Slide 1

First of all, I would like to thank the organising committee for inviting me to talk about bone to kidney people. So my goal will be to try to show you how bone is a fascinating tissue and I hope you will share my enthusiasm at the end of my talk.

Slide 2

Bone is a composite tissue with various functions.  The first function of course, is as you know calcium and phosphate metabolism  it’s also related to locomotion and this is related to bone mass and bone quality and when it’s altered, patients have fractures .  Bone is finally a shelter for hematopoiesis.
The inputs: Bone is submitted to calcium intake and nutrition, mechanical stress, aging, infections, inflammation and tumors and this list is not finished. There are lots of controllers including of course, the hormones of the phosphorous and calcium metabolism as well as the immune system and the hormones that control energy intake and the adipokines.
In the centre is a vessel. Blood vessels are very important for bone remodelling control. Blood vessels are also very important for the differentiation of the haematopoietic stem cells because they control the partial pressure of oxygen and we do know now that the partial pressure of oxygen controls the differentiation of haematopoietic stem cells.
We also know now and since the very nice work by Chantal Chenu that nerves are in a dense network around the bone cells, the osteoblasts as well as the osteoclasts. So you see that bone is in the middle of a network of controls.

Slide 3

So let’s go to the theatre and let’s see what the play of bone remodelling is about. Let’s see for the actors.

Slide 4

This is an image of a very intense secondary hyperparathyroidism and here you can see that there are lots of osteoclasts eating and resorbing bone and here are the osteoblasts that will synthesise the bone matrix just at the level where bone was resorbed by the osteoclasts. The little red dot here is the osteocyte. Osteocytes have been historically ignored cells for many, many years for several reasons. First of all it’s a very difficult cell to study because it’s embedded in the bone matrix and it was rather impossible to study it in vitro. Now we have some cell lines that look like osteocytes and it has become a more and more known cell and you will see that it’s becoming probably the most important cell in bone remodelling control.

Slide 5

The activity of the cells that resorb bone, the osteoclasts and the ones that make the bone, the osteoblasts and the balance between these activities will lead to the bone mass and we do know that bone mass is responsible for the mechanical resistance to fracture.
One thing you have to keep in mind is that the amount of bone that will be resorbed depends on two things; the first one is the number of osteoclasts that are working meaning that the recruitment, the proliferation of the pre-osteoclasts and also the lifespan of the osteoclasts. The more the osteoclast live at the surface of the bone, the more it will eat the bone up. It’s exactly the same for osteoblasts, bone formation depends on the number of osteoblasts at the surface and on the time they spend synthesising collagen. We’re going to see that the intracellular molecular mechanisms that regulate the number and the survival of the cells at the surface of the bone are very important for controlling bone remodelling.

Slide 6

So what’s the play about?

Slide 7

This is a remodeling period: you begin the story by a smooth surface of the bone that was covered by lining cells that have been retracted a little bit and then you have an activation and pre-osteoclasts are recruited at the surface of the bone, then you have fusion of pre-osteoclasts and the osteoclasts will resorb the bone and eat it up leading to a lacuna. Then you have a not very well known phase which is called the reversal phase where monocytic cells that smoothen down the bottom of the lacunae and will deposit some matrix and make what we call the cement line which is the bottom of the future osteon that will be built up by the osteoblasts. Then you have the osteoblasts that are recruited exactly where the bone has been resorbed and they will synthesize the matrix and the matrix will be mineralised. The first mineralization is called the primary mineralization. Then once the osteoblasts have covered the lacunae and filled up the lacunae they disappear, they become either lining cells or osteocytes or they die of apoptosis. Then the process which is called secondary mineralization takes place which is accumulation of minerals within the bone matrix in a cell independent manner. Then we will go back to the quiescent phase.

Slide 8

Here you can see a bone remodeling unit at work in cortical bone. Here is the resorption phase with TRAP positive osteoclasts here at the front of the perforating tunnel. Then the bone formation with the osteoblasts. Osteoblasts are making the bone and then you will have a narrowing of the hole because the matrix is synthesized and then mineralized and then you will have the process of secondary mineralization. So one thing to keep in mind again is the fact that this perforating tunnel is working in the bone, is going forward and the speed of this movement has been calculated. Keep in mind that this perforating bone remodeling unit is centred by a vessel. And why is the vessel so important? Because the vessel will bring the precursors, further precursors of the osteoclasts but also precursors of osteoblasts. There is no bone marrow around and you will see here that a precursor of osteoblasts will be brought by the vessel. You also have here another cell which is called a pericyte which is a cell that belongs to the smooth muscle cell family and which is able to transdifferentiate into real osteoblasts.

Slide 9

What about the set now?

Slide 10

First of all we’re going to talk about the molecular mechanisms that control the osteoclastogenesis. So you know the OPG RANKL / RANK system.The stem cells here and then there is a binding of M-CSF with the c-fms receptor and this will go either through the osteoclast lineage, either through the survival and the proliferation of the monocyte-macrophage lineage. Then you have these cells here that bear the receptor RANK which binds RANK ligand. RANK stands for receptor activator of NF-kB and it is a TNF family transmembrane molecule. So when RANK ligand binds to RANK, you have an activation of TRAF6 an adaptor molecule inside the pre-osteoclast and this will lead the commitment of the osteoclast to the osteoclastic lineage. Then you have proliferation of the pre-osteoclasts that then will be fused and then resorbing osteoclasts occur at the surface of the bone.

Slide 11

Fortunately for our bone there is a counter actor for this system and this is osteoprotegerin which is also synthesised by stem cells or pre-osteoblasts and when OPG binds to RANK ligand, then you can’t have any longer binding between RANK ligand and RANK and you won’t have anymore osteoclasts around.
So something which is important to understand is that most of the hormonal systems like TGF-β or oestrogen that decrease osteoclastogenesis or TNFalpha  also or other hormones which stimulates osteoclastogenesis signal through this RANK ligand RANK osteoprotegerin system.

Slide 12

What happens when you knockout osteoprotegerin? There is nothing to prevent RANK ligand and RANK binding and then you will observe very severe osteoporosis. On the opposite when you knock out either RANK ligand or RANK, you won’t have any binding between both the transmembrane molecules and the receptor and you won’t have no longer osteoclasts. Here is a TRAP staining of the wild type and you see on the KO that there is no pink labelling here and you have a very high bone mass with osteosclerosis, osteopetrosis and this is basically the same when you knock out the RANK receptor.

10.05.865

The precise molecular mechanisms that control eventually the induction of the osteoclast specific genes like the calcitonin receptor, the β 3 integrin and so on and so forth are now well known. M-CSF binding first will activate RANK expression and then you will have associated other signalling molecules that will activate the NFAT molecule which is a transcription factor , a major factor for inducing these osteoclast specific genes. Something which is important to bear in mind is that NFAT depends on the activation of cFOS which is an AP-1 member and also depends on the intracellular calcium and the calcineurin signalling ..

Slide 14

Few years ago when we were talking about the control of osteoblasts and osteoclasts it was a little bit frustrating for the people from the osteoclast field because things happened to be like the osteoblast was the king of remodelling because it controlled bone formation of course but also bone resorption because  osteoblasts synthesise both RANK ligand and osteoprotegerin. 2 years ago Brennan Boyce wrote this nice review in Nature Medicine about a paper that showed that osteoclasts were no longer osteoblasts’ slave for a couple of reasons. First of all, it was shown that this molecule which is a subunit of the ATPase which is basically a proton pump unit, the proton pump is very important for the mature osteoclast function because it spits out the protons in the lacunae here and helps the bone to be resorbed but very surprisingly the subunit was very important not for the function of the osteoclasts but for the fusion and it showed that when you knockout this subunit, you have a decreased osteoclast recruitment and osteoclast fusion. Unexpectedly this showed that there was an increase in osteoblast proliferation because in the absence of this subunit you have an increased bone formation with lots of osteoblasts and pre-osteoblasts at the surface of the bone. The second argument was the binding between ephrin B2 and the ephrin receptor B4 and this is a very interesting signalling system because you have forward and backward signalling events. For instance, when you go reverse from the ephrin receptor to the ephrin B2 you inhibit osteoclast formation through the inhibition of NFAT and on the other hand, when signaling goes from ephrin B2 to ephrin receptor B4, then you observe an increase in osteoblast formation, pre-osteoblast formation and osteoblast differentiation. These are two mechanismswhich show that osteoclasts are also able to control somehow osteoblast differentiation and proliferation and therefore bone formation.

Slide 15

What about osteoblast differentiation? A few years ago we did not know what was  the master gene of osteoblast differentiation. We knew that, for instance, MyoD was the master gene for muscle differentiation, that Sox 9 was the master gene for chondrocyte differentiation and eventually researchers found that the master genes of osteoblast differentiation was Runx2 at this time it was called Cbfa1. When mice are  Runx2 deficient, you can see here that there is no pink in this skeleton in these Runx2 knockout mice because there are no osteoblasts around. You can’t have an osteoblast without Runx2. So Runx2 is very important for the osteoblast differentiation but also for the osteoblast function. It’s not the only one, there are other genes (I don’t have the time to go through them) but we do know that Runx2 is definitely a major target for all the hormonal and local factor systems that control the bone formation.
Another important point to keep in mind is that the mesenchymal cell here which is able to differentiate in several different cells like muscle cells, chondrocytes or osteoblasts can also differentiate into adipocytes. We do know now that there might be a balance between the osteoblast differentiation and adipocyte differentiation and maybe when you have an increase in adipocyte differentiation, this might be at the expense of the osteoblast pool. This is a matter of concern because you know that we have now treatments for diabetes, the glitazones, that are PPAR gamma agonists and might, at least in animal models, be able to enhance adipocyte differentiation and therefore, decrease osteoblast differentiation. It has been shown recently that diabetic patients under glitazones may have an increased risk in fracture.

Slide 16

What about the new signaling pathway that controls bone formation? I have to say that the biggest breakthrough in the last years was the involvement of the wnt signalling pathway. The importance of this pathway was discovered when some research teams found out that this very rare metabolic bone disease that induces troubles in the eye development and a severe osteoporosis was due to the mutation, the inactivating  mutation of Lrp5 which is an LDL receptor related protein 5 and this Lrp5 is a co-receptor of the receptor of wnt, frizzled.

Slide 17

So mice were generated that lacked the Lrp5and : these mice exhibit very low bone formation and very low bone mass.

Slide 18

So what happens in the cells when the wnt signalling pathway is triggered?
wnt the signalling ligand here, (there are something like 12 types of wnts), binds to the receptor frizzled (there are several types of frizzled. When wnt binds to frizzled, then the canonical activation pathway is ON and then you have a binding with a co-receptor Lrp5 (which is mutated in the metabolic diseases I showed earlier) and this stabilizes the βcatenin which is the molecule that is going to go to the nucleus and activate with transcription factors the genes that are specific of osteoblast differentiation and survival. In contrast, when the pathway is OFF, or when there is an inhibitor of wnt around, the βcatenin which is phosphorylated by a complex that includes GSK3β, an enzyme which is inhibited by lithium that is the current treatment for some psychotic disorders. When β catenin was phosphorylated, it is addressed to the proteasome and it’s degraded.
So you can see that here we have a potential switch on/off of osteoblast differentiation and of bone formation going through the stabilisation or non-stabilisation of βcatenin.

Slide 19

There are a lot of inhibitors of this system. Here you have the normal system with activation of the canonical pathway but here is DKK1 which binds to Lrp5 and therefore, there is no binding with the wnt bound to frizzled. So β catenin degraded. DDK1 can also bind to another co-receptor which is called Kremen and this will induce endocytosis and also induce the degradation of β catenin. Of course things are not complicated enough:  DDK 2 can bind to Lrp5 and stabilise β catenin independent of the wnt pathway.
So you see it’s a very complicated system and DDK1 is one of the family members of inhibitors but you have other families we’ll go back to this later.

Slide 20

So why am I insisting on DDK1? Because DDK1 was shown to be involved in, several joint diseases, in rheumatoid arthritis and in osteoarthritis.
You can explain most of the abnormalities, maybe it’s a simple model, just by the balance between two systems: DDK1/wnt and osteoprotegerin / RANK ligand. So in rheumatoid arthritis you will have a decrease in bone formation due to DKK over expression and increase in bone resorption that explains the resorption of the joints in rheumatoid arthritis and on the opposite in osteoarthritis the wnt pathway is activated, with an increase in bone formation that will explain the osteophytes and no bone resorption because of low levels or relative levels of RANK ligand.
So we can now try to have a global universal explanation and a link between metabolic bone disease and joint disease and for a rheumatologist that is a dream come true.

Slide 21

I was talking about inhibitors I talked about DDK1 but you know there is also another inhibitor which is called SOST. SOST stands for sclerostin and sclerostin interestingly is only synthesised by osteocytes and you can see here that the immunohistochemistry shows that only osteocytes express the SOST molecule.

Slide 22

SOST is like DDK1, it inhibits the wnt pathway and of course, when you over express SOST you observe an expected decrease in bone mass at the expense of bone formation. You can see here there is no tetracycline labelling any longer when SOST is over expressed in bone in these mice and you can see that they have a very low bone mass.

Slide 23

SOST inactivating mutation was described in humans: it’s called the van Buchem disease and these patients have very, very thick bones as you can see here at the level of the skull, the wrist and the fingers. Sometimes these patients, unfortunately, experiment cranial nerve impairment like hearing loss or blindness because of the strain induced by the bone on nerves. All the biotech and the pharmaceutical companies have jumped on this potential therapeutic target and there are indeed phase 3 clinical trials ongoing with anti-SOST antibodies, inhibitors of inhibitors of bone formation which, in animal models, have improved bone formation . We definitely need treatments that increase bone formation. We only have currently intermittent PTH to treat osteoporosis with very severe bone loss.
In summary here, you can see that the osteocyte, which I want to remind you, is also the cell responsible for mechanosensing in bone, for the synthesis of FGF23 and is a phosphate sensor in the bone too, synthesizes SOST. SOST is an inhibitor of bone formation. SOST downregulates the wnt signalling pathway and thus,  decrease bone formation. If there were no osteocytes, bone formation would go on and on.

Slide 24

Most interestingly, it was also shown that osteocytes may also be the controllers of bone resorption. Recently, researchers “killed” osteocytes in bone. How did they do that? They just put killer genes driven by the promoter of DMP-1  which is  specific for osteocytes. When a drug is given to the mice, the promoter is activated, the gene product kills the osteocytes and what happens? Most intriguingly bone is invaded by osteoclasts. The osteoclasts do not resorb bone on the surface, they dig a tunnel in the trabeculae and they dig holes. You can see this here, this is the bone that doesn’t bear any living osteocyte anymore. So the osteocyte appears to be a very important cell that controls bone formation and bone resorption.

Slide 25

I’ve talked about this wnt signalling pathway. I don’t have the time to go through all the hormonal systems that control to bone formation. I would just like to show you that not only soluble molecules are able to control the bone remodelling but also some matrix molecules can do that as well. I just want to show you a paper from our lab that was just released last week and shows the results of the BSP knockout. BSP stands for bone sialoprotein which is a member of the sibling family, a protein that is synthesized by osteoblasts and binds calcium. When we knocked out BSP in mice, we observed a decreased bone formation and decreased bone resorption and we found a higher bone mass as compared to controls and an impaired osteoclast function..

Slide 26

So what about the direction now, the direction of the play?

Slide 27

I’m not going to go through every system but you can see that bone is in the middle of a network of different controllers. Calcium and phosphate metabolism I won’t talk about it, load and mechanical strain, sex steroid, reproduction hormones, nervous system. Since we are in a bone transplant session I will talk about immunity and I will also talk about the nutrition and energy metabolism and begin with this.

Slide 28

Bone is in the centre of a network that regulates energetic intake and expenditure and this involves also the brain and the nervous system and the fat tissue.

Slide 29

Eight years ago it was shown that leptin, an adipokine, which is synthesized by adipocytes induces anorexia. Leptin Knock out induces obesity in mice .These animals exhibit a low fertility and they have high cortisol serum levels. Despite the fact they have low sex hormones, despite the fact they are obese, these mice have a higher bone mass than the wild type and this is explained by a higher bone formation with very low or very subtle alterations in bone resorption.

Slide 30

When one injects leptin within the brain, one induces bone loss. Thus leptin exerts negative effects on bone mass when administered it in the brain (third ventricule).

Slide 31

It was shown later on that the leptin effect goes through the β adrenergic sympathetic system because the osteoblasts, as you can see here, express β adrenergic receptor and when one gives the mice an agonist of the β2 adrenergic receptor, bone loss occrus. When a β blocker is given, one prevents the deleterious effects on bone of intracerebral administration of leptin. So this was a real breakthrough in the bone field I can tell you.

Slide 32

This was confirmed later on by the fact that the mice that have a knockout for the β 2 adrenergic receptor have expectedly a higher bone mass and also different abnormalities.

Slide 33

In contrast, we showed that, while leptin has a negative central effect on bone through the β adrenergic system, it exerts positive effects when it is administered at the periphery (subcut or intraperitoneal injections) .

Slide 34

Others showed similar results in other models.  But Leptin is not the only nutritional adipokine that is able to control bone formation.

Slide 35

Indeed, when you look at all the adipokines; adiponectin, visfatin, resistin, they all have receptors on bone cells and thus, potential effects on bone mass. Not only the adipokines but all the hormones involved in nutrition and energy intake control have effects on both brain and bone. This is true for insulin and amylin, ghrelin which is synthesised by the stomach, incretin which are hormones that are also involved in the glucose metabolism. Very soon, pharmacological treatments of diabetes manipulating the incretin pathway, that might also have an effect on bone, are going to be released. I won’t talk about the cannabinoid system.

Slide 36

Recently, it was shown that Bone could in turn control glucose and energy metabolism by releasing molecules in blood circulation such as ostéocalcin.

Slide 37

Slide 38

So since we were in the bone transplant session I wanted to show you that also it was a breakthrough when Pacific’s team showed us 8 years ago that T cell deficient mice don’t lose bone when they are ovariectomised. This was very interesting and what they showed in this paper was that the lack of oestrogen induced osteoclastogenesis through an increase in proliferation in T cells. This was the birth of this new research field is called osteoimmunology.

Slide 39

So now we do know that T cells are able to activate osteoclastogenesis because they do express RANK ligand and we also do know that in some diseases some subsets of T cells are able to synthesise some osteoclast activators like IL-17. IL-17 is able to activate macrophages and synovial cells to express RANK ligand and induce increased osteoclastogenesis.

Slide 40

Many immuno-modulatory molecules are unexpectedely involved in bone metabolism and remodeling 

Slide 41

So this is very important because we know that for instance, in diseases like rheumatoid arthritis bone loss occurs and is linked to the increase in inflammation and it was recently shown by our team and colleagues in Lyon that when you give anti-TNF antibodies, you block bone loss induced by inflammation whether the patients are responders or non-responders in terms of inflammation to the anti-TNF antibody which was infliximab.

Slide 42

So in conclusion we have bad news:  it’s getting more and more complicated I agree but the good news are that we know better bone remodeling, so we will have more therapeutic targets. We know that the osteocyte is the master of bone remodelling and definitely became a therapeutic target. We are able to manipulate bone remodelling not only with pharmacological molecules but also with things like nutrition or mechanical load. This opens wide therapeutic prospective not only for osteoporosis but also for other metabolic boen diseases including bone transplantation bone disease.