Cystinuria

A number sign (#) is used with this entry because of evidence that cystinuria can be caused by mutation in the SLC3A1 amino acid transporter gene (104614), which encodes the heavy subunit of the renal amino acid transporter and is located on chromosome 2p, and/or by mutation in the SLC7A9 gene (604144), which encodes the light subunit and is located on chromosome 19. A classification scheme of cystinuria based on the molecular genetics of the disorder has been proposed (see NOMENCLATURE).

Description

Cystinuria is an autosomal disorder characterized by impaired epithelial cell transport of cystine and dibasic amino acids (lysine, ornithine, and arginine) in the proximal renal tubule and gastrointestinal tract. The impaired renal reabsorption of cystine and its low solubility causes the formation of calculi in the urinary tract, resulting in obstructive uropathy, pyelonephritis, and, rarely, renal failure (summary by Barbosa et al., 2012).

Nomenclature

Rosenberg et al. (1966) described 3 types of cystinuria based on excretion patterns of presumed heterozygotes, e.g., the parents and children of affected individuals, with type I heterozygotes showing normal amino aciduria, and type II and type III heterozygotes showing high and moderate hyperexcretion of cystine and dibasic amino acids, respectively. In contrast to types I and II homozygotes, type III homozygotes showed a nearly normal increase in cystine plasma levels after oral cystine administration.

Dello Strologo et al. (2002) pointed out that the traditional classification system of cystinuria patients, based on excretion of cystine and dibasic amino acids in obligate heterozygotes, was not supported by evidence (see later) that all 3 types of the disease are caused by mutation in only 2 genes, SLC3A1 and SLC7A9. Dello Strologo et al. (2002) proposed a classification system based on the molecular genetics of the disorder: type A, due to mutation in the SLC3A1 gene; type B, due to mutation in the SLC7A9 gene; and type AB, due to a mutation in both the SLC3A1 and SLC7A9 genes, respectively.

Clinical Features

Wollaston (1810) first described a cystine stone. He found that a glistening yellow bladder stone was composed of an unusual substance, which he called cystic oxide since it came from the bladder. Later analysis showed this to be a sulfur-containing amino acid and so this stone ultimately gave its name not only to cystinuria but also to the amino acids cystine and cysteine. Marcet (1817) showed that cystine stones occur also in the kidney. He suspected that the condition might be familial since 2 of his patients were brothers. Cystinuria was one of the 4 inborn errors of metabolism discussed by Garrod (1908).

Rosenberg et al. (1966) described 3 forms of cystinuria, each due to presumed homozygosity of a particular mutant allele at 1 locus. In cystinuria I, the homozygote excretes relatively large amounts of cystine, lysine, arginine and ornithine in the urine. Heterozygotes (e.g., parents) have no abnormal amino aciduria. Urinary stones form in all 3 types of cystinuria because of the limited solubility of this amino acid. Cystinuria II is incompletely recessive because heterozygotes have a moderate degree of amino aciduria, mainly cystine and lysine, and may occasionally form cystine stones. Observations in kindreds in which both cystinuria I and cystinuria II are segregating demonstrate that the genes for these are allelic (Hershko et al., 1965). In cystinuria III, intestinal transport of all dibasic amino acids is retained by heterozygotes, and homozygotes excrete cystine in slight excess. Rosenberg (1966) and others observed families in which 'doubly heterozygous' persons (I-II, I-III, or II-III) had full-blown cystinuria. The findings were best explained on the basis of allelism of the genes responsible for the 3 types.

Brodehl et al. (1967) reported a 2-year-old girl who was discovered to have isolated hypercystinuria with normal urine levels of arginine, lysine, and ornithine during an evaluation for candidiasis. A younger brother had the same pattern of urinary amino acid excretion. Their unrelated parents and an older sister had normal cystine excretion. The female proband was also noted to have isolated hyperparathyroidism, which was suspected to be familial because another sister and brother had died due to hypocalcemic tetany.

Scriver et al. (1970) presented evidence indicating that cystinuria patients are at increased risk for impaired cerebral function. Weinberger et al. (1974) demonstrated an unusually high frequency of type II or III cystinuria among Libyan Jews.

Kelly (1978) concluded that the excretion rates of obligate carriers among the relatives of cystinurics suffice to determine the type of cystinuria in the proband. Among 17 patients he studied, type I was the most frequent type and often occurred in compound heterozygotes with type III. When obligatory heterozygotes showed normal amounts of cystine and dibasic amino acids in the urine, they were called type I. When up to twice the normal range was excreted in the urine, they were called type III. When carriers excreted large amounts of cystine and lysine (9-15 times the normal range but less than in most stone-formers), they were called type II. On the basis of a study in Brazil, Giugliani et al. (1985) concluded that there is an increased frequency of heterozygotes for types II and III cystinuria among urinary stone-formers and that heterozygosity for these genes is a risk factor for urinary stones.

Inheritance

The inheritance pattern of cystinuria is complex. Some patients show classic autosomal recessive inheritance. However, the urinary excretion of cystine and dibasic amino acids can vary considerably among heterozygotes, and may result in nephrolithiasis. Heterozygotes have been classified biochemically into type I, in which there is a normal pattern of amino acid excretion (consistent with autosomal recessive inheritance), and non-type I, in which there is urinary hyperexcretion of cystine, sometimes resulting in stone formation. Manifesting heterozygotes suggests that the disease can also be transmitted in an autosomal dominant pattern with incomplete penetrance (summary by Barbosa et al., 2012).

Clinical Management

On the basis of the extensive experience at St. Bartholomew's Hospital in London, Stephens (1989) indicated that in many people cystine stones can be dissolved and new ones prevented by a high fluid intake; that in those in whom this measure does not succeed, regular treatment with penicillamine will be effective; that side effects of penicillamine are rarely severe enough to prevent its use; and that because of the effectiveness of treatment, early diagnosis is important with consideration of cystinuria in all persons, regardless of age, who form urinary stones.

Mapping

In a mentally handicapped 3-year-old child with cystinuria, Sharland et al. (1992) found an apparently balanced de novo translocation with breakpoints at 14q22 and 20p13. Family studies suggested that the child was a type I/type II compound heterozygote for cystinuria. Sharland et al. (1992) suggested that the cystinuria locus might be on either 14q22 or 20p13.

Pras et al. (1994) showed linkage between a panel of 17 cystinuria families (8 of which were of Libyan Jewish origin) and markers on the short arm of chromosome 2. They had directed their attention to this region of the genome because the SLC3A1 amino acid transporter gene, whose protein product is involved in cystine, dibasic and neutral amino acid transport, had been mapped to that site. However, refined linkage studies limited to the Libyan Jewish families by Wartenfeld et al. (1997) excluded the disease locus from the region of the SLC3A1 gene. Pairwise linkage analysis revealed a maximum lod of 9.22 at theta = 0.00 with the marker D19S882. Further analysis placed the gene in an 8-cM interval between D19S409 and D19S208. Quantitative urinary amino acid analysis in these families demonstrated non-type I disease; final determination as to whether these families had type II or type III remained to be determined by the findings of oral loading tests.

The 3 types of cystinuria were thought to be due to allelism of the SLC3A1 gene, although the possibility of 2 distinct loci for type I and type III cystinuria had been suggested by Goodyer et al. (1993) and others. Calonge et al. (1995) performed linkage analysis in 22 families with type I and/or type III cystinuria and found that type I/I families showed homogeneous linkage to SLC3A1, whereas types I/III and III/III were not linked. Calonge et al. (1995) concluded that type III cystinuria results from mutation in a gene other than SLC3A1.

By linkage analysis after a genomewide search, Bisceglia et al. (1997) mapped the CSNU3 locus to 19q13.1. Pairwise linkage analysis in a series of type III or type II families previously excluded from linkage to the CSNU1 locus, i.e., the SLC3A1 gene, revealed a maximum lod score of 13.11 at a recombination fraction of 0.0 with marker D19S225. Preliminary data on type II families seemed to place the disease locus for this rare type of cystinuria at 19q13.1 also.

Stoller et al. (1999) studied a kindred of 39 persons with cystinuria, in which one branch demonstrated type II cystinuria and the other had type III disease. Linkage analysis demonstrated linkage of both types to the 19q13.1 region. Two individuals in the pedigree were shown by haplotype analysis to have inherited a copy of the disease haplotype from each branch of the pedigree (both type II and type III alleles) and had an extreme stone-forming phenotype. The authors concluded that phenotypic differences between type II and type III cystinuria are likely due to allelism at this locus.

Biochemical Features

Pras et al. (1998) described the biochemical and clinical features of the Libyan Jewish cystinuria shown to be linked to 19q13.1. The levels of cysteine and the dibasic amino acids in heterozygotes supported previous data that cystinuria in Libyan Jews is not a type I disease. Oral loading tests performed with lysine showed some degree of intestinal absorption, but less than that in normal controls. Previous criteria for determining the disease type, based solely on urinary amino acid levels, proved useless due to a very wide range of cystine and the dibasic amino acids excreted by heterozygotes. Urinary cystine levels were useful in distinguishing between unaffected relatives and heterozygotes but were not helpful in distinguishing between heterozygotes and homozygotes. Urinary levels of ornithine or arginine, and the sum of urinary cystine and the dibasic amino acids, could distinguish between the last 2 groups. Among stone-formers, 90% were homozygotes and 10% were heterozygotes; 15% of homozygotes were asymptomatic.

Molecular Genetics

Calonge et al. (1994) sought mutations in the SLC3A1 gene because of its plausible candidacy as the site of the defect in cystinuria. In affected individuals from 8 different families, they identified 6 missense mutations in the SLC3A1 gene (which they referred to as rBAT), which segregated with cystinuria and accounted for 30% of the cystinuria chromosomes studied. Homozygosity for the most common mutation, met467-to-thr (104614.0001), was detected in 3 cystinuric sibs. This M467T mutation nearly abolished the amino acid transport activity induced by rBAT in Xenopus oocytes. Kastner (1994) also found mutations in the SLC3A1 gene; 1 patient was a genetic compound of a deletion in the maternal chromosome and a single base substitution in the paternal chromosome.

Gasparini et al. (1995) pointed out that all mutations identified in the SLC3A1 gene to that point belonged to cystinuria type I alleles, accounting for approximately 44% of all type I cystinuric chromosomes. After analysis of 70% of the FLC3A1 coding region, they had detected normal sequences in cystinuria type II and type III cases. The mutant alleles occurred in homozygous type I/I and in heterozygotes of type I/III, indicating genetic heterogeneity of cystinuria. They referred to linkage data also supporting genetic heterogeneity of cystinuria. Their studies were done in Italians and Spaniards, which may explain their conclusion that genetic heterogeneity of cystinuria exists; Pras et al. (1994) failed to find linkage evidence of heterogeneity in Jewish families which may have come from a more homogeneous background. On the basis of biochemical data, Goodyer et al. (1993) had suggested a complementation model with an interaction and expression of mutated alleles of 2 different genes, 1 for cystinuria type I and the other for cystinuria type III. First-degree relatives of type I patients had no abnormal urinary amino acid excretion, while type II and type III heterozygous individuals showed increased amounts of cystine in the dibasic amino acids in their urine. Moreover, oral cystine loading fails to raise serum cystine levels in type I and type II patients but results in nearly normal elevation of plasma cystine levels in type III patients, thus demonstrating a different intestinal defect. Gasparini et al. (1995) suggested that the genes involved in cystinuria types II and III may be genes coding for cystine transporters expressed in the S1 and S2 segments of the proximal tubule and/or a functionally associated subunit of an oligomeric rBAT transporter.

In Libyan Jewish, North American, Italian, and Spanish patients with non-type I cystinuria, the International Cystinuria Consortium (1999) identified mutations in the SLC7A9 gene. The Libyan Jewish patients were homozygous for a founder missense mutation (604144.0001) that abolished b(0,+)AT amino acid uptake activity when cotransfected with rBAT in COS cells. In other patients, they identified 4 missense mutations and 2 frameshift mutations. The authors were not able to fully differentiate between type II and type III phenotypes: according to the urinary amino acid profile, most of the patients described seemed to have inherited type III cystinuria from both parents, but there were exceptions. The results suggested that types II and III, and in some cases type I, represent allelic differences in SLC7A9. Other factors, genetic and environmental, were probably involved. In 1 patient, mutations in SLC7A9 (604144.0002) and in SLC3A1 (104614.0001) were found. These preliminary results suggested that cystinuria is a digenic disease in some of the mixed type I/non-type I patients and supported the hypothesis of partial genetic complementation (Goodyer et al., 1993).

The International Cystinuria Consortium (1999) offered 2 hypotheses as to why mutations in the SLC3A1 gene are recessive, whereas mutations in the SLC7A9 gene are incompletely recessive. First, if the active b(0,+) transporter is constituted by more than 1 rBAT and b(0,+)AT subunit, 1 mutated allele of the light subunit might produce a dominant defect, whereas 1 mutated allele of the rBAT heavy subunit would produce a trafficking defect. Second, the light subunit might associate with a protein other than rBAT and express cystine transport activity in a different proximal tubular segment. In situ hybridization and immunolocalization studies showed expression of the light subunit in the epithelial cells of the proximal straight tubule, like the heavy subunit, but higher expression in the proximal convoluted tubule. Most of the renal cystine reabsorption occurs in the proximal convoluted tubule via a low-affinity system not identified at the molecular level. If the SLC7A9 gene also encodes this transport system, a partial defect in this major renal reabsorption mechanism would explain the incompletely recessive phenotype of non-type I cystinuria.

Font et al. (2001) reported the genomic structure of SLC7A9 and 28 new mutations in this gene that, together with 7 previously reported, characterized 79% of the mutant alleles in 61 non-type I cystinuria patients. Therefore, SLC7A9 appears to be the main non-type I cystinuria gene. The most frequent SLC7A9 missense mutations found were gly105 to arg (G105R; 604144.0002), val170 to met (V170M; 604144.0001), ala182 to thr (A182T; 604144.0003), and arg333 to trp (R333W; 604144.0008). Among heterozygotes carrying these mutations, A182T heterozygotes showed the lowest urinary excretion values of cystine and dibasic amino acids, correlating with significant residual transport activity in vitro. In contrast, mutations G105R, V170M, and R333W were associated with a complete or nearly complete loss of transport activity, leading to a more severe urinary phenotype in heterozygotes. SLC7A9 mutations located in the putative transmembrane domains of b(0,+)AT and affecting conserved amino acid residues with a small side chain were associated with a severe phenotype, while mutations in nonconserved residues gave rise to a mild phenotype. The authors presented a genotype-phenotype correlation in non-type I cystinuria, and hypothesized that a mild urinary phenotype in heterozygotes may be associated with mutations permitting significant residual transport activity.

Dello Strologo et al. (2002) studied the amino acid excretion patterns of 189 heterozygotes with mutations in either SLC3A1 or SLC7A9. All SLC3A1 carriers and 14% of SLC7A9 carriers showed a normal amino acid urinary pattern (type I phenotype). The remainder of the SLC7A9 carriers showed the non-I phenotype: 80.5% were type III and 5.5%, type II. Dello Strologo et al. (2002) concluded that the traditional classification of cystinuria patients was imprecise and proposed a new classification based on genotype: type A, due to mutation in the SLC3A1 gene; type B, due to mutation in the SLC7A9 gene; and type AB, due to a mutation in both the SLC3A1 and SLC7A9 genes.

Leclerc et al. (2002) identified 2 missense mutations in the SLC7A9 gene (see 604144.0009 and 604144.0010) linked to type I alleles in a type I homozygote and in a patient with mixed (I/II) cystinuria, respectively. They also found that a single SLC7A9 mutation (799insA; see 601411.0011) was present on 2 type II and 2 type III alleles in 4 patients with mixed cystinuria, suggesting that type II and type III cystinuria may be caused by the same mutation and, therefore, that other factors must influence urinary cystine excretion.

Harnevik et al. (2003) analyzed the SLC3A1 and SLC7A9 genes in 16 unclassified Swedish cystinuria patients, 15 of whom were stone-forming. In 1 of the stone-forming patients, Harnevik et al. (2001) had previously identified compound heterozygosity for mutations in the SLC3A1 gene (see 104614.0001 and 104614.0008); this patient was found by Harnevik et al. (2003) to have a mutation in the SLC7A9 gene (604144.0010) as well. In 9 patients, only 1 mutation in SLC3A1 was found; 1 patient had only 1 mutation in SLC7A9; and in 4 patients, no mutations were identified. Harnevik et al. (2003) suggested that other mechanisms of gene inactivation, such as gene silencing, or additional genes may contribute to the pathogenesis of cystinuria.

Font-Llitjos et al. (2005) classified 164 unrelated cystinuria patients and their relatives on the basis of urine excretion of cystine and dibasic amino acids by obligate heterozygotes and screened for mutations in the SLC3A1 and SLC7A9 genes. They identified phenotype I heterozygotes with mutations in SLC7A9 (e.g., 604144.0001-604144.0004 and 604144.0012) and phenotype non-I heterozygotes with duplication of exons 5 to 9 of the SLC3A1 gene (104614.0007). Font-Llitjos et al. (2005) also identified 2 individuals of mixed phenotype and digenic inheritance with 3 mutations each: 1 had a mutation in each SLC3A1 allele and a mutation in 1 SLC7A9 allele (104614.0001, 104614.0007, and 604144.0013, respectively), and the other had a mutation in each SLC7A9 allele and a mutation in 1 SLC3A1 allele (604144.0002, 604144.0012, and 104614.0001, respectively).

In a brother and sister with isolated hypercystinuria previously reported by Brodehl et al. (1967), Eggermann et al. (2007) identified a likely causative mutation in the SLC7A9 gene (T123M; 604144.0014). Both sibs, who had never formed urinary stones, also carried an I260M variant in the SLC7A9 gene that was not found in more than 100 controls; however, it was also present in their healthy older sister who had normal aminoaciduria values, and Eggermann et al. (2007) concluded that I260M is a rare polymorphism. The female proband had also been noted to have isolated hypoparathyroidism by Brodehl et al. (1967); during follow-up with Eggermann et al. (2007), she was diagnosed with autoimmune polyendocrinopathy type I (APS1; 240300) and found to be compound heterozygous for 2 common mutations in the AIRE1 gene (607358.0001 and 607358.0003), which were not found in her healthy sister or her brother.

Barbosa et al. (2012) reported 12 Portuguese probands with cystinuria, who were classified as homozygous (7 patients) or heterozygous non-type I (5 patients) according to the concentration of cystine in the urine. Among the 7 homozygous patients, 6 had onset of lithiasis or urinary tract infection in the first or second decades. The seventh patient was ascertained in infancy due to neonatal hypotonia. The 6 patients with urinary symptoms all had relatives with lithiasis. Among the 5 non-type I patients diagnosed as children, 2 presented with lithiasis and 2 were ascertained during workup for developmental delay and autism spectrum disorder, respectively. Molecular analysis showed that 6 of the 7 homozygous patients had 2 mutations in the SLC3A1 gene (see, e.g., 104614.0001; 104614.0007; 104614.0009). Three of the patients were compound heterozygous for the exon 5-9 duplication (104614.0007) and another pathogenic SLC3A1 mutation. The seventh patient had 1 mutation in the SLC7A9 gene and another variant in the SLC7A9 gene that may have contributed to the disorder. Four of the 5 non-type I patients had a mutation in the SLC7A9 gene; 1 patient was heterozygous for the exon 5-9 duplication in SLC3A1. Overall, the most common pathogenic mutations in both genes were large genomic rearrangements (33.3% of mutant alleles) and M467T in SLC3A1 (104614.0001) (11.1% of mutant alleles).

Pathogenesis

Bartoccioni et al. (2008) analyzed assembly of wildtype SLC3A1 and type I cystinuria SLC3A1 mutants with SLC7A9 and found that most of the transmembrane domain L89P-mutant SLC3A1 did not heterodimerize with SLC7A9 and was degraded, but a few L89P mutant/SLC7A9 heterodimers were stable, consistent with assembly rather than folding defects. Mutants of the SLC3A1 extracellular domain (e.g., M467T, 604144.0001; M467K, 604144.0002; T216M; and R365W) efficiently assembled with SLC7A9 but were subsequently degraded. Bartoccioni et al. (2008) suggested that biogenesis occurs in 2 steps, with early assembly of the subunits followed by folding of the SLC3A1 extracellular domain, and that defects in either of these steps lead to the type I cystinuria phenotype.

Population Genetics

The overall prevalence of cystinuria is approximately 1 in 7,000 neonates, ranging from 1 in 2,500 neonates in Libyan Jews to 1 in 100,000 among Swedes (review by Barbosa et al., 2012).

Animal Model

McNamara et al. (1989) found that the cystine defect in cystinuric stone-forming dogs is reflected in isolated brush-border membranes, whereas the alteration responsible for the cystinuria of Basenji dogs with Fanconi syndrome did not appear to have a membrane locus.

In an N-ethyl-N-nitrosourea mutagenesis screen for recessive mutations, Peters et al. (2003) identified a mutant mouse with elevated concentrations of lysine, arginine, and ornithine in urine, displaying the clinical syndrome of urolithiasis and its complications. Positional cloning of the causative mutation identified a missense mutation in Slc3a1, leading to an amino acid exchange D140G in the extracellular domain of the rBAT protein. The mouse model mimics the etiology and clinical manifestations of human cystinuria type I.