Introduction:
Juvenile familial hyperuricemic nephropathy (FJHN) is a rare hereditary pathology transmitted in the autosomal dominant mode. It is characterized by a defect of excretion of the urates responsible for a hyperuricemia often complicated by early gout and by a chronic tubulointerstitial nephropathy progressively evolving towards the terminal renal insufficiency. The group of inherited tubulointerstitial nephropathies also includes nephronophthis, as well as cystic medullary disease (MCKD). Autosomal recessive nephronophthis manifests itself mainly during childhood by a polyuropolydipsic syndrome and a terminal renal failure occurring on average at the age of 12-13 years in the most frequent form. MCKD is also characterized by poorly specific tubulo-interstitial involvement, leading to terminal renal failure in adulthood. The presence of predominant cysts at the corticomedullary junction, although classic, is far from constant. It is sometimes accompanied by hyperuricemia. The FJHN and the MCKD have long been regarded as two distinct entities. Progress in recent years, including the discovery of mutations in the uromodulin gene, which encodes Tamm-Horsfall (TH) protein, in families with both diseases, have challenged this distinction.
History:
The FJHN was first described by Duncan and Dixon in 1960. They reported a case of a 19-year-old man hospitalized for renal failure and severe hypertension and who had recurrent seizures since the age of 14 of drop affecting many joints. All the investigations showed hyperuricemia, urine concentration disorder, albuminuria oscillating between 0.5 and 1.6 g / 24 hours, and small kidneys. No renal biopsy could be performed. The study of her family showed that the mother and four of the seven brothers and sisters also had hyperuricemia and kidney damage. For the first time, the authors evoked the concept of hyperuricemic nephropathy. Since the initial description, many other cases of families from around the world have also been reported. The prevalence of FHHN is not known, but the pathology is considered rare. For some authors, it is underestimated because of the lack of specificity of its clinical presentation and the inconstant character of gout. More than 50 families have now been studied in Britain.
The cystic disease of the medulla, for its part, was initially confused with nephronophthis because of their clinical and histopathological similarities. It was not until the late 1960s that Goldman, then Gardner, described large families in which the disease was clearly autosomal dominant and end-stage renal disease much later than in nephronophthis.
Genetics:
Progress in genetics has brought the FJHN and MCKD closer together. The existence of mutations in the uromodulin gene in patients with either disease has led some authors to suggest that these two entities are the different facets of the same disease.
Loci known:
If we consider families with MCKD / FJHN, that is to say, a table of autosomal tubulo-interstitial nephropathy, which is associated variable with sometimes symptomatic hyperuricaemia and kidney cysts, three different loci have currently identified. These loci: 1q21, 16p12 and 1q41 nevertheless account for only a little more than half of the patients affected. Only one of the genes responsible has been identified to date at the 16p12 locus. This is the uromodulin gene encoding the TH protein.
Locus 1q21:
This is the first locus highlighted by Christodoulou et al. within two broad Cypriot families labeled MCKD using linkage studies. This locus was subsequently confirmed by various authors and then the interval of interest, reduced to less than 650 kb, allowing the sequencing of candidate genes. The HAX-1 gene, coding for a protein associated with the actin cytoskeleton, was a good candidate, but the search for mutation in an MCKD family proved unsuccessful. The gene encoding the interleukin 6 receptor also constitutes a candidate gene. This protein stimulates the expression of the hepatocyte growth factor, involved in the suppression of cyst formation. The gene encoding the glycoprotein of the Rhesus B blood group has also been mentioned. No significant mutations were found by Wolf et al. in three families affected by the disease. Finally, the gene encoding the atrial natriuretic factor receptor, also located in the 1q21 region, was studied by Koptides et al., With no anomaly found in five Cypriot families.
Locus 16p12:
This locus is highlighted for the first time by Scolari et al. in 1999 in an Italian family. This family presents a picture associating polyuropolydipsis, renal insufficiency, arterial hypertension and evocative renal pathology.
There are no kidney cysts, but the family suffers from premature gout. In 2001, Hateboer et al. confirm the existence of the 16p12 locus within a large Welsh family with a typical MCKD pattern without hyperuricaemia.
In 2000, Stiburkova et al. and Kamatani et al. describe a binding between the FJHN and a chromosomal region located at 16p12. This link is confirmed in 2001 by Dahan et al. in a large Belgian family. Several candidate genes are mentioned (MIR16, Ks1, GPRC5B, etc.) and tested unsuccessfully by the Stiburkova team. In 2001, Pirulli et al.sequencing the uromodulin gene encoding the TH protein and located in the region of interest in two patients with MCKD. They find a silent polymorphism, but no coding mutation is identified. Since then, Hart et al. in 2002, and several teams described mutations in the uromodulin gene in many families labeled FJHN or MCKD. There was also a mutation of uromodulin in a family labeled glomerular cystic disease. More and more authors have now abandoned the terms of FJHN and MCKD to speak of nephropathy associated with uromodulin.
To date, some 40 different mutations have been reported. The vast majority of them are located in exon 4 of the gene (almost 85% of the cases). The other mutations are located in exon 5. Rampoldi et al. have demonstrated a mutation in exon 6 within a family. Most of the mutations found are point mutations, but some deletions or deletions-insertions have been described. Exon 4 of the gene contains a sequence rich in highly conserved cysteines, as well as three EGF domains binding calcium. More than two thirds of the mutations result in the replacement of a cysteine with another amino acid and are therefore likely to affect the conformation of the protein.
This disease is transmitted in the autosomal dominant mode.
However, in a Spanish inbred family with a mutation of uromodulin (C255Y), three individuals were homozygous for the mutation. These three individuals reached adulthood, but the authors noted a more severe phenotype with an earlier onset of hyperuricemia and a more rapid progression to end-stage renal disease compared to heterozygotes in the same family.
Locus 1q41:
Hodanova et al. studied in 2005 a large Belgian family which presented a table close to the FJHN, but for which the binding studies had excluded the known loci. This family, comprising nine patients over four generations, presented a chart combining goutless hyperuricemia, chronic renal failure with small echogenic kidneys without a cyst, reaching the terminal stage in three patients between 50 and 68 years, and finally a disproportionate anemia in relation to the level of renal insufficiency. In addition to decreasing the excretion fraction of uric acid, patients showed decreased urinary excretion of calcium and uromodulin. The renal histological study curiously showed a decrease in the expression of uromodulin at the level of the cove of Henle. The binding studies revealed a new locus located at 1q41. The authors studied eight other families not related to loci 1p21 and 16p12 but did not find any binding to this new locus.
The FJHN / MCKD table therefore corresponds to various genetic anomalies, yet to be discovered.
Tamm-Horsfall protein:
TH protein is the most abundant protein in normal urine excreted at a level of about 50 mg / d. It is a renal membrane glycoprotein whose functions are still poorly known. It was initially demonstrated in 1950 by I. Tamm and F. Horsfall who isolate a mucoprotein responsible for the inhibition of viral haemagglutination in the urine of healthy subjects. In 1985, Muchmore and Decker demonstrated a glycoprotein in the urine of pregnant women called uromodulin because of its immunomodulatory properties. It was in 1987 that cloning of the complementary deoxyribonucleic acid (cDNA) of uromodulin made it possible to link it to the protein of TH.
Electron microscopy studies of the protein showed that it polymerized to form a network of protofilaments. The carbohydrates represent approximately 30% of the weight of this glycoprotein. Three epidermal growth factor domains, 48 cysteine residues, eight potential N-glycosilation sites, a zona pellucida-like domain responsible for the polymerization of the protein in filaments and a hydrophobic zone at the site C-terminus. This sequence acts as a signal for the endoplasmic reticulum, which after cleavage anchors the protein to a new C-terminal membrane end.
It is a protein with only renal expression, synthesized at the cove of Henle. Ultrastructural studies by Bachmann et al.in the rat showed a prevalent localization of the TH protein in vesicles close to the Golgi cisterns, fusing with the apical pole of the cells of the cove of Henle. They also visualized a weak marking at the basolateral pole.
Its functions are still poorly known even if the genesis of KO mice for the uromodulin gene has improved our knowledge.
Defense against urinary infections:
It has been shown in vitro that the TH protein binds certain types of E. (glycoproteins to which the bacteria binds to colonize the urinary tract), probably preventing their progression. These data were reinforced by the existence of a greater susceptibility of KO mice to infections with certain types of E. coli.
Urolithiasis:
It was shown in vitro that the TH protein inhibited aggregation and growth of calcium oxalate and hydroxyapatite stones. Similarly, formation in calcium crystal urine was investigated in a model of mouse KO and found to be significantly increased compared to control mice.
Permeability of the cove of Henle:
The TH protein has the reversible ability to aggregate and tends to form a gel. Given its exclusive location at Henle’s loop, some authors have suggested that TH protein plays a role in the impermeability of this tubular segment to water.
Protein Interactions:
The TH protein interacts with other proteins present in the urine mainly during pathological situations.
Rhodes et al. have demonstrated an interaction of uromodulin with immunoglobulins of the immunoglobulin G (IgG) type. It has also been shown that the protein interacts with the immunoglobulin light chains forming the crystals responsible for myelomatous tubulopathy.
Immunomodulatory properties of the Tamm-Horsfall protein:
Muchmore and Decker in 1985 show that the protein inhibits in vitro the proliferation of lymphocytes in response to various antigens as well as the cytotoxicity of monocytes without altering the functions of the B cells. Several studies have shown that the TH protein binds and activates the leukocytes: monocytes, lymphocytes and polynuclear neutrophils. Su et al. have shown that the uromodulin-induced response to mononuclear cells in culture was a dose-dependent increase in tumor necrosis factor a (TNF- a ) and interleukin (II) -1 b secretion, pro-inflammatory cytokines , initially followed by IL-1 receptor antagonist and TNF receptor type II, and anti-inflammatory cytokines in a second stage. It has recently been shown that TH protein activates myeloid cells via toll like receptor 4, resulting in activation of the NF-kappaB pathway.
KO mouse models for uromodulin:
Two groups have so far generated mice deficient for uromodulin, by homologous recombination. Mo et al. have deleted the first four exons of the resulting gene in the total absence of protein. They did not report any effect on embryonic development and renal histology. They observed a predisposition of KO mice to urinary E. coli infections after inoculation. Similar results on susceptibility to E. coli infections were found by the Bates team in a second model of KO mice also resulting in complete absence of protein.
The Mo team has also been interested in the role of TH protein in preventing the formation of calcium calculations.They observed a greater formation of calcium urinary crystals in the KO mice compared to the controls spontaneously and after diet rich in oxalate and calcium.
No particular nephrological phenotype was noted by the two teams otherwise. The Bates team studied 10 KO mice and 10 matched control mice receiving a normal diet for 3 years. There was no difference in weight, growth or physical activity. The kidneys of mice sacrificed at 3 years showed no cystic or fibrous zones in any of the groups.
Physiopathology:
The protein of TH is therefore, in the state of our knowledge, the only protein involved in the genesis of the FJHN. The KO mouse models, which lead to the total absence of protein, have not made much progress in understanding the pathophysiology of the disease. It seems that it is more the existence of a pathological protein, localizing itself abnormally and / or not able to interact correctly with its usual partners, which is at the origin of the disease.
The vast majority of mutations observed in the uromodulin gene result in the replacement of a highly conserved cysteine with another amino acid. These mutations are thus capable of altering the formation of disulfide bridges and thus the tertiary structure of the protein. These structural abnormalities can result in abnormalities of function, degradation or localization of the protein. Some mutations do not affect these cysteines, but are located within the EGF-like domains and can also result in structural abnormalities of the protein. The study in immunohistochemistry by Dahan et al. of patients’ renal biopsies demonstrated an increase in TH protein expression at tubules with Henle’s loop markers. The labeling of the protein was, moreover, diffuse cytoplasmic and not apical as in the normal kidney.
The authors concluded that there was intracellular retention of the protein, then excreted in lower amounts in the urine.These data were subsequently confirmed by Rampoldi et al. which transfected four mutants of the protein into renal tubular cell cultures and showed that the mutations resulted in a delay in export of the protein to the plasma membrane by retention in the endoplasmic reticulum. These same authors have recently completed their study by transfection in a cell model of 12 different mutants, confirming anomalies in intracellular trafficking of mutated proteins, while anchoring to the cell membrane from the Golgi apparatus did not appear unnatural. The Jennings team recently studied the transfection of uromodulin mutants in a model of polarized kidney cells (LLC-PK1). They show profiles of cellular expression of similar mutated and wild proteins. They did not find an anomaly of secretion of the mutated protein at the basolateral pole of the cell, but a decrease of the secretion at the apical pole. They therefore conclude more with an anomaly of secretion than with addressing the apical membrane of the mutated protein. The authors found no difference in caspase 3 activity in the lysates of the different cell types, indicating the absence of increased apoptosis by the mutated protein in this cell model. The consequences of intracellular retention of the protein are not currently known and could cause tubular dysfunction or lead to an immune reaction responsible for interstitial nephritis.
The physiopathology of tubulo-interstitial nephropathy in the FJHN is therefore imperfectly known and probably not univocal.
There is at present no satisfactory explanation for the almost constant presence of hyperuricemia. Renal metabolism of urates is still imperfectly known, but the majority of renal urate reabsorption occurs at the proximal tubule upstream of the place of production and thus of the action of the TH protein. The only hypothesis is that water and sodium reabsorption anomalies at the level of Henle’s loop, leading to hypovolemia, would cause an increase in the absorption of urates in the proximal tubule. This remains unsatisfactory. It is unlikely that hyperuricemia is responsible for tubulo-interstitial nephropathy as evidenced by various genetic pathologies, leading to long-term hyperuricaemia not complicated by renal insufficiency.
Clinical presentation:
The age at diagnosis is extremely variable, between 3 and 67 years.
Hyperuricemia:
The lack of excretion of urates is one of the first abnormalities detected and is, for some, a good means of screening in the relatives of a patient in whom the diagnosis was made. Mac Bride et al. studied 34 apparently healthy children, but members of a family with FHHN. They found that 17 asymptomatic patients were hyperuricemic and of these, 42% did not yet have renal insufficiency. The occurrence of a gout attack, a fortiori in a young man or a woman, is often revealing of the disease. The age of the first gout attack is variable, but often occurs between the ages of 10 and 30 years. However, hyperuricemia can remain asymptomatic until the stage of terminal renal insufficiency.
In the various families reported in the literature, there is a clinical manifestation of gout in nearly two thirds of the cases. Hyperuricemia does not appear to be constant: within the large family described by Hart et al., 8% of patients have no hyperuricaemia, Hildebrandt et al. reported the case of a family with a mutation of MCKD-labeled uromodulin and no hyperuricaemia. More than the increase in uricemia, varying with the degree of renal insufficiency, it is often the decrease of the excretion fraction of uric acid (FE AU ) that is decisive. The FE AU corresponds to the clearance of urates relative to the clearance of creatinine. The standards used are 8.1 plus or minus 3.2% for men, 12.8 plus or minus 2.9% for women and 12-30% for children. During chronic renal failure, FE AU increases progressively (as soon as the glomerular filtration becomes less than 50 ml / min) to reach 85%. The points represent the data of patients with FHHN, illustrating the reduction of their FE FE.
Renal involvement:
Clinical features:
Renal involvement is far from specific. Chronic renal insufficiency has the characteristics of tubulo-interstitial nephropathy. Proteinuria is low or zero and, as a rule, less than 1 g per 24 hours. The urinary sediment is banal: a leucocyturia is incontestably found, there is no hematuria. There may be a urinary concentration disorder, usually subclinical, discovered during physiological tests. It is sometimes symptomatic, resulting in a polyuropolydipsic syndrome. The existence of an enuresis which can testify to a disorder of concentration of the urine has been reported by Hart et al. at a frequency of about 25%. Cesari found 70 to 100% of urinary concentration disorders in three families studied.
High blood pressure is often absent at the time of diagnosis, but may appear during the progression of renal failure and sometimes be severe. There is a frequency in the literature ranging from 35% to 70% of cases.
Several cases of pregnancy toxemia have been reported as the first manifestation of the disease.
Renal insufficiency appears most often during the second decade. It can sometimes be premature (6 years for the youngest patient reported). The evolution is that of a chronic interstitial nephropathy and takes place over 15-20 years in the vast majority of cases. The age at which patients reach terminal renal insufficiency is extremely variable, usually between 30 and 60 years. Dahan et al. described a Belgian family in which the age of dialysis ranged from 28 to 63 years, illustrating the high intra-familial variability of the progression of renal insufficiency.
Some authors have investigated the incidence of urinary tract infections in patients with FHHN. For Hart et al., They do not appear to be more frequent in men (no cases) or in women (some mild episodes in half of them) or more readily complicated with pyelonephritis. Wolf reports repeated urinary infections in a patient; nevertheless, it had a vesico-ureteral reflux.
Morphological aspect:
The renal morphological study is often reduced to the realization of a renal ultrasound, sometimes carried out in the stage of advanced renal insufficiency. The kidneys are of normal size or reduced according to the stage of the disease.They are sometimes referred to as hyperechoic. Thomson et al. reported in 1978 a family evocative of FJHN presenting to the anatomopathological analysis of the kidneys multiple medullary cysts 1 to 15 mm in diameter. Other authors have since reported the presence of single or multiple cysts, location and size in patients with FHHN.Ultrasound is often insufficient to diagnose them and only the most effective radiological techniques (scanning, magnetic resonance imaging (MRI)) can confirm or invalidate the existence of small cysts. They appear to be present in nearly half of the patients.
Pathological anatomy:
The histological study of the kidneys of patients with FJHN typically finds a chronic interstitial nephropathy with areas of focal tubular atrophy within a range of inflammatory interstitial fibrosis. There is thickening and often laminating of the tubular basal membranes predominant in distal circumferential tubules and collecting ducts, sometimes seats of cystic dilatation. Hart et al. have described the existence of acellular hyaline dense material embedding certain tubes.Zager et al. have described similar hyaline deposits in patients with MCKD and immunofluorescence studies have shown that it is a TH protein. However, this anomaly is not reported by the other authors. Dahan et al. , who studied in immunohistochemistry the kidneys of patients with uromodulin mutation, found excessive expression, as well as an abnormally cytoplasmic localization of the TH protein. They do not notice any signal within the interstitium or around the tubes. Rampoldi et al. have also demonstrated the formation of cytoplasmic globular masses on the kidneys of patients affected by electron microscopy corresponding to the accumulation of TH protein within the endoplasmic reticulum.
Hodanova et al., Studying a family with FHHN unrelated to a mutation of uromodulin, but at the 1q41 locus, demonstrated, conversely, a reduction in the expression of the TH protein. The glomeruli have no particular abnormalities. The existence of a particularly marked arteriolar hyalinosis is sometimes noted.
There are no immunoglobulin deposits in immunofluorescence.
Dosage of uromodulin:
Several authors have measured the rate of urinary excretion of uromodulin in patients with a mutation of the gene.This is constantly decreased, often found at a value less than 50% of that expected. The excreted protein is normal.Jennings et al. measured plasma uromodulin: patients with decreased urinary uromodulin had either a high or a lowered dose of plasma uromodulin. A young patient with normal renal function and urinary uromodulin nevertheless had a very high uromodulin plasma level. No conclusions can be drawn from these observations. Recently, Vylet’al et al. studied about fifty patients with a FHHN array and showed a decrease in urinary uromodulin in most cases while only 14 had a mutation of the gene.
Extraneous:
Some authors have noted a higher incidence of hyperparathyroidism, requiring surgical management in these patients.The limitation of the secretion of calcitriol by hyperuricaemia would be the explanation for some.
Austrian authors reported a family with FHHN in which two patients also had severe cardiac malformations (congenital stenosis of the pulmonary artery and aortic stenosis). This association was probably fortuitous.
Differential Diagnosis:
Certain genetic diseases can give clinical pictures close to the FHHN and must be evoked.
Mutations in the HNF1 b gene are responsible for an autosomal dominant disease with a variety of clinical manifestations. They are known to give MODY type diabetes; however, it is not always the first manifestation of the disease.
Cystic dysplasia and tubulo-interstitial nephropathy with glomerular cysts are part of the kidney damage described.
Hyperuricemia is also frequently found in this pathology, which is the main differential diagnosis.
Mitochondrial cytopathies by mitochondrial DNA mutation A3243G may cause renal tubulo-interstitial involvement and sometimes cysts. Extrarenal damage can often guide the diagnosis.
Lead poisoning may be responsible for chronic tubulo-interstitial nephropathy with gout. Family exposure could mimic a hereditary disease.
Treatment:
The treatment of FHHN is essentially symptomatic.
The correction of high blood pressure when it exists is paramount. The prescription of antiproteinuric treatment in patients in whom proteinuria exceeds 0.5 g / 24 hours seems justified.
The treatment of hyperuricemia with allopurinol is not discussed when gout manifestations are present, but its efficacy in slowing the progression of nephropathy is not yet certain. There are no prospective controlled studies due to the scarcity of the disease and the difficulty of constituting a control group.
Several retrospective studies have investigated the effect of allopurinol on the development of nephropathy.
Some regain stabilization of renal function, others no effect. Mc Bride et al. are convinced of the efficacy of early treatment with allopurinol. Faibanks et al. report their experience in a retrospective long-term follow-up study in 2002.They compare their treated patients to the evolution of previous generations or untreated relatives. They stratify patients according to the degree of renal insufficiency at the start of treatment.
In patients with serum creatinine greater than 200 μmol / l, progression of nephropathy was rapid despite treatment. In patients with serum creatinine less than 200 μmol / l, renal function remained relatively
stable for up to 34 years. The authors therefore insist on the necessity of the early introduction of the treatment before the installation of the renal insufficiency. Other molecules may also be effective in the future.
Choi et al. suggest, on the basis of in vitro studies, an efficacy of colchicine on the progression of renal fibrosis.
These authors transfected wild-type or mutated uromodulin cDNA into immortalized cells of Henle’s Loop. They studied the effect of colchicine, as well as sodium 4-phenylbutyrate (chaperone molecules to redirect proteins abnormally addressed to the endoplasmic reticulum) and showed an increase in the secretion of TH protein from the reticulum to the cell membrane and the culture medium and an improvement in cell viability.
The benefit of long-term colchicine therapy has not yet been tested in patients.
The management of complications of chronic renal insufficiency is not different in patients with FHHN.
A very small number of renal transplants in patients with FHHN have been reported. In the seven patients reported by the Cameron team, no recurrence of hyperuricemia or gout in the three patients with a functioning graft at 10 years was noted. The Pirson team also reported the absence of recurrence of hyperuricemia in five transplant patients and normalization of FE AU in a patient with normal renal function. A patient whose renal function had deteriorated underwent a biopsy 5 years after the transplant, no sign of recurrence of the initial nephropathy was found.
