by which the local pH can be changed, so I don’t know if that exists or not.

Question: Thanks for a wonderful lecture. I wanted to ask you how you felt then the relative responsibilities are if you like for phosphate concentration in plasma versus the inhibitory mechanisms because it seems to me that the changes in plasma phosphate that we see in the associated studies aren’t very great really and it’s rather surprising that we should have such a dramatic effect and I wondered what you thought.

Prof. O'Neill: Well that’s interesting if you look at the correlative studies, there are very small changes you know the effect of phosphate in those studies on calcification is very small. Let me mention in terms of inhibitors the point of how potent pyrophosphate is it inhibits calcification 3 orders of magnitude lower concentration than ambient phosphate concentrations. So it’s extremely potent, it’s not competing with phosphate because it’s present at a thousand fold less concentration. If you look at studies for instance I think what the studies are illustrative of this are the studies that have used non-calcium based phosphate binders, sevelamer showing changes in calcification in patients. If you look at the data you really know significant change, no real significant difference in the phosphate levels between the calcium based phosphate arm and the non-calcium based phosphate arm. If you look carefully you’ll see a very small change in calcium. Obviously I don’t want to talk about calcium here because we’re talking about phosphate but we see the same thing in animals and in the cultured vessels that very small changes in calcium can have a major impact on calcification. So I think that we focus on phosphate but I think that there’s an important story with calcium there as well. So clearly I think phosphate is a funny thing because we know we need to have the elevated phosphate levels to get the calcification but once the calcium levels are elevated do small changes in the calcium make a big difference? It’s unclear and I think one has to be really careful here and I think this gets back to the first lecture that we tend to look at fasting levels. We rarely look at postprandial levels and there has been some recently published data that suggests that we perhaps need to be looking at the postprandial levels that they may have more to do with what’s going on than the fasting levels. I don’t know if that answers your question.

Question: Again most of the work you talked about was in normal aortas that have been injured or whatever. Do you feel that after a sclerosis actually has a different mechanistic effect? I’m thinking primarily of the translational model that Keith Ruska has with the LDL receptor knockout where phosphate absorption, it doesn’t matter whether you use calcium based or non-calcium based binders tends to really ameliorate that calcification that’s done there, your comments.

Prof. O'Neill: Yes there’s a lot about -- I didn’t mean to imply that these mechanisms didn’t have any role in atherosclerotic calcification I just wanted to point out that it’s a totally different process that’s associated with inflammation. There are phenotypic changes in the smooth muscle cell which may be really why the cultured smooth muscle cell model may be a better model for atherosclerosis than for medial calcification. So the cell type is different. There are a lot of inflammatory cells around and the big difference of course is that we get calcification in atherosclerosis in anybody with severe atherosclerosis develops calcification. Whereas the medial we only see that in specific situations, renal failure, ageing, diabetes and genetic disorders. Now could pyrophosphate be an inhibitor there as well? It certainly could be. So all these inhibitory mechanisms could be the same. The role of phosphate could be the same. We know that’s hydroxyapatite that’s being deposited there. So again the phosphate could play a role there as well. I’m not meaning to say that phosphate doesn’t play a role in medial calcification it’s just that when you look in people and in animals we can’t see big differences in the phosphate concentration in the plasma despite large differences in calcification. So there is certainly something else going on.

Question: -- from Spain. I have a question. Do you know something about phitic acid or phitate and the role inhibitor of arterial calcification or renal stones?

Question: ‘Phitates’ they are in the diet, in beans or fibre. Do you know something about it?

Prof. O'Neill: I’m not understanding. Phitates? Oh phitates? Ok I’m sorry. Well, there are a number of polyphosphates that can inhibit calcification. This was shown way back in the work by Flaish and colleagues way back in the early 60s. They used a number of polyphosphates including pyrophosphate but longer ones and number one, most of them will inhibit vascular calcification. Phitates certainly could too as well the question is how much are they getting absorbed? I don’t know very much about phitates but I think there have been in vitro studies showing that phitates will inhibit hydroxyapatite formation. I’m not absolutely certain about that but a whole host of polyphosphates will do that. I don’t know about the role in vivo because I don’t know if they get absorbed.

Prof. O'Neill: They’re clearly in a diet and they clearly bind calcium. It’s well known. Do they get absorbed I don’t know that they get absorbed?

Question: Yes we get all phitates by diet and we think that the diet we put our patients on with low phosphate has low phitic acid we think there is a role.

Chairman: Thank you very much. Ok I think we have to close the session I’d like to thank the speakers and I’d like to thank the audience for their participation.