Pediatric Nephrology

, Volume 24, Issue 10, pp 1967–1973

Locus heterogeneity of Dent’s disease: OCRL1 and TMEM27 genes in patients with no CLCN5 mutations

Authors

  • Enrica Tosetto
    • Division of Nephrology, Department of Medical and Surgical SciencesUniversity of Padua
  • Maria Addis
    • Department of Biomedical Sciences and BiotechnologyUniversity of Cagliari
  • Gianluca Caridi
    • G. Gaslini Pediatric Institute
  • Cristiana Meloni
    • Department of Biomedical Sciences and BiotechnologyUniversity of Cagliari
  • Francesco Emma
    • Division of Nephrology and DialysisBambin Gesù Pediatric Hospital
  • Gianluca Vergine
    • Division of Nephrology and DialysisBambin Gesù Pediatric Hospital
  • Gilda Stringini
    • Division of Nephrology and DialysisBambin Gesù Pediatric Hospital
  • Teresa Papalia
    • Department of Nephrology‘Annunziata’ Hospital
  • Giancarlo Barbano
    • G. Gaslini Pediatric Institute
  • Gian Marco Ghiggeri
    • G. Gaslini Pediatric Institute
  • Laura Ruggeri
    • Department of PediatricsUniversity of Brescia
  • Nunzia Miglietti
    • Department of PediatricsUniversity of Brescia
  • Angela D′Angelo
    • Division of Nephrology, Department of Medical and Surgical SciencesUniversity of Padua
  • Maria Antonietta Melis
    • Department of Biomedical Sciences and BiotechnologyUniversity of Cagliari
    • Division of Nephrology, Department of Medical and Surgical SciencesUniversity of Padua
Original Article

DOI: 10.1007/s00467-009-1228-4

Cite this article as:
Tosetto, E., Addis, M., Caridi, G. et al. Pediatr Nephrol (2009) 24: 1967. doi:10.1007/s00467-009-1228-4

Abstract

Dent′s disease is an X-linked renal tubulopathy caused by mutations mainly affecting the CLCN5 gene. Defects in the OCRL1 gene, which is usually mutated in patients with Lowe syndrome, have recently been shown to lead to a Dent-like phenotype, called Dent’s disease 2. About 25% of Dent’s disease patients do not carry CLCN5/OCRL1 mutations. The CLCN4 and SLC9A6 genes have been investigated, but no mutations have been identified. The recent discovery of a novel mediator of renal amino acid transport, collectrin (the TMEM27 gene), may provide new insight on the pathogenesis of Dent’s disease. We studied 31 patients showing a phenotype resembling Dent’s disease but lacking any CLCN5 mutations by direct sequencing of the OCRL1 and TMEM27 genes. Five novel mutations, L88X, P161HfsX167, F270S, D506N and E720D, in the OCRL1 gene, which have not previously been reported in patients with Dent’s or Lowe disease, were identified among 11 patients with the classical Dent’s disease phenotype. No TMEM27 gene mutations were discovered among 26 patients, 20 of whom had an incomplete Dent’s disease phenotype. Our findings confirm that OCRL1 is involved in the functional defects characteristic of Dent’s disease and suggest that patients carrying missense mutations in exons where many Lowe mutations are mapped may represent a phenotypic variant of Lowe syndrome.

Keywords

Dent’s disease 2Genotype-phenotype correlationLowe syndromeOCRL1 mutationsTMEM27 gene

Introduction

Dent’s disease (DD) is an X-linked renal tubular function disorder characterized by low-molecular-weight proteinuria (LMWP) and a number of other features of Fanconi syndrome, such as glycosuria, aminoaciduria and phosphaturia, but not proximal tubular acidosis. Hypercalciuria is common among patients with DD, as are nephrocalcinosis and kidney stones. No extrarenal manifestations have been recognized other than rickets, which presents in a minority of patients [13]. CLCN5 gene mutations have consistently been reported in patients with DD [Dent’s disease 1 (DD 1), OMIM 300009].

The CLCN5 gene, located in Xp11.22, encodes the ClC-5 Cl-/H+ exchanger, which is believed to be crucial to acidification of the endosomes participating in solute reabsorption and plasma membrane receptor recycling in the proximal tubule [4]. Disruption of the mouse homolog of this gene produces a phenotype resembling the form of the human disease, thereby confirming the role of this gene in the human disease [5, 6].

In all, 107 different mutations have been reported to date in families with DD, but 30–40% of patients with classic DD symptoms do not have CLCN5 gene mutations [79]. Recent investigations have shown that defects in the OCRL1 gene encoding a phosphatidyl-inositol 4,5-bisphosphate 5-phosphatase are found in some families with DD, indicating a genetic heterogeneity for this disorder [10]. Dent’s disease associated with OCRL1 mutations is now termed Dent’s disease 2 (DD2, OMIM 300555).

Mutations in the OCRL1 gene were initially found to cause Lowe syndrome, an X-linked recessive multisystem disorder (OMIM 309000) affecting the eyes, nervous system and kidney [11, 12]. Renal manifestations in Lowe syndrome are very similar to those of DD, although the characteristics of the tubular dysfunction are different. Tubular acidosis is one of the cardinal signs of Lowe syndrome, whereas in DD, this condition is rare. Features of Fanconi syndrome are also observed more frequently in patients with Lowe syndrome than in patients with DD [13]. OCRL1 mutations, including frame shift mutations, splice defects and missense mutations, have been found in about 15% of DD patients, whereas other, as yet unidentified genes may be involved in the remaining 20% of DD patients [10, 14]. The CLCN4 and the SLC9A6 genes (located on Xp22.3 and Xq26.3, respectively) were investigated by Hoopes et al. [10], but no mutations were identified.

We have therefore explored the possibility of another gene harboring mutations. We investigated the TMEM27 gene that maps to chromosome Xp22.2, encoding a transmembrane glycoprotein called collectrin. This gene is expressed mainly on the apical membrane of proximal tubuli and is considered to be a novel mediator of amino acid transport through an as yet incompletely understood mechanism also related to vesicular trafficking [1517].

We identified 31 patients with a phenotype resembling DD and who were determined to lack CLCN5 mutations. These patients were screened for any mutations in the OCRL1 and TMEM27 genes, and five new OCRL1 mutations were discovered.

Methods

Subjects

A total of 31 male patients with all or some of the symptoms of DD, but no CLCN5 gene mutations, were enrolled in this study. Eleven patients with the classical DD phenotype, according to the criteria adopted in our previous work [14], were studied for OCRL1 mutations. These patients presented with at least three hallmarks of DD, including LMWP, hypercalciuria and at least one of the following: nephrocalcinosis, nephrolithiasis, hypophosphatemia, renal failure, rickets or other features of a proximal renal tubular defect (e.g. hypophosphatemia, aminoaciduria, glycosuria, a positive family history). Twenty-six patients were selected for TMEM27 mutation analysis: six patients with three hallmarks of DD who had no OCRL1 mutations, and 20 with an incomplete DD phenotype, i.e. only two of the above hallmarks [14]. All patients were of Italian origin except for two siblings from the Cape Verde islands. Blood samples were drawn from all individuals after obtaining their informed consent, and these were used for TMEM27 and OCRL1 gene mutation analyses.

Mutation analyses

Genomic DNA was extracted from the patients’ peripheral blood using NucleoSpin Blood Quick Pure minicolumns (Macherey-Nagel GmbH, Duren, Germany).

Six pairs of primers were used for PCR amplification of the human TMEM27 gene. Primer sequences and amplified product lengths are given in Table 1.
Table 1

Primer pairs used for TMEM27 gene mutation analysis

Exon

Primers

Size of PCR product (bp)

Amplification cycles (n)

1

f: 5′-TGGTGGCTCGCTTGTTTC-3′

300

30

r: 5′-ACTGCCCCAACTTTCAAGC-3′

2

f: 5′-CAGGAGTATTTGGGGCTGTT-3′

277

30

r: 5′-TGCAGACAGGCCCATTATTAG-3′

3

f: 5′-GGAGAGCCACTGTGGGTACA-3′

284

30

r: 5′-AAAGTGTGAATCCCTTTGAAAAA-3′

4

f: 5′-TGATTTTGAAACAATGGGAATTT-3′

488

30

r: 5′-TGCCCCTGGATGAGAACTAC-3′

5

f: 5′-GCCAGGAGGATGCTTTGTT-3′

394

30

r: 5′-GGAAAATCCTCTCCTGATTTG-3′

6

f: 5′-TGGTCTTTGAAATTCGTTTGA-3′

495

30

r: 5′- TTCAGTGGTGTTGGTGGGTA-3′

f, Forward; r, reverse

PCR Master mix (Roche Diagnostic, Mannheim, Germany) was used to amplify 50 ng of DNA in a final reaction volume of 50 µl. The amplification profile was the same for each primer set and consisted of initial denaturation at 94°C for 1 min, followed by 30 amplification cycles (30 s at 94°C, 30 s at 58°C, 30 s at 72°C), with a final extension at 72°C for 7 min. PCR products were analyzed using the Agilent bioanalyzer technology (Agilent Technologies, Waldbronn, Germany) and were purified using the QIAquick DNA Purification kit (Qiagen, Hilden, Germany). Direct automated PCR products were sequenced using the ABI PRISM GENESCAN 373A DNA sequencer and the BigDye Terminator v1.1 Cycle Sequencing kit (PE Applied Biosystem, Foster City, CA). Sequences of all 23 coding exons and the flanking intronic sequence of the human OCRL1 gene were analyzed according to Addis et al. [18]. The nomenclature of the mutations is based on the cDNA sequence described by the National Human Genome Research Institute (NHGRI: http://research.nhgri.nih.gov/), choosing Met-18 as the amino acid number 1 in accordance with previously conducted similar studies [10, 13, 19, 20].

All of the nucleotide variations observed were confirmed by independent PCR reactions.

The missense mutations were confirmed by studying 100 normal X chromosomes by direct sequencing of exons 10, 15 and 19.

Results

Of the 31 male patients with all or some of the symptoms of DD but no CLCN5 gene mutations, 11 with the classical DD phenotype were selected for OCRL1 gene analysis, and five new mutations were identified.

The five novel OCRL1 mutations were found in six patients from five families: three were missense mutations, F270S, D506N and E720D, in exons 10, 15 and 19, respectively, where many known OCRL1 mutations map. The other mutations were a L88X nonsense mutation and a P161HfsX167 frame-shift mutation. The latter two mutations occurred in OCRL1 exons that had never been known to be involved in Lowe syndrome. According to the OMIM nomenclature, these patients were considered as having DD 2.

Table 2 lists the clinical and biochemical characteristics of the renal tubulopathy of DD 2 patients in detail. None of the patients had nephrolithiasis, but two of them had nephrocalcinosis based on the results of an ultrasound examination. None had clinical rickets, though one had a mildly reduced bone mineral density. As is commonly seen in DD, these patients had incomplete Fanconi syndrome—i.e. three patients had aminoaciduria and one had glycosuria.
Table 2

Clinical and biochemical features of patients (P1–P6) carrying OCRL1 mutations

Clinical and biochemical features of patients

P1

P2

P3

P4

P5

P6

Normal range (child)

Age at diagnosis (years)

13

5

10

4

10

11

 

Serum component

  Ca (mmol/l)

2.40

2.40

5.28

2.42

2.47

2.30

2.1–2.6

  P (mmol/l)

1.26

1.45

2.5

1.58

1.13

1.13

0.85–1.55

  PTH (ng/l)

18

3

36

na

20

25

10–65

  Uricemia (µmol/l)

244

125

244

125

107

161

119–416

  HCO3 (mmol/l)

26

26

26

25

17

17

22–28

Urine component

  β2 microglobulin (μg/24 h)

12 500

15 300

81 567

53 900

14 500

16 300

<300

  Ca/Cr (mg/mg)

0.7

  

0.73

0.29

0.31

<0.25a

  Ca (mg/kg/24 h)

5.8

4

<4

  Proteinuria (mg/24 h)

1300

623

690

960

2100

1530

0–200

  Aminoaciduria

+

+

-

+

-

+/-

-

  Glycosuria

+/-

+

-

-

+/-

+/-

-

  Uricosuria (F. Cl. %)

14

na

na

na

32

26

<12a

  Ccr (mL/min per 1.73 m2)

94

108

na

114.4

105

99

65–150 (4–14 years)

  Hypophosphatemia

-

-

-

-

+/-

+/-

 

  Nephrocalcinosis

-

-

+

-

+

-

 

  Nephrolithiasis

-

-

-

-

-

-

 

  Rickets

-

-

-

-

-

-

 

  Acidosis

-

-

-

-

+

+

 

  Family history

+

+

+

-

+

+

 

Extrarenal involvement

  Increase in serum CK levels (IU/l)

368

336

-

238

-

-

<229

  Ocular involvement

-

-

-

+

+

+

 

  Cognitive/ behavioral impairment

-

+

+

-

-

-

 

  MUTATION

L88X

P161HfsX167

F270S

D506N

E720D

E720D

 

Ca, Calcium; P, phosphate; PTH, parathyroid hormone; HCO3-, bicarbonate; Cr, urine creatinine; F. Cl., fractional clearance; Ccr, creatinine clearance; +, present; -, absent; na, not available

aThe normal range in children varies with age

Ocular anomalies were found in the two brothers (P5 and P6) and in patient P4; some degree of neurological and behavioral impairment was reported in patients P2 and P3, while patients P1, P2 and P4 also had high creatine kinase (CK) levels. All patients underwent slit-lamp examination after their OCRL1 mutations were discovered, but none underwent any formal developmental testing.

No TMEM27 gene mutations were detected among the six of 11 patients with the classical DD phenotype who had a normal OCRL1 sequence, nor among the 20 with an atypical DD phenotype.

Discussion

The five newly identified mutations extend the molecular spectrum of the OCRL1 mutations associated with DD 2. These mutations were found among 11 patients selected as having a classical DD nephropathy (according to the criteria adopted in our previous studies [14]), but no CLCN5 mutations. These mutations had not been reported earlier in patients with Dent’s or Lowe disease, thus confirming that allele heterogeneity accounts for the phenotypic heterogeneity. It is worth noting, however, that although the L88X and P161HfsX167 mutations are novel, they introduce a premature termination of translation at the same codons as the frame-shift mutations Q70RfsX88 and Q152RfsX167 identified in two DD 2 patients by other authors [13, 19]. Similar to the patients reported in Utsch et al’s [19], and Cho et al.’s [13] studies, our patients (P1 and P2) had no ocular anomalies, and only one had a mild cognitive impairment, despite a severely truncated OCRL1 protein.

The functional significance of the three novel missense mutations, F270S, D506N and E720D, appears less clear. The E720D mutation is thought to produce a rather conservative substitution, whereas the F270S and D506N mutations resulted in a non-conservative amino acid replacement, thereby leading to a disrupted charge distribution in the highly conserved exonuclease–endonuclease–phosphatase family domain of the OCRL1 protein. It is noteworthy that this domain has always been involved in all of the missense mutations detected in DD 2 patients to date.

Confirmation that these three mutations were missense mutations was obtained by directly sequencing exons 10, 15 and 19 of 100 normal X chromosomes: these mutations were not detected in any of the exons analyzed. Moreover, the SIFT (http://blocks.fhcrc.org/sift/SIFT.html) [21] and PolyPhen (http://genetics.bwh.harvard.edu/pph) [22, 23] computer programs predict that these missense mutations affect protein function. Taken together, these results indicate that these missense mutations are very likely pathogenic.

The F270S mutation was found in one patient (P3) who had some learning difficulties and behavioral impairments. Judging from the first studies by Hoopes et al. [10], mild cognitive or behavioral impairment seems to be a relatively common feature of DD 2 patients, as are mildly elevated serum levels of muscle enzymes. Elevated CK levels with no muscle weakness and mild cognitive or behavioral impairments were found in three and two, respectively, of our six patients. It is worth adding, however, that two of the six DD patients in our series with no OCRL1 mutations also had cognitive and/or behavioral impairments. Table 3 summarizes the clinical features of our DD patients with and without OCRL1 mutations and shows that there were no phenotypic differences, although extrarenal manifestations and a positive family history seem to occur more frequently in OCRL1-positive patients.
Table 3

Clinical features in Dent’s disease patients with or without OCRL1 gene mutations

Clinical features

DD patients with mutations

DD patients without mutations

Age (years)

Median 9 (range 4–13)

Median 14 (range 9–38)

LMW proteinuria

6/6

6/6

Hypercalciuria

6/6

6/6

Nephrocalcinosis

2/6

3/6

Nephrolithiasis

0/6

1/6

Renal insufficiency

0/5

1/6

Rickets

0/6

1/6

Other proximal tubular defects

6/6

5/5

Positive family history

5/6

2/6

Extrarenal manifestations (CK not included)

5/6

2/6

DD, Dent’s disease; LMW, low molecular weight

The case of the E720D missense mutation seems intriguing as it is the most conservative mutation and occurs in exon 19, which is considered not to be part of the phosphatase domain, but of the COOH-terminal, catalytically inactive ASH-RhoGAP domain. The mutation was found to segregate in two brothers affected with LMWP, hypercalciuria and tubular acidosis (P5 and P6); one of the brothers also had nephrocalcinosis. The patients had no neurological signs, but they did present with ocular anomalies on slit-lamp examination (mild bilateral cortical lens opacity). The presence of acidosis in these patients is striking, since it contrasts with all other reported cases of DD 2. These patients are also the only reported cases of DD 2 with a mutation in the ASH–RhoGAP domain, suggesting that they may have a mild Lowe syndrome phenotype. Since both patients also carry the silent P568P mutation in the CLCN5 gene [14], another possibility to consider is that the nucleotide substitution in the CLCN5 gene may be a modifier allele, exerting its effect by preventing a complete Lowe phenotype. Indeed, in an earlier study, we demonstrated that, by affecting the strength of a cryptic splice site in exon 10, the silent P568P mutation (1195 C > T transition) was able to induce changes in the expression of an alternative, spliced, shorter ClC-5 transcript [14]. However, the functional significance of the short isoform that we found remains to be elucidated. The findings of the investigation reported here would support a potentially protective role of its over-expression in relation to the signs of the Lowe phenotype. These results may mean that the phenotype variability in Lowe syndrome cannot be attributed to allele differences in the OCRL1 gene alone, but that modifying loci could significantly alter the phenotypic consequences of an OCRL1 protein deficiency. Further research is needed to confirm—or not—this hypothesis.

It is noteworthy that our patients with mutations in exons 10, 15 and 19 (where many known Lowe mutations are mapped) had, in addition to typical Dent’s tubulopathy, acidosis and/or ocular or neurological impairments. We hypothesized that these patients may represent a subgroup of patients with very mild symptoms of Lowe syndrome.

To test this hypothesis, we examined the OCRL1 mutations of the DD 2 patients described to date in earlier studies and the associated phenotypes [10, 13, 19, 20]. In total, 12 different OCRL1 mutations have been described in 15 unrelated DD 2 patients. All of the mutations located in exons 5–7, including intron 7, are frame-shift or splice-site mutations leading to premature stop codons, while the missense mutations are concentrated in exons 9–15. No cases of cataract or tubular acidosis were identified, while cognitive and/or behavioral impairments were recorded in 5/15 patients, and high CK levels in 9/14. When we classified these patients into two groups according to type of genetic lesion (premature stop codon vs. missense) and by OCRL1 gene domain (exons 5–7 vs. exons 9–15), we found that four of the seven patients harboring missense mutations, but only one of the eight carrying stop codon mutations, have cognitive/behavioral impairments. High serum CK levels are equally distributed between the two groups. The situation in our patients is consistent with these findings, suggesting that the worst DD phenotype is associated with OCRL1 mutations occurring in the protein domains involved in Lowe syndrome, thereby supports the hypothesis that this might characterize a subgroup of patients with very mild symptoms of Lowe syndrome.

Surprisingly, although all of the premature stop mutations occurred in the first exons of the gene and thus led to a protein being absent, they were associated almost exclusively with the renal phenotype, not with any ocular and/or mental impairments. Indeed, the premature stop codons due to frame-shift and splice-site mutations were shown to induce the absence of the OCRL1 protein by Western blot analysis, whereas the missense mutations cause reduced protein levels and a diminished PtdIns(4,5)P2 5-phosphatase activity [10].

These unexpected results seem to suggest that the target cells tolerate the absence of the protein better than the presence of an abnormal protein. As such, this behavior resembles the dominant negative effect of missense mutations in multimeric proteins, such as collagens and some transcription factors and receptors. A mutant molecule that packs or interacts erroneously poses more problems than one that simply goes unrecognized or is absent. OCRL1 is not a multimeric protein, but its capacity to bind numerous proteins has recently been demonstrated. In fact, it has been found to be capable of linking several proteins of the endocytotic apparatus [2427]. In particular, OCRL1 was also found to bind to, and become co-located on endosomes with the endosomal adaptor protein, APPL1 [28]. Very recent evidence has shown that, among the interactions targeting OCRL1 to membranes of the endocytic pathway, binding to APPL1 is the only one abolished by all known missense mutations causing Lowe syndrome in the ASH–RhoGAP domains of the protein (where exon 19 belongs) [29, 30].

Not all CLCN5-negative DD patients carry mutations in the OCRL1 gene. Other, as yet unidentified genes may be involved in 20% of DD patients. We decided to explore the possibility of TMEM27 gene involvement, given its location on the X chromosome and its instrumental role in proximal tubular function [17]. We analyzed 26 patients, i.e. six with the classical DD phenotype but no CLCN5 or OCRL1 mutations, and 20 with an incomplete DD phenotype. Direct TMEM27 gene sequencing analysis failed to identify any DNA sequence abnormalities in our series of patients.

Our findings confirm, once again, that OCRL1 is relevant to the functional defects characteristic of DD. They also suggest that: (1) although premature stop mutations occur in the first exons of the gene, they are associated almost exclusively with the renal phenotype; (2) patients with missense mutations in exons typically involved in Lowe syndrome may represent a phenotypic variant of Lowe syndrome (i.e. a very mild Lowe syndrome); (3) modifying loci may significantly alter the phenotypic consequences of an OCRL1 protein deficiency; (4) TMEM27 gene mutations are unlikely to be a cause of DD in this cohort of patients.

Acknowledgements

This study was supported by grant No. 2005063822_004 from the Italian Ministry of Education, University and Research.

This work received an award at the ERA-EDTA XLIV Congress (Barcelona, Spain 21–24 June 2007) as one of the best abstracts presented by young authors.

Copyright information

© IPNA 2009