
WNK KINASES AND HYPERTENSION: ION TRANSPORT AND BEYOND |
Xavier Jeunemaitre, Paris, France
|
Chair:
Pierre Corvol, Paris, France
|
Friedrich C. Luft, Berlin, Germany |
|
Prof X. Jeunemaitre
INSERM Unit 772, Collège de France, Faculté de Médecine Paris-Descartes, Paris, France |
Slide 1
Thank you Fred and thanks to the organisers for inviting me to just share with you some of our knowledge about the WNK kinases.
Slide 2
As it has been discussed before as you see the monogenic forms of hypertension are easier to decipher than the usual essential hypertension and the few laboratory in this case has been very successful to actually identify most of the genes that are involved in the mendelian form of hypertension.
Slide 3
As you see, most of the cases were more or less known or could be predicted, if you are looking at the Liddle’s syndrome the fact that these patients were very sensitive to amiloride, for example, and made the ENaC and all these subunits alpha, beta, gamma subunits very good candidates for this syndrome. This is not exactly the case for the so-called Gordon’s syndrome since we have identified with Rick Lifton, 2 new genes which are usually not good candidates for hypertension at all.
Slide 4
Looking at the clinical description of this syndrome when you’re looking at the literature, it is quite puzzling. Actually from one report to another one you always find some sort of autosomal dominant transmission, we saw some rare recessive cases have also been described. You have always hyperkalemia and these levels of potassium are quite variable from 5.5 to 6.5 despite normal glomerular function, so if you have normal creatinine levels, you have metabolic acidosis quite variable too with hyperchloremia always plasma renin activity which are relatively low and aldosterone which re variable depending on potassium levels and a very high sensitivity to thiazide diuretics. This is probably a very constant feature when you’re looking at the different cases that have been reported. Other abnormalities have been reported in very few cases and I’m not sure that it’s really the problem of the Gordon’s syndrome itself. So a phenotype which is variable when you’re looking at a family like this one for example, you’re looking at an autosomal dominant transmission, you see that the level of blood pressure is evidently higher than in the normal population for this brother and his mother but it’s not a huge hypertension and his daughter and his son who are affected actually have normal blood pressure at this young age but obviously they have the biological feature with hyperkalemia, hyperchloremia and a tendency for metabolic acidosis and all of them responded very well to very small doses of thiazides.
Slide 5
When you have looked at the different clinical features of these 20 or 25 families or cases that we have gathered in the last 5 years, obviously you have good relationships and a strong correlation between the different abnormalities, between the bicarbonate levels and potassium, the chloremia and potassium or between diastolic and systolic blood pressure. The thing which is interesting is that when you’re looking at the age and blood pressure, you see that you have an increase of blood pressure according to age but again before the age of 20 you don’t have hypertension per se like for example, in Liddle’s syndrome.
Slide 6
So variation in blood pressure, variation in the intensity of ionic abnormalities, variation in some other abnormalities like for example calciuria while you have hypercalciuria in some defects and not in others. Possibly variation in transmission.
Slide 7
As you see, this corresponds to some genetic heterogeneity and for the moment you have 3 loci that have been described, one on chromosome 1 and it was already almost 10 years ago by Rick Lifton and colleagues and it has not been proved yet that this locus is a real one with a real gene. These 2 loci on chromosome 17 and 12 which correspond to this kinase and we will see more on that and probably for most of the families that we have studied for the moment other loci and other genes are involved but we don’t know yet these genes.
Slide 8
The story began by a large family like this one from the North of France where you have a nice autosomal dominant transmission and we found just by classical reversion I think that we had a positive linkage on chromosome 12. This positive linkage was also found in this pedigree studied by Wilson and the thing which was interesting in this particular pedigree that they ended up with a new allele and this one was as you see, corresponding to a deletion and this deletion was a very particular deletion, 40kB deletion in intron 1 of this with no lysine kinase gene. After that we showed that this deletion was also present in the first pedigree but a smaller one which was not taking the first microstatic marker within it. We also found that there was a change in expression of the WNK1 gene within leukocytes of controls and cases, suggesting that this deletion corresponded to a change in expression of the gene in periphery or at the kidney level.
Slide 9
Coming from this gene was relatively easy knowing the human genome because of the similarities between the WNK1 and WNK4, this was a second locus on chromosome 17 and on this chromosome 17 you have that gene a homology which was between 75 and 50% depending on the domains between these 2 kinases. The particular mutation that was found in WNK4 was this one was a missense mutation in very peculiar domains which are close to CC domains so probably domains interacting with different other proteins. Itself close to an auto inhibitory domain and the kinase domain. So for this gene also not some sort of loss of function mutation usual in genetics but rather particular missense mutations that saying that it was playing probably again a function somewhere in the cascade.
Slide 10
So what are these with no lysine kinase genes? The WNK1 gene has been cloned in rats by Melanie Cobb in 2000. It was the first study at that point and it was corresponding to a new family of 7 tyrosine kinase and this family is corresponding to something which is in terms of kinase activity peculiar because of the absence of lysine usually located in the subdomain 2 but which is now located for this family on subdomain 1. So that’s why she called this family, with no lysine kinase which is funny, interesting and probably good in terms of memory. These two kinases WNK1 and WNK4 as expressed in the kidney, WNK 4 is mostly present in the kidney, WNK1 is expressed in the kidney but in other tissues because of the 2 transcripts that we are going to see now.
Slide 11
Indeed, WNK1 is more complex in terms of study than WNK 4 because you have these two isoforms, the first one is a full length isoform depending on proximal parameters with the kinase domain which is here in red, the auto inhibitory domain and the 2 coils domains. You see this isoform is expressed in most of the tissues but mainly in skeletal muscle, heart and brain and you see 2 bands corresponding to 2 probably different polyA sites. The other one, the short kinase specific isoform is devoid of any kinase activity because you do not have any kinase domain there. You have the auto-inhibitory domain, the two coiled coil domains and probably this kinase is playing a role in terms of interaction but not at all in phosphorylation. This kinase here, kidney specific is expressed in the kidney not in other tissues and mainly in the distal tubule.
Slide 12
For the other ones you see that it was interesting to observe and it was preliminary results which were very important for us to establish a mouse model that the pattern of expression in mouse is mainly similar to what we observe in humans and that the organisation of the WNK1 gene is similar to what we observe in the mouse. You have a big hWNK1 here 60kB, here 40 kB. You have a particular renal promoter which is upstream of a particular exon which is called exon 4A and as you see this exon 4a is very homologous when you’re looking at the mouse and human gene, this is the same thing when you’re looking at the dog sequence, it’s the same thing when you’re looking at the rat sequence, for example.
Slide 13
When you’re looking at these two isoforms WNK, WNK1 and WKN short and long isoform, you see that the kidney specific short kidney is predominant, it’s approximately 90% of the expression when you’re looking at the whole kidney by qPCR and when you are looking by in situ hybridisation you see it very well in the cortex here and you see also there that it’s mainly located in the distal tubule here close to the glomerulus. Actually when we performed qPCR with macrodissected tubules it was interesting to observe that you have a low expression of the long isoform of WNK1 all along the tubule, that you have an expression of WNK4 which is all along the tubule but mainly in the distal tubule and collecting tubule and collector tubule and that you have a very short expression of the kidney specific WNK1 there in the DCT1, DCT2 and CNT. When you’re looking at that particular portion of the tubule, as you see you have many other partners which are of interest either regulators like AGK1 or the M receptor or different transporters like ROMK, NCC or the icaC for example ENaC.
Slide 14
So as soon as the genes were known, it was very interesting to test if this particular kinase could interact with these different transporters either directly or indirectly. This possibility of interaction with NCC because Gordon’s syndrome is more or less a mirror of the Gitelman syndrome possible interaction with ROMK, with ENaC or even on the NCC.
Slide 15
Actually, I would say that it’s very astonishing that the studies were so successful. Most of the groups that have studied the WNK4 have shown that at least in – site you can find an interaction between WNK4 and most of the transporters that were good candidates to explain the fact that mutation of WNK4 could explain the Gordon’s syndrome.
Slide 16
When you’re looking at the different studies it seems that WNK4 in vitro inhibits the sodium chloride cotransporter NCC, stimulates the internalisation of ROMK, increases the paracellular chloride permeability, interacts with ENaC and possibly interacts with CRP4 and icaC. So obviously a role which is very important in terms of regulation of the different transporters and the regulation between the sodium, potassium and even chloride or even calcium balance.
Slide 17
This is more or less explained there by this review that was published in Kidney International a few months ago. When you are looking at the effects of WNK4 the interesting thing is that not only this serotonin kinase seems to have a role by phosphorylation of other partners especially SPAK and OSR1 and then possibly by this kinase playing a role on NKCC 1 or different potassium channels and also by phosphorylating the claudins but also it seems that per se not by the phosphorylation, it seems to have an effect on the endocytosis of NCC, ROMK, Trp4 or even Trp5. So a very complex role not only on one transporter but on different transporters and this role is played by not only phosphorylation but also interaction with different partners which might explain the consequence of the mutation.
Slide 18
When we’re looking at WNK1 and the studies that have been performed on WNK1 more or less say the same thing even though it’s even more complex but in epithelial cells, groups are found an effect of WNK1 either L-WNK1 or the KS-WNK1, an effect on NCC, ROMK, ENaC and these kind of transporters.
Slide 19
If you want to summarise this effect it’s more or less for WNK1 an inhibition of WNK4 and this interaction itself playing an interaction on NCC. Possibility of also interaction on ECaC. For L-WNK1, so the form which has a phosphorylating activity you can find, at least in vitro, a phosphorylation on AGK1 which might explain itself a further phosphorylation of WNK4 and then an effect on ENaC.
Slide 20
This L-WNK1 could also interact with the internalisation of ROMK and the fact which is interesting in the studies that have been performed is that it seems that the KS-WNK1 which has no kinase activity could interact with L-WNK1 by just multimerization and this multimer between KS-L-WNK1 other proteins probably WNK4 for example, could itself regulate their activities on transporters or other proteins. So when you’re looking at the gene itself, the fact that you have proximal promoters, the possibility of L-WNK1 expression or the possibility of the short kidney expression or the balance between KS and L-WKN 1 will be very important in terms of regulation of this particular protein on the transporters. Actually, more or less two studies at least and our group also have found that when you put people on high potassium diets or people, rats actually rats or mice you increase the ratio between KS and L-WNK1 so you increase KS-WNK1 not L-WNK1 at least in gene expression not on protein because we had some difficulties in terms of antibodies and if it’s true, this increase of KS-WNK1 might well explain the fact that in that case you have an inhibition of L-WNK1 and by different inhibition or positive action you would have in that case in case of a high potassium diet an increased activity of ROMK because of an increased number of potassium channels at the membrane.
Slide 21
When we are trying to hypothesise what could be the mechanism of the mutation corresponding to the FHH deletion and from the preliminary results that we have in vivo in mice what we could hypothesise is that in that case because of, I’m not going to enter into details in terms of transcription but because of important sites that are within the intron 1 you would have an expression of L-WNK1 and KS-WNK1 which would be over expressed in the kidney but more L-WNK1 than KS-WNK1. In that case the equilibrium between the two isoforms would be different and if you increase L-WNK1, you are going to inhibit WNK4 more and then WNK4 being itself an inhibitor of NCC, you will increase NCC activity. Same thing when you’re looking at L-WNK1 its possible an effect on ENaC and ROMK, it would be logical then to think that you would have a decreased number of ROMK channels at the membrane and an increased ENaC activity. So all this would fit with the physiopathology that we have for the moment for the disease.
Slide 22
I’ll be taking 5 minutes just to show you the expression in heart, vessels and brain of WNK1 which I think might be of interest in terms of pathophysiology.
Slide 23
What we have done in the lab is to construct a transgene with a bac corresponding to a 5' sequence of about 50 kB upstream and 10 kB downstream and to put a reporter gene which was corresponding to the L-WNK1 isoform with a stop codon for the short isoform then we would not have any expression of the short isoform. Different founders were found, different lines were studied and more or less what we found during development is that you have a very strong expression of the L-WNK1 isoform.
Slide 24
And this early expression is ubiquitous but what was surprising is that when this expression is in the heart as in adults it’s not very surprising but when you’re looking at the expression of WNK1 in different vessels like that here in the aorta it’s quite a strong expression.
Slide 25
Same thing when you are looking at the adult mice, you find that you have a strong expression here like the in the arteries and veins and mainly in the endothelium but also in the smooth muscle cells and you see here the strong expression in the aorta.
Slide 26
Here the expression of LacZ in one artery within the kidney. So quite surprising is that we have a strong expression in the vessel. Same thing when we look at the mesenteric vessels here which is all blue and this corresponds with the in situ hybridisation that was performed by Celine Delaloy and you find that you have exactly the same antigenic expression that in the transgene. Same thing when you’re looking in the muscle. So this expression of L-WNK1 in the cardiovascular system might be of interest since at least in terms of blood pressure the knockout model is deleterious when you have a knockout complete model but the heterozygous mice have a decrease in blood pressure which is not explained it seems by the effect on kidney.
Slide 27
Last thing, it’s possible on the cerebellum.
Slide 28
It was very surprising to us also to observe that the long isoform of WNK1 is not expressed in the brain but mainly expressed I don’t know if you see it very well there in the cerebellum. This is very located to a particular domain which is a granular layer of the Purkinje cells and this expression is not only found in the transgenes but also by in situ hybridisation of the endogenous gene.
Slide 29
Same thing when you look at the antigenic expression by different probes and this one for example where you see a very nice expression of this in toto cerebellum in situ hybridisation.
Slide 30
This is of interest when you’re looking at L-WNK1 in a more larger way than only blood pressure, ionic reabsorption but also on possible effects of ion transport in large and it was published two years ago that by just a double hybrid system that a particular protein which is synaptotagmin 2 is important in terms of transport of different proteins in roles and this synaptotagmin 2 was interacting with WNK1 and when you are looking at the expression of WKN1 and synaptotagmin 2, you see that at least in this panel, for example, it’s very well co-localised and we have very nice results on that too.
Slide 31
So I will end up my talk just saying that the Gordon’s syndrome or familial hyperkalemic hypertension was very interesting in showing two new genes that maybe are involved in the transport of ion in the kidney and this transport of ion in the kidney is complex and set up a new kind of regulation but this story has not ended yet on the kidney, it maybe and it is probably the case that these kinases and especially WNK1, it might be the case also of WNK3 that these kinases are binding other proteins, are phosphorylating other proteins as this one for example, in the cerebellum.
Slide 32
I would like just to thank the people working in the hospital Pompidou who have done a great job trying to identify the families. People in the clinical investigation centre, a particular thanks to Celine Delaloy and Juliette Hadchouel for the results that I showed on WNK1 and our collaborators. Thanks again.
Slide 33
Chairman: This is terribly complicated.
Prof Jeunemaitre: Life is like that you know.
Chairman: WNK1, how does the autosomal dominant disease work? You’ve got a deletion in one of the genes of which each individual has two copies. What happens that this is an autosomal dominant disease?
Prof Jeunemaitre: To my view the fact is that as I showed in one slide, the fact that you have an over expression of the L and KS-WNK1 but more say a disturbance in the equilibrium between the two isoforms by one allele it is sufficient to not be counteracted by the other one and that between this default of equilibrium you just have after that a cascade which is more in one way than in the other one but it’s just obviously a hypothesis at the moment.
Chairman: The WNK1 patients don’t seem to have a cerebella phenotype.
Prof Jeunemaitre: Not yet we have not studied that in detail but obviously it will be studied in mice and we are planning to do that also in humans. We have no particular phenotype, spontaneous phenotype..
Question: With WNK 4 you mentioned gain of function mutation, did I understand that properly?
Prof Jeunemaitre: Well, it depends at which level you think of gain of function, gain of function, loss of function in terms of interaction but I think that it’s not a classical loss of function if you, for example, destroy the complete WNK4 as a classical genetic mutation.
Question: I assume both of these kinases have been knocked out in mice.
Prof Jeunemaitre: Yes.
Question: Tell me very briefly what do the phenotypes look like if these kinases are gone completely?
Prof Jeunemaitre: The only model that I’ve seen published on WNK1 is Liddle because of the deletion of L-WNK1. We have now in our laboratory a knockout of KS-WNK1 which is not Liddle.
Chairman: Ok. Other questions? Well, I’d like to thank all the speakers and thank you all for listening and coming. Real science wonderful afternoon thank you.