Slide 1

Thank you very much. I want to thank the organisers for the invitation to this meeting and the possibility to talk about the mechanisms of injury in uromodulin associated kidney disease. Uromodulin was discovered over 50 years ago and so its history is older than many of the investigators working on it including myself. However, the cellular function of uromodulin is still not very well understood. So, I want to invite you to a journey through time showing you what research over the last decades has revealed about uromodulin.

Slide 2

58 years ago 2 microbiologists Tamm and Horsfall discovered a novel nuclear protein in human urine which inhibited viral hemagglutination. Subsequently this protein was named Tamm-Horsfall protein. In 1985 Pennica et al published in Science that the cDNA sequence of Tamms-Horsfall protein corresponded to the cDNA sequence of another urinary protein named uromodulin. They concluded that both Tamm-Horsfall protein and uromodulin are identical. In order to avoid any further confusion, I will only refer to this protein as uromodulin in this talk.

Slide 3

So what are the characteristics of uromodulin? Uromodulin is a complex 80-90 kDa renal glycoprotein which is heavily modified by N- and O- glycosylation. Uromodulin is primarily synthesised in the kidney where it is expressed in the thick ascending limb of the loop of Henle. Here it is soldered to the apical membrane.

Slide 4

Additional characteristics of uromodulin are that it’s a transmembrane protein which is secreted into the urine by a cleavage of a so-called GPI-anchor which is the abbreviation for glycosylphosphatidylinositol anchor. Uromodulin is the most abundant protein in human urine with a secretion of up to 100 mg/day. Uromodulin is not very well conserved in evolution it’s only conserved ---.

Slide 5

So if this protein is so strongly expressed in urine, what is its function? Over the last 58 years several aspects have been considered. In the literature you can find involvement in the following diseases. Uromodulin functions as a protective factor against the development of nephrolithiasis. It serves as a host defence mechanism against urinary tract infections. Moreover, uromodulin has been found to be involved in the formation of myeloma and cast nephropathy. Finally, it has been hypothesised that uromodulin is involved in concentrating urine by forming a gel-like structure which results in the water impermeability of the thick ascending limb. All together we have a wide spectrum of suggested functions and the conclusion is that the cellular function is still not very well understood.

Slide 6

However, in the context of uromodulin’s role and the development of urinary tract infections several immunostimulatory activities have been found. Uromodulin is able to activate human granulocytes and monocytes. We have a TLR4, uromodulin can activate human dendritic cells. Uromodulin can stimulate TNF-α production and can also interact with other parts of the immune system like IL-1, complement 1 and 1q and so interaction of uromodulin with the immune system has been suggested.

Slide 7

Clinical and experimental studies had indicated involvement of uromodulin in the pathogenesis of several inflammatory kidney diseases. A role was suggested for acute and chronic pyelonephritis for reflux nephropathy and for interstitial nephritis. In 1980 Hoyer, a paediatrician from Boston, showed an active role for uromodulin in the pathogenesis of interstitial nephropathy. He immunised rats with uromodulin and he discovered that these rats had tubulointerstitial inflammation and microscopic scarring. The immunised rats showed large PAS positive deposits at the cortical thick ascending cells and mild peritubular leukocytic infiltration. In addition, he discovered antibodies to uromodulin and cytotoxic T cells with uromodulin specificity indicating an immune response to uromodulin.

Slide 8

In 2002 the scientific field of uromodulin changed significantly when mutations in uromodulin were discovered. The group of Hart et al discovered three mutations in three different kindreds with a disease named familial juvenile hyperuricemic nephropathy abbreviated as FJHN and one mutation in a kindred with medullary cystic kidney disease type 2 abbreviated MCKD2. In the mean time 38 different mutations have been published. The mutations are heterozygous indicating an autosomal dominant inheritance. The identified mutations are mostly missense mutations and only a few in frame deletions have been published. No non sense mutations have been found yet. The mutations cluster in two exons, exon 4 and 5 and both of them encode for calcium binding, epidermal growth factor like domains. The vast majority of mutations, about 60% are affecting the proper encoding of the aminoacid cysteine. There are a total of 48 cysteine residues in uromodulin and they are involved in the formation of 24 intramolecular disulphide bridges. These disulphide bridges are crucial for the proper conformation of uromodulin.

Slide 9

Mutations in uromodulin are responsible for a spectrum of previously differentiated diseases and I’ve already named MCKD2 and FJHN. There’s a third disease which is called glomerulocystic kidney disease abbreviated GCKD which is very infrequently caused by mutations in uromodulin. All three diseases are inherited in an autosomal dominant way and all three diseases result more or less in progressive tubulointerstitial nephropathy. Over the last few years it has become more and more difficult to actually differentiate these diseases. So the current tendency is to summarise all 3 diseases under the name of uromodulin associated kidney disease abbreviated UAKD. But what have been the characteristics of the previously differentiated diseases? MCKD2 was characterised by urinary salt wasting, decreased urinary concentration, development of hypertension, formation of medullary cysts along the cortical medullary border and ESRD at the age of 40-50 years of age. Familial juvenile hyperuricemic nephropathy was characterised by very early onset of hyperuricemia sometimes as early as 3 years. These patients developed later on gout and ESRD. In this disease a decreased urinary fractional excretion of uric acid has a very supportive value. Finally GCKD. These patients are characterised by their renal biopsy which shows cystic dilatation of the Bowman’s space. These patients very frequently have hypoplastic kidneys.

Slide 10

In 2007 we realised that one of the mutations that we had discovered earlier occurred quite frequently in our patients. We found the C744 gene mutation in a total of 5 different kindreds coming from 4 different countries. We performed haplotype analysis and were able to confirm our hypothesis of a founder mutation. Here you can see the pedigrees of all 5 different kindreds and the affected individuals are shown in black. The black bar represents the haplotype carrying the ---mutation. Interestingly, 2 of these 5 kindreds presented with symptoms of MCKD2. Another two presented with symptoms of FJHN.

Slide 11

And the 5th kindred in retrospect had symptoms had GCKD. So one of the same mutations could actually result in all 3 different entities showing that there must be a common pathogenic pathway and this justifies that we summarise all 3 diseases under the term of UAKD.
UAKD is actually characterised in the histologic triad of tubular basement membrane disintegration, tubular atrophy and cyst development along the cortical medullary border and interstitial cell infiltration and fibrosis. These findings are similar to the rats that were immunised by Hoyer with uromodulin.
Interestingly, uromodulin associated kidney diseases morphologically and clinically are very similar to recessive nephrolithiasis. Recessive nephrolithiasis shares the microscopic features. However, UAKD is a dominant disease versus nephrolithiasis that is recessive. The age of ESRD is much later in UAKD.

Slide 12

Here you can see a renal ultrasound of a patient with a uromodulin mutation and you can see the hyperechogenic kidney and the formation of cysts along the cortical medullary border. These cysts can be found in 20-50% of patients.

Slide 13

So what are the consequences of uromodulin mutations? Typically uromodulin is distributed along the thick ascending limb and the distal convoluted tubules with a staining pattern characteristic of the apical membrane and a diffused intracellular reactivity. Luca Rampoldi and Karinda – performed studies and renal biopsies of patients with uromodulin mutations and they discovered cross accumulation of uromodulin with patchy deposits within the thick ascending limb cells and a subset of tubules that appeared to be dilated. This resulted in the hypothesis that the uromodulin mutation results in a misoflding of the protein which then results in retraction of this protein within the thick ascending limb cells.

Slide 14

These findings were somehow confirmed by additional urinary analyses. Here you can see the pedigree of a kindred with a uromodulin mutation. The affected individuals are shown again in black and you can see that every generation is affected indicating an autosomal dominant inheritance.
At the bottom you can see the urine Western Blot analysis for uromodulin and you can very easily see that the affected individuals had a decreased uromodulin secretion in urine compared to the wild type or not affected individuals. 

Slide 15

However, in the past decreased urinary uromodulin secretion was published as a non-specific finding in patients with renal failure of diverse aetiology. Karinda Howe was able to show that patients with uromodulin mutations actually had a lower uromodulin excretion in urine than patients with renal failure of other aetiology. She performed urine analysis with Western Blot in a control individual in patients with uromodulin mutations and in patients with renal failure of other aetiology. She analysed the setting in 3 different groups of creatinine clearance and she was able to persistently demonstrate that in all 3 groups the patients with uromodulin mutations had a decreased urine uromodulin excretion compared to patients with renal failure of other aetiology.
She confirmed this by densitometry analysis and she also showed that the patients with uromodulin mutations secreted only wild type uromodulin in the urine indicating that the mutated uromodulin is retained within the thick ascending limb cells.

Slide 16

Luca Rampoldi analysed 12 different mutations in uromodulin concerning their effect on intracellular trafficking and subcellular localisation. He confirmed that all 12 mutations, even though to a different extent, resulted in an accumulation of mutated uromodulin in the endoplasmatic reticulum and in a defective endoplasmatic reticulum to Golgi apparatus trafficking. Shortly after transfection of the wild type uromodulin you can see a strong signal in the Golgi apparatus and the plasma membrane, whereas transfection of the different mutant isoforms of uromodulin resulted in a strong signal in the endoplasmatic reticulum with only little if any signal in the Golgi apparatus or in the plasma membrane.

Slide 17

Jennings et al analysed 2 different uromodulin mutations concerning their affect on apical membrane targeting and direction of secretion. They discovered that both mutations still are resulting in a targeting of the apical membrane and that they are still secreted from the apical membrane. However, this group also discovered that besides the apical secretion of uromodulin there’s also a basolateral secretion of uromodulin and this is new.
This basolateral secretion of uromodulin is maintained in the cells that are transfected with a mutated uromodulin and so the hypothesis of this group is that the mutated uromodulin is actually secreted from the basolateral side of the cell into the interstitium and may there trigger the immune system which then may result in the tubulointerstitial inflammation.

Slide 18

Choi et al discovered that accumulation of mutated uromodulin in the endoplasmatic reticulum actually resulted in an increased rate of apoptosis. Treatment with chaperones in some diseases can actually rescue misfolded proteins from the endoplasmatic reticulum and so they were wondering if treatment with chaperones of transfected cells could perhaps changes the accumulation of altered uromodulin in the endoplasmatic reticulum. They transfected thick ascending limb cells with wild type uromodulin and you can see here an increased staining of the plasma membrane. They also transfected thick ascending limb cells with mutated uromodulin and here you can see an increased staining of the cytoplasm. They then treated these cells with chaperone sodium 4PBA and with colchicine and interestingly the staining pattern changed. They found an increased staining pattern within the plasma membrane and a decreased staining pattern of the cytoplasm comparable to the wild type and indicating that there is an increased intracellular secretion and improved transport of uromodulin from the cytoplasm to the plasma membrane.
Interestingly, these treated cells also showed a decreased rate of apoptosis and so this was the first group to provide a new therapeutic target for potential treatment in UAKD.

Slide 19

So please let me summarise.
Uromodulin is a renal glycoprotein and the most abundant protein in human urine. Uromodulin is involved in the pathogenesis of nephrolithiasis, urinary tract infection and interstitial nephritis. Mutations in uromodulin can cause a broad spectrum of symptoms and this can be very challenging for the clinician.
Uromodulin mutations result in a misfolded protein and this undergoes a slower processing in the endoplasmatic reticulum.
Accumulation of uromodulin, mutated uromodulin in the endoplasmatic reticulum may cause apoptosis.

Basolateral excretion of mutated uromodulin may actually activate the immune response in the renal interstitium. Chaperones and colchicine can alleviate the intracellular damage that is caused by uromodulin mutations.

Slide 20

Hopefully we will have soon additional therapeutic options for our patients. Thank you so much.

Slide 21

Chairman: Thank you for this excellent talk. It’s open for questions. Maybe while we’re waiting I want to ask you, there was a brief report a year ago or so by Rick Johnson’s group in a knockout of uromodulin that they reported that there were some problems with sodium reabsorption kind of implicating that uromodulin had some effects on sodium reabsorption and worst. This was to explain the pathogenesis of hyperuricemia and that I think only report where the knockout has a phenotype other than predisposition to a stone formation or infection. I was wondering whether you have any comments about it.

Dr Wolf: You’re referring to the hyperuricemia as a – wide developing in the patients?

Question: Well, the question is that in the knockout basically there is no phenotype that is related to the uromodulin associated kidney disease but Rick Johnson’s group again, it was a very brief communication reported, that there were some problems with the wasting of sodium, so I wonder whether you have any comments about that or whether you have any confirmation of these findings?

Dr Wolf: So in 2006 there was a study that tried to analyse if there’s any direct interaction with uromodulin and uric acid and they found that there is no interaction. So there was a search for a more indirect mechanism and Scolari and his group actually discovered that there’s a very interesting relationship between urine osmolality and uric acid. So the lower the urine osmolality is, the higher the uric acid was found to be and this indicated for Scolari and his group that there’s polyuria, volume contraction and most likely impaired sodium reabsorption in the thick ascending limb. He compares the situation to treatment of patients with loop diuretics where we also have an impaired sodium reabsorption in the thick ascending limb and as a compensatory mechanism we have then increased sodium reabsorption in the proximal tubule with an increased reabsorption of uric acid. This causes the side effect of hyperuricemia in the treatment of loop diuretics and this looks very reasonable as a comparison to our situation here and it could very well explain why we have this decreased urinary fractional excretion of uric acid. However, the mouse has an enzyme it’s called urease and therefore the mice will never develop hyperuricemia.

Question: Yes but that’s what they reported that actually they do that they measured the ratio of sodium and uric acid in the urine and they found it to be very alleviated which is difficult to understand.

Chairman: For the remaining aficionados who are in the room a last question on the mechanism and the speculation of the hypothesis of Tucker’s group on the basolateral secretion of the mutated uromodulin reaching maybe the systemic circulation. What are the arguments in favour of this hypothesis? Are there antibodies against uromodulin found in the serum? Are there increased levels of uromodulin found in the serum? What are the arguments in favour of this hypothesis?

Dr Wolf: Are you referring to Scolari’s hypothesis or ….?

Chairman: No, what are the arguments in your opinion favouring this hypothesis of systemic activation against uromodulin?

Dr Wolf: Oh, that’s a good question.

Chairman: Is it only a speculation or are there solid arguments?

Dr Wolf: There is no real data, I mean we don’t have renal biopsies where we have real time PCR measurements of concentrations or whatever. It’s a speculation, it’s a hypothesis and it looks pretty reasonable but !

Chairman: At this stage interesting speculation. Thank you very much.