CME Slides Forum - J. Silver

A joint Congress by ERA-EDTA and ISN

BONE AS AN ENDOCRINE ORGAN: FGF 23 AND THE PARATHYROID

Justin Silver, Jerusalem, Israel

Chair: Tilman Drueke, Paris, France

Markus Ketteler, Coburg, Germany

silver

Prof Justin Silver
Department of Nephrology and Hypertension
Hadassah University Hospital
Ein-Karem, Jerusalem

Mr Chairman, first of all I would like to thank the organising committee for inviting us to present this talk. I would like to discuss with you bone as an endocrine organ, FGF23 in the parathyroid and I’ll start with my first slide showing you the Chagall windows which are large vitreous windows at Hadassah Hospital.

Well, mineral metabolism has its origins in very ancient history, in fact pre-history because prokaryotes, unicellular prokaryotes are able to synthesise vitamin D under the influence of ultraviolet light provided by the sun. So we’ve utilised the same systems to make vitamin D for ourselves. We all know vitamin D’s further metabolism and particularly it’s geared to increase calcium and phosphorous absorption so we can then mineralise our bones and have the normal physiological functions of calcium. But there are other factors involved in these networks of this mineral metabolism like vitamin D acting on the parathyroid to inhibit its synthesis, and more recently this recently described phosphatonin, fibroblast growth factor 23 which decreases PTH gene expression and serum PTH. We’ll discuss that in greater depth.

Well the parathyroids in rodents are very small, in the rat there are 2 parathyroids perched in the upper poles of the thyroid and in fact they are about 1 mm in size. So they present a challenge to understand them.

Well, they’re really endocrine glands and are classically very small. So zebrafish in fact use bone as an endocrine gland. They have PTH shown in this work from Geoff Zajac of Melbourne and the PTH is present in bone. We’re not sure if there is circulating PTH in fish but the parathyroid glands themselves, the presence of the parathyroids is important for the development of the gills from the pharyngeal pouches. So there is a relationship. But it brings us to bone as an endocrine organ.

Well, who would imagine that this bone of an elephant, of a dinosaur if you want, is really an endocrine organ? It is, it so turns out to be and a very important one involved in phosphorous metabolism.

How does bone do this? It does this by the use of osteocytes and osteoblasts as endocrine organs. Osteocytes present within the bone substance have long been known to be important for the recognition and response to mechanical stimuli and also micro fractures. How do they act as endocrine organs? Well, they’re actually -- cells and they communicate with the capillaries as well as the surface lining cells of bone. They respond to phosphorous and vitamin D and secrete FGF23.

So before we can carry on to the main topic subject of our talk, we should really remind ourselves of our bone cells. So pre-osteoblasts mature under the influence of DMP-1 that is a transcription factor to mature osteoblasts where PDMP-1 now helps the mineralisation of matrix. P DMP-1 is an osteocyte gene and protein and it’s upstream together with the phosphatonin MEPE of PHEX, an endopeptidase which we know is mutated in x-linked hypophosphatemic rickets. PHEX downregulates FGF23 of which we know its function on phosphorous and on vitamin D.

So I’ll show you another form. So we have DMP-1 and MEPE stimulating PHEX which itself acts to increase the activity of a protease which we don’t know as yet, it’s unidentified which breaks down FGF23, a well-defined 6 aminoacid sequence and FGF23 is responsible for the serum phosphorous.

So this is the classical knowledge that we have, the classical type of pathways which Jorge Cannata presented earlier with low calcium actually increasing PTH and so does high phosphorous and phosphaturic PTH and retains calcium from both acting on bone as well as on the kidney but in addition phosphorous acts stimulating the secretion of FGF23 which is a major regulator of phosphorous metabolism.

Well, things became much more interesting, -- was in Tokyo at the university of Tokyo described klotho when he mutated mice and found that they died early and the gene that was mutated coded for a protein which he called klotho. Well, why klotho? Well if you go to the Victoria and Albert museum in Cromwell road in London, you’ll see this beautiful Flemish tapestry and klotho is one of the 3 daughters of Zeus and they determine how long we live.

So we all live in trepidation of klotho. So when the mouse died early, he called them klotho mutations and the gene is called klotho gene. You saw earlier today that klotho is the co-receptor for FGF23 and determines where FGF23 is going to act. So where we have klotho FGF23 can act. So klotho is mainly present in the kidney, a bit in the brain and otherwise it’s really not present. FGF23 itself is increased in CKD similar to PTH and there’s a tremendous increase in FGF23 on a logarithmic scale as patients go into CKD more progressively.

The levels of FGF23 are markedly increased in dialysis patients. It was of interest that they correlate mainly with serum inorganic phosphate levels as shown by Imanishi and colleagues.

What was most rarely important was the work from Gutierrez, from Wolf’s laboratory where he showed that patients with the highest FGF23 levels starting dialysis, these patients here it doesn’t matter how you analyse for them, they have the highest mortality. So I think we need to understand what FGF23 is doing in our patients and in experimental systems.

Well if we had known this article earlier, we would have looked at it earlier but we didn’t. What we have here is the expression of the beta-GAL driven by the klotho promoter in transgenic mice shown by Takeshita and colleagues where they’ve shown the abdomen, the kidneys there’s a lot of klotho expression, in the sino-atrial node of the heart there’s klotho and it’s also present in this thing here apart from the parathyroid and the choroid plexus. This is in the neck and actually that’s the parathyroid. In the heart if you stress klotho knockout mice, they get nervous and they die of arrhythmia. So don’t stress your klotho mice.

So what about in the parathyroid? To cut a long story short we showed that the parathyroid is a target organ for FGF23 in rats and this has also been shown in mice. This work was done by Iddo Ben-Dov and Hillel Galitzer in our laboratory together with Tally Naveh-Many and I.

First we showed that klotho protein is expressed in the rat parathyroid by immunohistochemistry lots of it, as well as the co-receptor for FGF23 which is FGFR1. So they are both present in the parathyroid but not present in the thyroid. What about the mRNA? Is that present in the parathyroid? Well, it is present, it’s present in quite large amounts. When we look at parathyroid klotho mRNA there’s a lot of it. There’s also a lot of course in the kidney where we do remember the amount of klotho mRNA present in the kidney is only in the distal tubules while in the parathyroid it is in all the chief cells.

What about the function of this FGF23 on the parathyroid? So we administered either vehicle or FGF23 up to 5 days or sometimes for much shorter time periods and then measured what happened at the parathyroid.

So FGF23 given intraperitoneally. At 20 minutes this is serum PTH there was no change in serum PTH. But at 40 minutes there’s a decrease in serum PTH after FGF23. This correlated with changes in early growth response genes which we measured in the parathyroid and there’s also a decrease in PTH mRNA levels as shown by real time PCR.

If we go for longer time periods and measure the serum PTH or as I show you here the mRNA levels, they were markedly decreased by FGF23.

Well, we utilised methodology developed by the students in our laboratory because the parathyroid in the rat is very superficial, you can get to it if you take away the sternocleido mastoids. So some of the parathyroid glands were submerged in inhibitors of the ERK-MAP kinase pathway. We used the ERK1/2 inhibitor, UO126. At the same time we gave FGF23 systemically. We now added the inhibitor topically, this diffuses into the gland and allows us to look at an in vivo situation as FGF23 signal transduction intact in the presence of the inhibitor.

To look at this further, we use an in vitro system, there is no parathyroid cell line to the chagrin of all of us working in the field but we adopted the parathyroid organ culture system developed by Mariano Rodriguez in Cordoba. We take parathyroid organ in culture and we measure PTH in the medium. With control or after FGF23 added to the medium. In control, this is continual increase in PTH in the medium but in the presence of FGF23 there’s no marked increase and this is dramatically decreased compared to the control.

So what happens? Well, if we look at serum PTH in these rats, control, if we give FGF23 intravenously I think, at 20 minutes there’s a decrease in serum PTH but if we give it together with a MAP kinase inhibitor, the ERK1\/2 inhibitor, we can abolish the effect of FGF23. So this indicates, suggests that FGF23 is acting on the parathyroid through the MAP kinase pathway.

We want to look at what happens, is the same signal transduction involved? So we looked at the MAP kinase inhibitor using the in vitro system and if we looked at control first of all with no FGF23 and no inhibitor added or just the MAP kinase inhibitor, there’s no effect upon PTH in the medium.

If we add FGF23 once again there was a decrease in the PTH medium but when we add PTH in the presence of the MAP kinase inhibitor, we can abolish the effect of FGF23. So FGF23 is acting through the MAP kinase pathway also in vitro.

So this is the system that we present to you that phosphate regulates FGF23 and PTH regulates phosphorous and now FGF23 acts to inhibit PTH gene expression and serum PTH.

To look at this in summary what happens we suggest. In a situation of high phosphate the direct effect of phosphate on the parathyroid in all probability and lots of work from many people present in the hall, phosphorous itself acts to increase FGF23 secretion as does 1, 25 (OH)2D. FGF23 acts on the dimer of klotho and FGFR1c to inhibit PTH gene expression and serum PTH. Vitamin D does the same, it inhibits PTH gene expression as does a high serum calcium and of course, calcium is the major regulator of the parathyroid. Because if you knockout the calcium receptor, the mice are just not viable.
So calcium and the calcium receptor are the major regulators of parathyroid function but fine tuning we suggest is done by vitamin D and FGF23 through klotho.

This is just to show you where the work is done, the work is done in my laboratory in Hadassah hospital in the Hebrew university medical school and most importantly these are the people who have done this work.

This work was done by Iddo Ben-Dov and Hillel Galitzer, two talented graduate students in our laboratory and of course Tally Naveh-Many, she actually directs all the work done. Thank you very much.

Chairman: Thank you Justin for the presentation of this very elegant work. We have time for some questions or comments. May I ask the first question? I have just a first question myself Justin. My question is, is there a direct effect of PTH on FGF23 secretion or synthesis?

Dr. Silver: Well, it’s a very good question. In patients if you perform a parathyroidectomy, either patients with primary hyperparathyroidism or patients with tertiary hyperparathyroidism, the serum FGF23 levels go down not very much but they do go down, so it is suggestive that there may be some counter regulatory mechanisms. But of course, we’re also regulating serum phosphorous which is a major regulator of FGF23 and there has been no study done that I know of where there have actually been very good control also of serum phosphorous. But these are the studies that obviously people are dreaming and thinking about. Perhaps the ASN will learn more about this.

Chairman: Wait a minute we are running a little bit late so first one let’s see.

Dr. Dusso: I am very confused. The problem is that in the patient there was a report in PANS of a patient that had a constitutive klotho activation and this patient had very high PTH, very high parathyroid hyperplasia and that was independent of phosphorous levels. My confusion is like also with the paper in Science showing that klotho stimulates PTH secretion.

Dr. Silver: Well thank you very much. There’s a report in PANS by Brownstein and colleagues that a patient with a mutation, a translocation actually involving the klotho gene had high serum PTH levels and they called it an activating mutation but they didn’t prove it. There was a much more relevant paper from the laboratory of Michael Econs which was in JCI and they showed that a single patient with an inactivating mutation of klotho had very high serum PTH levels. So it depends which paper you choose, so I prefer the Econs paper.

Question: Justin, Mariano Rodriguez Cordoba. If FGF23 was invented to get rid of phosphate why would FGF23 decrease PTH which is a good companion to decrease phosphate in the body?

Dr. Silver: Well thank you. Obviously Mariano I don’t know the answer but there are always counter regulatory mechanisms. How does vitamin D or PTH act to increase serum calcium and then act to increase 1, 25 vitamin D which then acts on a self to decrease it? But if you look at the time sequence these things are probably more logical. Nature was there before us it seems that even -- was there before us.