|Year : 1997 | Volume
| Issue : 3 | Page : 227-234
|Hereditary Renal Diseases
Hopital Necker Paris, France
Click here for correspondence address and email
|How to cite this article:|
Grunfeld JP. Hereditary Renal Diseases. Saudi J Kidney Dis Transpl 1997;8:227-34
| Epidemiology|| |
The great diversity in hereditary renal diseases is illustrated in [Table - 1],[Table - 2],[Table - 3],[Table - 4]. This paper mainly focuses on the most prevalent of these diseases in adults and on those in which recent advances are noteworthy. Progress in molecular genetics has greatly contributed to our understanding of the mechanisms of these diseases, as well as to that of more general mechanisms of disease.
Autosomal dominant polycystic kidney disease (ADPKD) is by far the most prevalent hereditary renal disease and one of the most prevalent inherited diseases in humans. Its prevalence ranges from 1 in 400 to 1 in 1000 individuals in white populations. The corresponding figure in blacks is not yet available. In the past decades, ADPKD accounted for 8 to 10 percent of all patients treated by renal replacement therapy in Western countries. In the 90s it represents only 5 to 7 percent of these because of the increasing number of elderly patients with diabetic and/or vascular renal diseases reaching end-stage.
Alport's syndrome is the second most prevalent inherited kidney disease (approximately 1 in 5000 individuals). It should be clearly understood that the incidence of patients reaching end-stage renal failure (ESRF) does not adequately reflect the true incidence of the disease, since many affected heterozygous females in X-linked Alport's syndrome will never progress to ESRF. This is also true in a substantial fraction of ADPKD patients (see below).
Many hereditary renal diseases manifest only late in adult life. This is a common occurrence in ADPKD. In 90 percent of the male patients with Alport's syndrome, ESRF is reached in adulthood, in some cases after 50 or 60 years of age. Inherited disorders are not restricted to children, and all physicians, not only pediatricians, are confronted with hereditary renal disorders. Of course, the most severe and invalidating diseases are revealed in childhood. However, those that are the most prevalent, such as non-insulin dependent diabetes, Huntington's disease, Alzheimer's disease, genetic amyloidosis and so on have their first manifestations in adulthood.
| Cystic Kidney Diseases|| |
The spectrum of genetic cystic kidney diseases is illustrated in [Table - 1].
| Autosomal Dominant Polycystic Kidney Disease|| |
The disease is genetically heterogeneous. Indeed, a first locus, PKD1, had been located in 1985 on the short arm of chromosome 16. The gene was identified nine years later thanks to the study of a Portuguese family with both ADPKD and tuberous sclerosis, and with a chromosomal translocation 22;16. Of great interest, the breakpoint of the translocation on chromosome 16p was within the PKD1 gene. In addition, the PKD1 gene was shown to be contiguous to the TSC2 gene, one of the genes involved in tuberous sclerosis  . Subsequently, a few families were described in which both PKD1 and TSC2 genes were deleted, giving rise to very early and severe polycystic kidney disease  .
The PKD1 disease is the most prevalent, accounting for approximately 85 percent of the cases in European countries. The protein encoded by the PKD1 gene has been called polycystin. It is a membrane protein with a large extracellular adhesive component, a series of 13 membrane-spanning domains, and, at the C-terminus, a cytoplasmic tail. This putative structure suggests that this protein interacts with other cell membrane or extracellular matrix molecules. The signal normally conveyed from the polycystin legends in the extracellular space to the internal milieu of the cell is disrupted by the PKD1 mutations. This probably leads to abnormal differentiation of tubular cells and finally to cyst formation.
A second locus had been localized on the long arm of chromosome 4, and the PKD2 gene was just identified in 1996  . There are some ADPKD families unlinked to either the PKD1 or the PKD2 locus. Thus, there must be at least one additional PKD gene. The PKD2 protein is a membrane protein but its extracellular domain is more restricted than that of polycystin. Its structure may be compatible with an ion channel.
The renal manifestations of ADPKD (pain, bleeding, infection, stone) have been well known for decades. The incidence of renal cell carcinoma is not increased in ADPKD. I will concentrate on renal failure and its progression, and on hypertension.
Progressive renal failure is the most serious and most frequent complication of ADPKD. Recent epidemiological data, however, have indicated that the probability of being alive without ESRF ranges from 23 to 50 per cent for patients with ADPKD. When renal failure occurs, end-stage is reached at a mean age of 55 years. Regular dialysis is initiated between 40 and 59 years of age in 75 per cent of the patients. When renal failure starts, the mean loss of glomerular filtration rate is 6 ml/min per year. There is no clear intrafamilial concordance with regard to age at ESRF, and no clear evidence for anticipation.
In very rare instances, ADPKD may be responsible for ESRF in young children. This may be ascribed to large deletions disrupting both PKD1 and TSC2 genes as indicated above, and these children are also affected by tuberous sclerosis. In the other families (usually PKD1), the reason for early progression is unknown. Possible risk factors are a parental new mutation (which is a very rare event in ADPKD) and maternal inheritance of the disease. The risk of recurrence in siblings is high.
The determinants of progression are both genetic and non-genetic. The rate of cyst development and progression of renal failure is slower in non-PKDl families (mostly PKD2) than in PKD1 families and ESRF occurs 15 years later than in PKD1 families (70 vs 55 years). Male sex affects negatively the renal prognosis; the rate of progression of renal failure is slower in females than in males, and end-stage is reached approximately six years later in women. Black patients, and more specifically those with associated sickle cell disease, have a more rapid progression than whites, but additional information is needed on this point.
Hypertension is a common and early complication of ADPKD. It is found in at least 33 per cent of the patients with normal renal function. Young people with ADPKD have slightly higher blood pressure and left ventricular mass than their unaffected relatives. In recent studies, no strong correlation between mean blood pressure and the rate of progression of renal failure has been demonstrated in ADPKD patients, most of them receiving anti-hypertensive therapy. This does not mean that earlier intervention would not have been beneficial and, of course, does not detract from the need to control blood pressure to prevent cardiovascular complications.
Pregnancy does not seem to alter significantly the course of ADPKD. Not surprisingly, hypertension and preeclampsia are more frequent than in the general population.
Among external manifestations, I will focus on liver involvement and intracranial aneurysm. Liver cysts develop later than renal cysts; they are not found before 20 years of age and their prevalence is approximately 80 per cent after 50 to 60 years of age. They are more prevalent, are observed earlier, and are more numerous and extensive in females than in males. Liver cysts are usually asymptomatic. Massive polycystic liver may lead to compression of the gastrointestinal tract, abdominal pain and discomfort. Posterior cysts may also compress the suprahepatic veins and intrahepatic vena cava, leading to hepatic vein outflow obstruction and resulting in portal hypertension with ascites. The management of these patients is not easy; therapeutic means include percutaneous puncture and sclerosis with alcohol or minocycline, laparascopic fenestration, hepatic resection, and even liver transplantation.
The rupture of intracranial aneurysm is one of the most severe complications in ADPKD patients. The prevalence of asymptomatic intracranial aneurysm has been found to be approximately 8 per cent in ADPKD patients, i.e., five times higher than in the general population. The frequency of rupture is also five times higher than in the general population. The mean age at rupture is 41 years, and 10 per cent of the patients were under 21 years of age when this accident occurred. Rupture of intracranial aneurysm is preceded in 20 to 40 per cent of the cases by-premonitory headaches, of sudden onset, of unusual character, often posterior, and results in subarachnoid hemorrhage. These patients should be managed in a neurosurgical unit; non-enhanced CT scanning is the first line diagnostic procedure. In the future, initial work-up will probably include contrastenhanced spiral CT which often provides additional information on the site of the ruptured aneurysm  .
Intracranial aneurysm rupture is a severe complication; it entails a 35 to 55 per cent risk of combined mortality and serious sequelae. Is it therefore worthwhile to screen asymptomatic ADPKD patients for intracranial aneurysm? This has been debated in recent years by using decision analysis. In short, screening is advised in patients who have a positive family history of intracranial aneurysm (because the risk of aneurysm is 2.6 times greater in these patients than in those without a definite family history). Screening has also been proposed in patients who undergo major elective surgery or in those with high-risk occupations. The best screening procedures are magnetic resonance-angiography and sprial CT scan. Screening should be repeated every five years (?) in high-risk patients and in those who had a previous accident of rupture since the cerebrovascular disease is progressive.
Other vascular changes have been reported in ADPKD patients, including mitral valve prolapse, the incidence of which has been claimed to be 10 times higher than in the general population.
| Autosomal Recessive PKD|| |
In this disease, renal cysts develop from collecting ducts (whereas they arise from any tubular segment in ADPKD), and renal involvement is associated with congential hepatic fibrosis, either asymptomatic or responsible for portal hypertension, and/or bacterial angiocholitis if dilatations of intrahepatic ducts are also found. This disease is usually symptomatic in childhood. However, in rare cases, ESRF may be reached after 20 years of age.
The gene, encoding for autosomal recessive PKD has been located on chromosome 6  .
| Juvenile Nephronophthisis-Medullary Cystic Disease Complex|| |
Juvenile nephronophthisis, along with cystinosis, is a major inherited cause of ESRF in children. Renal involvement is characterized by diffuse interstitial fibrosis with thickened and multilayered tubular basement membranes. Medullary cysts are detected late in the course. It may be associated with retinal involvement (Senior-Loken syndrome). It is an autosomal recessive disease whose gene has been located on chromosome 2. A homozygous deletion has been identified in affected subjects. This is also found in sporadic cases , .
A similar disease with prominent renal tubulointerstitial changes has been described in adult patients. The renal cysts predominating at the corticomedullary junction or the medulla can be detected earlier in the course, accounting for the denomination "medullary cystic disease". Renal failure progresses to end-stage at about 35 years of age. In fact, the disease differs in adults: (i) no extrarenal manifestation occurs and (ii) inheritance is autosomal dominant. The gene has not been localized so far.
| Renal Cystic Involvement in Phakomatoses|| |
Multiple renal cysts, sometimes mimicking ADPKD, may be found in patients with phakomatoses, namely tuberous sclerosis and von Hippel-Lindau (VHL) disease. However, other renal and extrarenal features are diagnostic.
In tuberous sclerosis (TSC), renal cysts are often associated with multiple bilateral angiomyolipomas. These renal lesions may lead to ESRF  . These patients are exposed to the risk of bleeding due to highly vascularized angiomyolipomas, and of developing renal cell carcinoma. Binephrectomy has therefore been recommended in end-stage patients. Two TSC genes have been identified: TSC1 on chromosome 9, and TSC2 on chromosome 16. In both forms of TSC, renal cysts may be present.
Renal involvement is more prominent in VHL disease. In addition to cysts, renal cell carcinoma, often multifocal and bilateral, develops in 60 to 70 per cent of the patients. This is presently the main cause of death in this disease. Nephron-sparing surgery is recommended but exposes to the risk of recurrence of further development of carcinoma. When the renal lesions are too diffuse or have recurred after conservative surgery, bilateral nephrectomy and renal replacement therapy are the only choice  .
The VHL gene has been localized on the short arm of chromosome 3, and cloned. Like the TSC2 gene, it is a tumour suppressor gene. The development of carcinoma is triggered by two events (the "two-hit" hypothesis of Knudson), the first being germinal mutation and the second, somatic (in the tumor). Of great interest, (somatic) mutations of the VHL gene have been found in at least 80 per cent of sporadic renal cell carcinomas. It has been recently shown that renal carcinogenesis is initiated by the lack of interaction between the mutated VHL protein and the elongin system which regulates transcription. This exemplifies how the information drawn from inherited disorders may lead to understanding much more general mechanisms of disease , .
| Alport's Syndrome|| |
Alport's syndrome is an inherited renal disorder characterized by familial occurrence, in successive generations, of progessive hematuric nephritis with ultrastructural changes of the glomerular basement membrane (GBM) and sensorineural hearing loss  . In the family he reported in 1927, Dr. A. Cecil Alport clearly noted that "males develop nephritis and deafness and do not as a rule survive" whereas "females have deafness and hematuria, and live to old age". This statement has since been largely confirmed: X-linked dominant inheritance has been shown to be the most frequent in this disorder (85 per cent of the families).
In recent years, knowledge on the genetics and pathogenesis of Alport's syndrome has progressed dramatically. It has been found successively that the primary lesion involves the GBM, and more specifically type IV collagen (which is a major component of basement membranes), and that the primary biochemical defects involve novel alpha chains of type IV collagen (whose distribution is restricted to certain tissues, including the kidney, ear and eye, and whose genes have been located, cloned and found mutated in the patients). Thus, in two decades, we have moved from a "syndrome" to different "diseases" whose clinical, biochemical and molecular features have in part been identified [Table - 2].
Classical X-lined dominant Alport's syndrome could be designated more adequately as X-linked dominant alpha 5 (IV) disease. Indeed, it is characterized by mutations involving COL4A5, the gene encoding for the oc5 chain of type IV collagen. Deletions are detected in 5 to 16 per cent of the kinderds. They are usually associated with juvenile-type Alport's syndrome in which all affected males progess to ESRF at 30 years of age or earlier. In the other kindreds, small mutations are found. De novo mutations are not rare.
X-linked Alport's syndrome associated with leiomyomatosis represents a separate entity. In addition to leiomyomatosis affecting the oesophagus, respiratory tract or genital tract in females, this entity includes lens opacities (whereas the eye abnormalities in classical Alport's syndrome are mainly bilateral anterior lenticonus and perimacular retinal changes). This entity is characterized by large deletions involving both COL4A5 and COL4A6, genes located contiguously on the long arm of the X chromosome.
The autosomal recessive form of Alport's syndrome is characterized by the following features: equally severe disease in males and females, or more specifically, severe renal disease in females leading to ESRF before 20 years of age, and appearance of the disease in the family after a consanguineous marriage. Molecular studies have clearly established autosomal recessive inheritance by showing mutations involving both copies of COL4A3 or COL4A4 genes in homozygous patients. Both genes are located contiguously on the long arm of chromosome 2.
In the autosomal dominant form(s) of Alport's syndrome, with or without macrothrombocytopenia, the molecular defect has not so far been identified.
The term thin-basement-membrane nephropathy has been coined to describe diffuse thinning of the GBM, associated with urinary abnomarlities. This entity is either sporadic or inherited. In fact, it does not correspond to a single clinical syndrome: a) Thin GBM can be seen in various renal diseases, including IgA nephropathy; b) Whereas thickened GBM with split lamina densa is the most characteristic ultrastructural lesion in Alport's syndrome, thin GBM can be found in typical families, more often in children; c) Patients with benign familial hematuria usually have thin GBM. In contrast, patients with thin GBM may present with various abnormalities and some of them may have abundant proteinuria and may progress to ESRF. Thus, thin GBM is an ultrastructural lesion, and does not correspond to a single disease. It should not be considered as the histopathological coun terpart of benign familial hematuria. Its prognosis is uncertain (except when the criteria for benign familial hematuria are met, not only in females but also in males), and careful follow-up is required.
The advances in Alport's syndrome since 1990 have been astounding. They have been decisive for genetic counseling to the families. The risk of transmission to the offspring is obviously different in X-linked and autosomal recessive families. Molecular genetics is very helpful in X-linked kindreds for identifying heterozygous carrier females who may have intermittent or no urinary abnormalities. It also allows identification of non-carrier females in affected families which is an often neglected, but essential consequence of genetic counseling.
| Inherited Metabolic or Non-Metabolic Diseases with Kidney Involvement|| |
I will not review these diseases in detail. I will only stress that the renal involvement manifests often late in life, in adulthood, and can reveal the hitherto undiagnosed disease [Table - 3].
Renal involvement is one of the cardinal features of Bardet-Biedl syndrome, associated with obesity, pigmentary retinitis, hypogonadism and polydactyly. It is generally apparent only in adults. In inherited urate nephropathy (or nephropathy with hyperuricemia), the renal symptoms appear only in adult patients. In other metabolic diseases, such as Fabry's disease, renal abnormalities develop in affected hemizygous adults. The diagnosis of the disease may have been established earlier in life, but it is not uncommon that this disease remains undiagnosed until the third decade of life. The diagnosis is much more difficult in heterozygous carrier females, in the absence of a positive family history. These women may, however, develop slight renal involvement after 60 or 70 years of age  . In type I glycogen storage disease (von Gierke's disease), symptomatic management allows survival of children until adulthood. We have therefore learned that a renal disease may develop progressively in these adult patients and is characterised by focal and segmental glomerulosclerosis.
The inherited forms of nephrolithiasis have been identified more precisely. The molecular defects responsible for cystinuria, Dent's disease or X-linked recessive nephrolithiasis, type I primary hyperoxaluria, and adeninephospnoribosyl transferase (ADPRT) deficiency (with 2,8 dihydroxyadenine stone formation) have been elucidated, at least in part. Whereas primary hyperoxaluria may present very early in life with recurrent urinary stones and nephrocalcinosis, it may-become symptomatic only in adulthood. The ADPRT deficiency may rarely lead to intrarenal precipitaion of 2,8 dihydroxyadenine crystals and to progressive renal failure. This has been reported in adults, including a 72-year-old woman.
These examples have been selected to demonstrate that the renal disease complicating inherited, metabolic or not, disorders may appear late in life and the diagnosis may be occasionally made only in elderly subjects. All physicians are exposed to hereditary diseases, even nephrologists.
| Inherited Tubular Disorders|| |
Some of these disorders are complicated by nephrolithiasis. Cystinuria due to a defect in dibasic acid tubular reabsorption, including cystine, leads to cystine stone formation. One of the genes involved in this autosomal recessive disorder has been identified, encoding for one of cystine transport systems in the proximal tubule.
Dent's disease and X-linked calcium nephrolithiasis have been described independently but are closely related, sharing a common gene defect involving a chloride channel CLCN5. It remains to be understood why this defect leads to calcium nephrolithiasis or nephrocalcinosis. Is it also a more general mechanism involved in some way in other forms of calcium nephrolithiasis? Dent's disease and X-linked nephrolithiasis are also characterized by proximal tubular dysfunction and progression to renal failure  .
The molecular basis of three inherited defects of sodium tubular reabsorption have been recently clarified. Autosomal recessive neonatal Bartter's syndrome is due to inactivating mutations affecting the gene encoding for the burnetanide/furosemide sensitive Na-K-2 C\ co-transporter, located at the apical membrane of the thick ascending limb of Henle's loop  . The manifestations of Bartter's syndrome comprise failure to thrive, growth and mental retardation, metabolic alkalosis and hypokalemia due to renal potassium wasting. Urine calcium excretion is usually high, accounting for the risk of nephrocalcinosis. Gitelman's syndrome has long been considered as a variant of Bartter's syndrome. Hypokalemia due to renal wasting is similar, but associated features are found; late onset after 8-10 years of age, often in adulthood, normal development, hypomagnesemia, sometimes chondrocalcino-sis, and low calcium excretion. In the autosomal recessive form of Gitelman's syndrome, homozygous inhibitory mutations of the gene encoding for the Na-Cl co-transporter, inhibited by thiazide, are found; this co-transporter is located at the apical membrane of the distal tubule  . This explains why this syndrome encompasses abnormalities reminiscent of long-term thiazide administration (hypokalemia, hypomagnesemia and hypocalciuria). Of great interest, in both Bartter's and Gitelman's syndromes, the blood pressure is low as occurs in chronic administration of diuretics. In contrast, activating mutations of these genes may induce high blood pressure. Is this the mechanism of some forms of salt-sensitive hypertension, which are in turn corrected by administration of diuretics?
Liddle's syndrome is due to activating mutations of the gene coding for β or γ subunits of the epithelial sodium channel, sensitive to amiloride, located at the apical membrane of the collecting duct cells. These mutations are therefore responsible for increased tubular reabsorption of sodium, positive sodium balance and hypertension (exquisitely sensitive to low salt diet and amiloride or triametrene administration). This syndrome was first described in 1964 by G. Liddle, and its hereditary autosomal dominant nature had been recognized. Early and severe hypertension, responsible for fatal cerebral hemorrhage in some patients, is accompanied by hypokalemia (inconstantly), due to renal wasting, and low renin and aldosterone. Three decades later, it was shown that kidney transplantation, made necessary by hypertensive nephrosclerosis, cured hypertension and hypokalemia, showing that the primary defect was intrinsic to the kidney. The amiloride-sensitive epithelial sodium channel might also be involved in other forms of hypertension. The information collected in rare inherited diseases may be applied to other domains of molecular medicine , .
| References|| |
|1.||Pirson Y, Chauveau D, Grunfeld JF. Autosomal-dominant polycystic kidney disease, in Oxford Textbook of Clinical Nephrology, A.M. Davison et al. editors, 2nd Ed., 1997, Oxford University Press: (in press). |
|2.||The polycystic kidney disease 1 gene encodes a 14 kb transcript and lies within a duplicated region on chromosone 16. The European Polycystic Kidney Disease Consortium. Cell 1994;77:881-94. |
|3.||Brook-Carter PT, Peral B, Ward CJ, et al. Deletion of the TSC2 and PKD1 genes associated with severe infantile polycystic kidney disease: A contiguous gene syndrome. Nat Genet 1994;8:328- 32. [PUBMED] [FULLTEXT]|
|4.||Mochizuki T, Wu G, Hayashi T, et al. PKD2, a gene for polycystic kidney disease that encodes an integral membrane protein. Science 1996;272:1339- 42. [PUBMED] [FULLTEXT]|
|5.||Pirson Y, Chauveau D, van Gijn J. Subarachnoid haemorrhage in ADPKD patients: how to recognize and how to manage? Nephrol Dial Transplant 1996;ll:1236-8. |
|6.||Zerres K, Rudnik-Schoncborn S, Mucher G. Autosomal recessive polycystic kidney disease: Clinical features and genetics. Adv Nephrol 1996;25:147-57. |
|7.||Konrad M, Saunier S, Heidet L, et al, Large homozygous deletions of the 2ql3 region are a major cuase of juvenile nephronophthisis. Hum Mol Genet 1996;5:367-71. |
|8.||Antignac C, Arduy CH, Beckman JS, et al. A gene for familial juvenile nephronophthisis (recessive medullary cystic kidney disease) maps to chromosome 2p. Nat Genet 1993;3:342-45. |
|9.||Schillinger F, Montagnac R. Chronic renal failure and its treatment in tuberous sclerosis. Nephrol Dial Transplant 1996;ll:481-5. |
|10.||Chauveau D, Duvic C, Chretien Y, et al. Renal involvement in von HippelLindau disease. Kidney Int 1996;50:944-51. |
|11.||Gnarra JR., Duan DR, Weng Y, et al. Molecular cloning of the von HippelLindau tumor suppressor gene and its role in renal carcinoma. Biochim Biophys Acta 1996;1242:201-10. |
|12.||Linehan WM, Lerman MI, Zbar B. Identification of the von Hippel-Lindau (VHL) gene. Its role in renal cancer. JAMA 1995;273:564-70. |
|13.||Grunfeld JP, Knebelmann B, Alport's syndrome, in Oxford Textbook of Clinical Nephrology, A.M. Davison, et al. Editors, 2nd Ed., 1997, Oxford University Press. Oxford (in press). |
|14.||Hillsley RE, Hernandez E, Steenbergen C, Bashore TM, Harrison JK. Inherited restrictive cardiomyopathy in a 74year-old woman: a case of Fabry's disease. Am Heart J 1995;129:199-202. [PUBMED] [FULLTEXT]|
|15.||Lloyd SE, Pearce SH, Fisher SE, et al. A commom molecular basis for three inherited kidney stone diseases (see comments). Nature 1996;379:445-9. [PUBMED] [FULLTEXT]|
|16.||Simon DB, Karet FE, Hamdan JM, DiPietro A, Sanjad SA, Lifton RP. Bartter's syndrome, hypokalaemic alkalosis with hypercaJciuria, is caused by mutations in the Na-K-2C1 cotransporter NKCC2. Nat Genet 1996; 13: 1S3-8. |
|17.||Simon DB, Nelson-Williams C, Bia MJ, et al. Gitelman's variant of Bartter's syndrome, inherited hypokalaemic alkalosis, is caused by mutations in the thiazide-sensitive Na-Cl cotransporter. Nat Genet 1996;12:24-30. [PUBMED] [FULLTEXT]|
|18.||Shimkets RA, Warnock DG, Bositis CM, et al. Liddle's syndrome: heritable human hypertension caused by mutations in the beta subunit of the epithelial sodium channel. Cell 1994;79:407-14. [PUBMED] |
|19.||Rossier BC. The renal epithelial sodium channel: new insights in understanding hypertension. Adv Nephrol 1996;25:275-86. |
Universite Rene Descartes – Paris V, Clinique Jean Hamburger, Hopital Necker Paris
[Table - 1], [Table - 2], [Table - 3], [Table - 4]
| Article Access Statistics|
| Viewed||4895 |
| Printed||53 |
| Emailed||0 |
| PDF Downloaded||417 |
| Comments ||[Add] |