Pediatr Nephrol (2006) 21: 1241–1250DOI 10.1007/s00467-006-0172-9

ORIGINAL ARTICLE

Michael Ludwig . Boris Utsch . Bernd Balluch .Stefan Fründ . Eberhard Kuwertz-Bröking .Arend Bökenkamp

Hypercalciuria in patients with CLCN5 mutations

Received: 10 January 2006 / Revised: 26 March 2006 / Accepted: 27 March 2006 / Published online: 29 June 2006# IPNA 2006

Abstract Hypercalciuria is regarded as a characteristicsymptom of Dent disease, an X-linked recessive tubulo-pathy characterized by low molecular weight (LMW)proteinuria, nephrocalcinosis/nephrolithiasis, and progres-sive renal failure due to mutations in the CLCN5 gene. Asthe presence of hypercalciuria may affect the decision toconsider aCLCN5mutation in the differential diagnosis, thephenotypic spectrum and the relative frequency of hyper-calciuria in patients with CLCN5 mutations was deter-mined. We assessed renal calcium excretion in 34 malepatients with proven CLCN5 mutations, who had beenreferred because of LMW proteinuria and at least oneadditional symptom of Dent disease. Hypercalciuria wasdefined as renal calcium excretion exceeding 0.1 mmol/kgper day. Data obtained were compared with all series ofCLCN5-positive patients identified by a systematic litera-ture survey. In 7 of our 19 families, at least 1 affected malehad normal calcium excretion. Hypercalciuria was observed

in 22 of 31 patients tested (71%) compared to 85 of 90(94.4%) in series from Europe and North America and74.4% from Japan. LMW proteinuria was present in allCLCN5-positive patients; 25% of the patients in Europeanand North American series, 45% of the Japanese, and 41%in the present series had only two of the four principalsymptoms of Dent disease. Therefore, a CLCN5 mutationshould be considered irrespective of the presence ofhypercalciuria in a patient with LMW proteinuria and oneadditional symptom of Dent disease.

Keywords Dent disease . Hypercalciuria .LMW proteinuria . Mutation . Nephrocalcinosis .Voltage-gated chloride channel-chloride/proton antiporter5 gene (CLCN5)

Introduction

Dent disease (OMIM 300009) is an X-linked recessiveFanconi-like syndrome [1, 2] due to mutations in CLCN5[3], the gene encoding the voltage-dependent chloridechannel and chloride/proton exchanger (ClC-5) [4, 5]. Thedisease is characterized by low molecular weight (LMW)proteinuria, nephrocalcinosis/nephrolithiasis, hypercalci-uria, and progressive renal failure in affected males [6,7]. Variable associated symptoms are hyperaminoaciduria,glycosuria, impaired urinary acidification as well as renalphosphate and potassium wasting [7].

Lloyd et al. [8] demonstrated mutations in the CLCN5gene in the phenotypically similar disorders X-linkedrecessive nephrolithiasis (XRN) and X-linked recessivehypophosphatemic rickets (XLHR). Familial idiopathicLMW proteinuria in Japanese patients (JILMWP)—alsoreferred to as Dent Japan disease [9, 10]—is a fourthclinical entity resulting from CLCN5 mutations [8, 11].Therefore, Scheinman proposed to summarize CLCN5-associated renal disease under the term “X-linked hyper-calciuric nephrolithiasis” [12].

In the kidney, ClC-5 is expressed in recycling earlyendosomes beneath the brush border of proximal tubular

M. Ludwig (*)Department of Clinical Biochemistry,University of Bonn,Sigmund-Freud-Str. 25,53105 Bonn, Germanye-mail: mludwig@uni-bonn.deTel.: +49-228-2875969Fax: +49-228-2875028

B. UtschDepartment of Pediatrics,University of Erlangen-Nuremberg,Erlangen, Germany

B. BalluchGottfried von Preyer’sches Kinderhospital,Vienna, Austria

S. Fründ . E. Kuwertz-BrökingDepartment of Pediatrics,University Children’s Hospital,Münster, Germany

A. BökenkampDivision of Pediatric Nephrology,Dept. of Pediatrics, Vrije Universiteit Medical Center,Amsterdam, The Netherlands

cells [13], where it is critical for acidification [14]. CLCN5knockout in mice disrupts both megalin-mediated andfluid-phase endocytosis [15]. This explains the massiveurinary loss of LMW proteins, including parathormone andvitamin D binding protein in ClC-5-associated nephropa-thies [14]. Besides protein ligands, megalin also bindscalcium ions and, hence, may be responsible for non-regulated calcium absorption in the proximal tubule [13].ClC-5 expression has also been documented in themedullary thick ascending limb of Henle’s loop, themajor site of calcium reabsorption along the nephron,and in α-intercalated cells of the collecting duct [13].

Although hypercalciuria was a characteristic symptomof Dent disease in the original descriptions [1, 2], findingsin a CLCN5 disruption model suggest that hypercalciuriamay be a secondary phenomenon reflecting the complexbalance between the loss of 25-hydroxyvitamin D alongwith vitamin D binding protein and the stimulation of 1,25-dihydroxyvitamin D synthesis by increased luminal para-thormone delivery [14].

This raises the question of whether the presence orabsence of hypercalciuria in a patient with other symptomssuggestive of X-linked hypercalciuric nephrolithiasisshould influence the decision to include a CLCN5mutationin the differential diagnosis. Therefore, the present studyset out to analyze the clinical picture of CLCN5 mutationswith special emphasis on the presence of hypercalciuria.

Patients and methods

A total of 35 families were brought to the authors’ attentionby their treating physicians because of clinical featuressuggesting the diagnosis of X-linked hypercalciuric neph-rolithiasis. In order to qualify for CLCN5 gene analysis atleast two of the hallmarks of the disease had to be present:LMW proteinuria, defined by excessive urinary loss of α1-microglobulin, β2-microglobulin, retinol binding proteinor lysozyme; hypercalciuria of more than 0.1 mmol/kg perday in a 24-h urine collection; nephrocalcinosis/nephro-lithiasis on ultrasound or computed tomography, and/orrenal impairment, defined as creatinine clearance below80 ml/min per 1.73 m2. Renal impairment was furtherdifferentiated into mild-moderate [glomerular filtration rate(GFR) 30–80 ml/min per 1.73 m2] and severe (GFR<30 ml/min per 1.73 m2 or renal transplantation). Patientswere regarded as not having hypercalciuria if calciumexcretion had been measured on at least three occasionsand had never exceeded 0.1 mmol/kg per 24 h.

After informed consent was given, ethylenediaminetet-raacetate (EDTA) blood samples were obtained fromfamily members. Isolation of genomic DNA was carriedout by the use of standard procedures and amplification bypolymerase chain reaction (PCR) and automated sequenceanalysis was performed as described in detail elsewhere[16]. In brief, primers—given in Ludwig et al. [17]—weredirected to all 17 exons of the human CLCN5 gene [3, 18],and resultant PCR products were subjected to directautomated sequencing (373A, Applied Biosystems, Foster

City, CA, USA). Initially, both strands from the indexpatient’s amplicons were sequenced. All nucleotide varia-tions observed were (1) confirmed by independent PCRreactions and (2) segregation of these variants in familymembers was investigated by sequencing the respectivePCR products. DNA from 75 (50 male samples and 25female samples=100 X-chromosomes) unrelated healthyCaucasian subjects served as normal controls.

A systematic MEDLINE search was carried out forpapers presenting the clinical phenotype of patients withproven CLCN5 mutations using the search strategy {“ClC-5 protein, human”[Substance name]) OR ((CLCN5) AND(sensitiv*[Title/Abstract] OR sensitivity and specifity[MeSH Terms] OR diagnos* [Title/Abstract] OR diagnosis[MeSH:noexp] OR diagnostic* [MeSH:noexp] OR diag-nosis, differential [MeSH:noexp] OR diagnosis [Subhead-ing:noexp])) Limits: Humans}.

Four reviews were excluded from the analysis but servedas additional reference to assess whether all relevant papershad been identified. The remaining 36 papers wereanalyzed independently by two of the authors (A.B. andM.L.) who identified 24 reports, which contained data onLMW proteinuria, renal calcium excretion, nephrolithiasis/nephrocalcinosis, and renal failure in patients with provenCLCN5 mutations. Row data on individual patients wereextracted and classified in analogy to those stated for ourown series. If no row data were available, the classificationused in the original paper was adopted.

Data were analyzed using JMP IN statistical software5.1.2 (SAS Institute, Cary, NC, USA).

Results

CLCN5 gene analysis was performed in patients from 35families with signs and symptoms typical of Dent disease.No mutation was found in 16 of these families and data ofsome of these patients were given in detail elsewhere [18].A single deviation from the normal sequence could beidentified in patients from 19 families presented in Table 1.These include 12CLCN5mutations that have been reportedpreviously [17, 19]. From the seven additional mutationspresented here (R34X, G88D, W95X, S244L, R347X,L521Rfs526X, K725E), three (G88D,W95X, K725E) havenot been observed in patients with Dent disease thus far.

When samples of additional family members were avail-able, the respective mutation segregated with the disease.Aside from the S244L mutation, a single private intronicvariation could be established in family F1: here, both thecarrier mother and her affected son showed an IVS6 -35Ginstead of a T in the heterozygous and hemizygous state,respectively. Moreover, the S244L mutation was found tohave arisen de novo since it could not be detected in thematernal grandparents. None of the control samples showedany of the mutations/variations observed.

The clinical features of the 19 index patients positive fora CLCN5 mutation and their affected family members aresummarized in Table 1. Mean (±standard deviation) patientage was 20.4±17.0 years with a median of 14 years. Index

1242

Table 1 Clinical features and CLCN5 mutations found in families with Dent disease

Origin Individuala Age(years)

Phenotypeb Exon andmutationc

Amino acidchange

Consequences of mutation and comments [reference]

Turkey A-1 13 1, 2, 3, 4 2, C391T R34X Also observed with hypercalciuria [20]; #Germany B-1 7 1, 2, 3 3, G425A W45X Also observed with hypercalciuria [21];

initially reported in [17]Germany C-I.1 7 1, 2 4, G554A G88D #; not reported previously

C-II.1 36 1C-II.2 39 1,3,5

Germany D-1 20 1, 2, 3 4, G575A W95X #; not reported previouslyGermany E-1 10 1, 3 6, T952C C221R Also observed with hypercalciuria [21]; initially

reported in [17] and shown to abolish ClC-5trafficking [17]

Turkey F-1d 7 1, 4 7, C1022T S244L/IVS6 -35Ge Also observed with hypercalciuria [8, 20–23];~30% residual ClC-5 currents [8, 24]; #

Germany G-I.1 9 1, 2, 3 8, T1262G L324R Initially reported in [17] and shown to abolishClC-5 trafficking [17]G-I.2 13 1, 2, 3

Austria H-I.4 6 1, 2, 3 8, G1286A W332X Initially reported in [17]H-I.5 8 1, 3H-II.2 42 1, 3, 5H-III.1 67 1, 3, 4

Finland I-I.3 15 1, 2, 3 8, C1330T R347X Also observed with hypercalciuria[21, 25, 26]; shown to abolishClC-5 trafficking [17]; family Iwas initially reportedin [17] and J-1 reportedin [19]; K-1: #

I-I.5 20 1, 2, 3I-II.3 37 1, 2, 3, 4I-II.5 43 1, 2, 3

Turkey J-1 4 1, 2Turkey K-1 14 1, 3Germany L-I.1 12 1, 3 9, G1676T G462V Initially reported in [17] and shown to abolish

ClC-5 trafficking [17]L-I.2 14 1, 2, 3L-I.3 15 1

Germany M-I.1 22 1, 2, 3 9, C1690T R467X Initially reported in [17]M-I.2 15 1, 2, 3

Germany N-1 24 1, 2, 3 10, C1837T R516W Also observed with hypercalciuria [21, 25];initially reported in [17] and shown to exhibit~30% residual ClC-5 currents and normalprotein trafficking to the plasma membrane [17]

Germany O-I.1 13 1, 3 10, 1842, ins T M517Ifs528X Initially reported in [17] and shown to abolishClC-5 trafficking [17]O-III.1 76 1, 3, 5

Finland P-1d 8 1, 2 10, 1853-54, del TT L521Rfs526X Initially reported in [17] and shown to abolishClC-5 trafficking [17]; Q-1: #Germany Q-1 14 1, 2

Germany R-I.1 18 1, 2, 3 10, C2200T R637X Also observed with hypercalciuria [27, 28]R-I.2 16 1, 2, 3R-III.1 60 1, 2, 3, 5

Germany S-1 9 1, 2 12, A2462G K725E #; not reported previouslyaIndividual numbering with index patients shown in bold is either according to Ludwig et al. [17] or as follows: I patient generation,II parent generation, III grandparent generation, 1 only one patient known

bPhenotypic features are as follows: 1 LMW proteinuria, 2 hypercalciuria, 3 nephrocalcinosis and/or stones, 4 glomerular filtration rate(GFR) 30–80 ml/min per 1.73 m2, 5 GFR <30 ml/min per 1.73 m2 or transplantation

cNucleotide (nt.) numbering is according to the cDNA (GenBank acc. no. NM_000084) with the A of the start-ATG at nt. 292dDe novo mutationePrivate polymorphism, also observed in the carrier mother#=this study

1243

patients were significantly younger (11.9±5.2 years) thanaffected family members (31.9±20.8, p<0.0009, Mann-Whitney U test). In 7 of the 19 families studied, renalcalcium excretion was normal in at least one male with aCLCN5 mutation. Three patients (C-II.2, H-II.2, O-III.1)were excluded from the analysis as calcium excretion hadonly been measured in end-stage renal failure. Therefore,the relative frequency of hypercalciuria in our patients was22 of 31 [71%, 95% confidence interval (CI): 53–84%]. Intwo patients (F-1 and H-III.1) calcium excretion wasnormal in the setting of stage 2 chronic renal disease [29].The age of hypercalciuric patients was comparable topatients with normal calcium excretion (15.9±9.8 vs 20.2±19.5, p=0.89, Mann-Whitney U test). Hypercalciuria wasobserved in 79% of index patients compared to 58% offamily members (p=0.22, chi-square test). Comparingclinical symptoms of patients positive for CLCN5 muta-tions with the 16 patients in whom no CLCN5 mutationwas detected in our laboratory (data not shown), no sig-nificant difference in the prevalence of hypercalciuria wasfound (83%, 95% CI: 55–95%, p=0.45, two-tailed Fisher’sexact test).

In addition, LMW proteinuria was the only symptom intwo patients (Table 1: C-II.1, L-I.3) in our series even at theage of 36 (C-II.1).

The systematic MEDLINE search revealed 24 reportsidentifying CLCN5 mutations as the molecular basis ofDent disease [8, 10, 11, 20–23, 25–28, 30–42]. In these,clinical data were retrieved from 118 families with a total of156 males with clinical signs of X-linked recessivenephrolithiasis/Dent disease, in whom a CLCN5 mutationwas demonstrated (Table 2). For 15 patients, no data onurinary calcium excretion were available, largely becauseurinary calcium had only been measured when the patientalready had advanced renal failure. Of the 141 patientstested, 123 had an increased calcium excretion (87.2%, 95%CI: 81–92%). In reports from Japan and Korea, the relativefrequency of hypercalciuria was lower (74.5%, 95% CI:61–84%) than from Europe and North America (E/NA)(94.4%, 95% CI: 88–98%). Although age distribution wasnot reported in all the papers, patients from the Japaneseseries were significantly younger than in the publishedseries from E/NA and our series (11.0±8.3 vs 19.0±15.7 vs20.4±17.0 years, p=0.01 Kruskal-Wallis test).

Reported cases of CLCN5 mutations without hypercalci-uria are presented in detail in Table 3. Of the 14 CLCN5mutations, 6 (R28X,R34X,H100QfsX106,G506E, R516W,R648X) were observed both with and without hypercalci-uria, even within the same family. In contrast to the otherthree principal symptoms caused by CLCN5 mutations,LMW proteinuria was present in virtually all patientsreported. Nephrocalcinosis was observed in 78% of theEuropean/North American patients, in 55% of the Japanesecases, and in 77% of our patients. Renal dysfunction waspresent in 55% of the cases from the E/NA series comparedto only 7% of the Japanese and 24% of our patients.

Comparison of the clinical phenotype between thepublished series from E/NA, Japan and our own data,with regard to the four principal Dent symptoms caused by

CLCN5 mutations, is given in Fig. 1. Due to the higherfrequency of renal failure observed in patients from the E/NA series, 43.5% (95% CI: 33–54%) of this group had allfour symptoms in contrast to only one patient from Japan(2.6%) and three cases (9.7%) in the present study.Inclusion of the three patients with end-stage renal disease,in whom calcium excretion had not been measured, wouldbring the relative frequency of the complete set of Dentsymptoms in our series to 19.4%. The triad of LMWproteinuria, hypercalciuria, and nephrolithiasis/nephrocal-cinosis was observed at comparable rates (E/NA: 33%,Japan: 50%, this study: 45%) in contrast to an oligosymp-tomatic presentation with only two of the four principalsymptoms (12.9% vs 31.64% vs 35%).

Discussion

Generation of transgenic mice (RZ) with reduced ClC-5expression [43] or targeted disruption of the CLCN5 genein mouse models [15, 44, 45] elucidated the pathogenesisof LMW proteinuria but these models varied as to thepresence [43–45] or absence [15] of hypercalciuria andnephrocalcinosis. Moreover, Günther et al. [14] foundevidence for absorptive hypercalciuria reflecting the bal-ance between the loss of 25-hydroxyvitamin D and thestimulation of 1,25-dihydroxyvitamin D synthesis byincreased luminal parathormone delivery.

According to Scheinman [12], low-grade hypercalciuriaof 4–6 mg/kg per day is characteristic for the X-linkedhypercalciuric nephrolithiasis syndromes, with highercalcium excretion being found in affected children. Halfof the patients are reported to have fasting hypercalciuriaand almost all have an exaggerated calciuric response tooral calcium loading [9, 46]. Therefore, severe hypercal-ciuria is commonly regarded as a characteristic finding inaffected boys, both in textbooks of pediatric nephrology[47] and online medical databases (UpToDate, 29 Novem-ber 2005). Reviewing seven reports presenting data onurinary calcium excretion in males with Dent disease,however, Yu [48] found that hypercalciuria had only beenobserved in three-quarters of the patients. Still, thediagnosis of Dent disease in his paper had been made onclinical grounds and may not apply to patients with provenCLCN5 mutations. Therefore, the present study set out todetermine the relative frequency of hypercalciuria inpatients with defined CLCN5 mutations.

We chose to study only patients with X-linked recessivenephrolithiasis and proven CLCN5 mutations, as manycases diagnosed on clinical findings are CLCN5 negative[18, 21, 25, 26, 32], indicating that other gene defects canmimic Dent disease, such as recently described mutationsin the OCRL1 gene [49].

In our series, about one-third of patients with CLCN5mutations (Table 1) did not have hypercalciuria on repeatedurine sampling and two patients exclusively presented withLMW proteinuria. Using only LMW proteinuria as ascreening parameter for CLCN5 mutations in children,Barbano et al. [50] recently reported a comparable fre-

1244

Tab

le2

Com

bineddata

ofstud

iesrepo

rtingDentpatientswith

CLCN5mutations.NANorth

America,

HC

hypercalciuria,MGFmaternalgrandfather,ESR

Dend-stagerenaldisease

Report[reference]a

Location

Age

bDiagn

ostic

grou

pcFam

ilies

(n)

Totalpatients

(n)d

Patientswith

HC(n)d

Patientswith

outHC

(n)e

Not

tested

forHC

(n)e

Mutations

inallfamilies

detected

d

Lloyd

etal.,

1996

[8]

Europ

e/NA

1– 60b

DentXRN

XLHR

1126

213(D

/III/2;1

0/92

,III.15;

10/92,

III.16

)2(ESRD)

Yes

Lloyd

etal.,

1997

[30]

Japan

3–19

JILMWP

46

60

0Yes

Oud

etet

al.,

1997

[22]

Europ

e8–21

XLHR

14

40

0Yes

Lloyd

etal.,

1997

[31]

Europ

e/NA

1–19

Dent

57

60

1(7.3/95-II.1,

ESRD)

Yes

Nakazatoet

al.,

1997

[11]

Japan

7bJILMWP

28

62(L3,I.1;

L3,III.1)

0Yes

Aku

taet

al.,

1997

[25]

Japan

7–20

JILMWP

912

(8mut.+

4mut.-)7(5

mut.+

2mut.-)

3[1

mut.+(6/95)2

mut.-(9.2/95;

20.2/95)]

2mut.+

No(4

mut.-)

Langloiset

al.,

1998

[32]

NA

2–36

XRN

33(2

mut.+

1mut.-)

3(2

mut.+1mut.-)

01(M

GFin

family

1,ESRD)

No(1

mut.-)

Kelleheret

al.,

1998

[23]

NA

13–

38XRN

15

50

0Yes

Morim

otoet

al.,

1998

[26]

Japan

3– 12b

XLHR

JILMWP

iLMWP

56(5

mut.+1mut.-)

2(1

mut.+

1mut.-)

3(II.1.B,II.2.B,II.3.

D)

1(II.2.C)

No(1

mut.-)

Hoo

peset

al.,

1998

[20]

NA

2–49

Dent

611

80

3(2

ESRD)

Yes

Igarashi

etal.,

1998

[10]

Japan

5–18

JILMWP

57

43(2,3,

11)

0Yes

Nakazatoet

al.,

1999

[33]

Japan

3– 55b

JILMWP

38

31(L1)

4Yes

Bosio

etal.,

1999

[34]

Europ

e2–7

Dent

12

20

0Yes

Cox

etal.,

1999

[35]

Europ

e2–36

Dent

88

71(4/96)

0Yes

Yam

amotoet

al.,

2000

[36]

Europ

e/Japan

1–43

Dent

66

51(26/97

)0

Yes

Igarashi

etal.,

2000

[37]

Japan

3–29

DentJILMWP

23

20

1(pat.4,

ESRD)

Yes

Takem

uraet

al.,

2001

[27]

Japan

2–3

JILMWP

22

20

0Yes

Carballo

-Trujillo

etal.,20

03[38]

Europ

e5–25

Dent

35

50

0Yes

1245

Report[reference]a

Location

Age

bDiagn

ostic

grou

pcFam

ilies

(n)

Total

patients

(n)d

Patientswith

HC

(n)d

Patientswith

outHC

(n)e

Not

tested

forHC

(n)e

Mutations

inallfamilies

detected

d

Claverie-Martin

etal.,20

03[39]

Europ

e7

Dent

11

10

0Yes

Hoo

peset

al.,

2004

[21]

NA

2–39

Dent

3232

(19mut.+13

mut.-)

32(19mut.+13

mut.-)

00

No(13mut.-)

Matsuyamaet

al.,

2004

[40]

Japan

1–12

JILMWP

36

42(1.II.1;

2.II.1)

0Yes

Brakemeier

etal.,

2004

[41]

Europ

e25

Dent

11

01

0Yes

Yanagidaet

al.,

2004

[42]

Japan

15JILMWP

11

10

0Yes

Cheon

get

al.,

2005

[28]

Korea

1–9

Dent

33

30

0Yes

Studies(n)24

Patients

tested

Total

118

175

19mut.-

Mut.+

156

123/14

1(87.2%

)18

/141

(12.8%

)15

Europe/NA

9785

/90(94.4%

)5/90

(5.6%

)7

Japan

5938

/51(74.5%

)13

/51(25.5%

)8

a Referencesaregivenin

chrono

logicalorder

bDataon

patient

ageareincompletein

therespectiv

erepo

rtc D

iagn

ostic

grou

prefers

to:XRN

X-linkedrecessiveneph

rolithiasis,XLHRX-linkedrecessivehy

poph

osph

atem

icrickets,JILMWPfamilial

idiopathic

LMW

proteinu

riain

Japanese

patients,iLMWPidiopathic

LMW

proteinu

ria

dTo

talnu

mberof

casesdividedin

patientswith

(mut.+)or

with

out(m

ut.-)CLCN5mutation

e Ind

ividualnu

mbering

isaccordingto

therespectiv

erepo

rt

Tab

le2(con

tinued)

1246

quency of hypercalciuria in Italian children as presentedherein. This indicates that the relative frequency ofhypercalciuria might be much lower than previouslyreported in European and North American series, where itwas around 97%, but similar to reports from Japan(Table 2). This most likely reflects differences in patientselection, as hypercalciuria was an inclusion criterion inmany of the series reported fromEurope andNorth Americabut not in the Japanese reports, Barbano et al.’s [50] cohort,and the present study. In Japan, many index patients aredetected through the systematic urine screening for hema-turia and proteinuria in schoolchildren [30]. The design ofthe present series shares some features of these reports: allthe index patients were identified by pediatric nephrologistsbased on the presence of at least two symptoms of X-linked

recessive nephrolithiasis, with LMW proteinuria being aconstant finding in all 34 patients. This referral bias isreflected by the younger age of index cases compared toaffected family members in the present series. Still,comparing the proportion of hypercalciuria between indexpatients and their affected family members, there was notrend towards less hypercalciuria in index cases. Addition-ally, hypercalciuria was not correlated with age in ourpatients. Thus, the lower incidence of hypercalciuriadetected here does not appear to be related to patient agebut most likely reflects the wider patient selection.

Renal calcium excretion is known to decrease withdeclining renal function [35]. We therefore excludedpatients with stage 4 and 5 renal failure from our analysisof the frequency of hypercalciuria. We cannot rule out that

Table 3 Dent patients with a proven CLCN5 mutation and without hypercalciuria reported in the literature

Origin [reference] Individuala Age(years)

Phenotypeb Exon andmutationc

Amino acidchange

Consequences of mutation and comments[reference]

Japan [26] II-1-B 6 1 2, T355G W22G Abolished ClC-5 currents [26]II-2-B 3 1

Japan [40] Family 2, II.1 6 1 2, C373T R28X Also observed with hypercalciuriain the same family [40]

North Europe [35] 4/96 35 1, 3, 4 2, C391T R34X Observed with [20, 32] and withouthypercalciuria (this study)

Japan [40] Family 1, II.1 12 1, 3 4, 591insA H100QfsX106 Also observed with hypercalciuriain the same family [40]

Japan [10] Family 5, II.1 16 1, 3 IVS4, A-2C noneJapan [10] Family 4, II.1 18 1 8, G1125C L278F ~30% residual ClC-5 currents [10]Japan [11] Family L3, I.1 ? 1 8, 1499delA Y403LfsX433

Family L3, III.1 7North America [8] Family 10/92,

III-15? 1, 3 9, G1808A G506E Also observed with hypercalciuria in

the same family [8]; abolished ClC-5currents [8] and abolishedtrafficking [17]

Family 10/92,III-16

? 1

Japan [33] Family L1, I.6 ? 1 10, 1834insG T515DfsX527Japan [25] 6/95, III.2 20 1 10, C1837T R516W Also observed with hypercalciuria in

the same family [25] and in [17, 21];~30% residual ClC-5 currents andnormal protein trafficking to theplasma membrane [17]

Japan [10] Family 1, II.1 14 1, 3 11, C2233T R648X Also observed with hypercalciuria[8, 33–35]; ~30% residual ClC-5currents and increase in surfaceexpression compared to wild-typeClC-5 [17]

Family 1, II.2 6 1, 3

Japan [36] 26/97 4 1 12, 2474-2496del23 bp

V728GfsX733

Japan [26] II-3-D 23 1, 5 >7 kb deleted, incl.exons 5–8

Germany [41] Unnamed 39 1, 3, 4, 6 del658 bp, ins49 bp,removing exon 5

aIndividual numbering is according to the respective reportbPhenotypic features are as follows: 1 LMW proteinuria, 3 nephrocalcinosis and/or stones, 4 GFR 30–80 ml/min per 1.73 m2, 5 GFR<30 ml/min per 1.73 m2 or transplantation, 6 ricketscNucleotide numbering is according to the cDNA (GenBank acc. no. NM_000084) with the A of the start-ATG at nt. 292

1247

in the two patients with stage 2 chronic renal disease,diminished glomerular filtration rate might have maskedhypercalciuria. However, although hypercalciuria wasfrequently observed in the series from Europe and NorthAmerica, 50% of these patients did still have hypercalciuriain the setting of mild to moderate renal failure (Fig. 1).

Comparing the relative frequency of hypercalciuria inour CLCN5-positive patients with the smaller groupwithout a mutation, no significant difference was found.Therefore, the presence or absence of hypercalciuria doesnot appear to be helpful in distinguishing CLCN5-positiveand CLCN5-negative patients. Of note, the proportion ofCLCN5-negative patients in this series (16 of 50) wascomparable with published series on clinical X-linkedrecessive nephrolithiasis in which also CLCN5-negativepatients were reported (19 of 54, Table 2).

Both in the present series and in the literature [31], thereis no genotype-phenotype correlation for the presence ofhypercalciuria, as for almost half of the mutations (Table 3)both individuals with and without hypercalciuria werereported even within the same family. This fits well withthe current concept of hypercalciuria associated withCLCN5 mutations [48]. Also, functional studies on mutantClC-5 channels provided no clue for a mutation-dependentphenotype as complete loss of ClC-5 function (C221R,M571IfsX528) as well as only a reduction of ClC-5 activity(R516W) were associated with both the hyper- and thenormocalciuric phenotypes [17]. Looking at intrafamilialvariation in calcium excretion, we also found discordantfindings in three of the seven pedigrees with at least onepatient having normal calcium excretion pointing todifferences in genetic background, environment, diet, orunidentified modifier gene(s) [14, 23].

Nephrocalcinosis and/or stone disease are almost invari-ably present in Dent disease [7] and distinguish it from otherforms of renal Fanconi syndrome [2]. Kidney stones are

composed of calcium phosphate, calcium oxalate, or acombination of both [2, 12]. Urinary excretion of citrate andoxalate are normal in most patients [2, 12, 50], as is urinaryacidification before patients develop nephrocalcinosis orrenal insufficiency [12]. Therefore, hypercalciuria is classi-cally regarded as the most significant risk factor forcalcinosis and stone formation [12]. This prompted Raya etal. [51] to compare thiazide and/or amiloride treatment ofhypercalciuria between patients with Dent disease andidiopathic hypercalciuria. They found a comparable reduc-tion in calcium excretion in both groups with chlorthalidonebut a less impressive reduction in calcium-oxalate super-saturation because of hypocitraturia induced by thiazides.Based on these findings, the authors recommended thiazidetherapy in Dent patients with recurrent nephrolithiasis inanalogy to idiopathic hypercalciuria.

Still, the data presented herein show that calciumexcretion is normal in a considerable subgroup of CLCN5-positive patients with documented nephrolithiasis/nephro-calcinosis. To solve this paradox, several mechanisms maybe considered: first, calcium excretion may fluctuate and/orcalcium excretion may have been underestimated due toincomplete urine sampling. This is rather unlikely, however,because normal calcium excretion was documented onseveral occasions in our patients. Second, although withinnormal limits, citrate excretion in Dent patients is low whencompared with other forms of renal Fanconi syndrome notprone to develop nephrolithiasis, thereby suggesting arelative deficiency especially when calcium excretion iselevated [2]. In a recent study, a delay in the progression ofrenal insufficiency on a high citrate diet was observed inClC-5 knockout mice [52]. Third, ClC-5 is expressed inproton-secreting α-intercalated cells in the collecting duct.As acid secretion in these cells occurs via an apical H+-ATPase, ClC-5 might play a role in the endocytic insertionand removal of these proton pumps [53]. However, urinary

0

10

20

30

40

50

60

70

1,2, 3

,41 ,2

, 31 ,2

, 4 1 ,21, 3

,4 1 ,3 1 ,4 1

Phenotype

rela

tive

fre

qu

ency

(%

)

Eur ope/ NA

Japan

Pres en t s t udy

Fig. 1 Relative frequency(+upper 95% confidence inter-val) of the four principal Dentsymptoms observed in patientswith CLCN5 mutations. Thediagram shows the clinical phe-notypes reported in publishedseries from Europe/NorthAmerica (NA), Japan, and ourown data with symptoms indi-cated as follows: 1 LMWproteinuria, 2 hypercalciuria,3 nephrocalcinosis and/orstones, 4 renal failure

1248

acidification is normal in Dent patients who retain normalrenal function [12].

The most likely explanation comes from recent reportsthat the collecting duct is able to clear nucleated crystalsadhering to apical cell surface by endocytosis [54]. ClC-5seems to play a prominent role in this process as Sayer et al.[55] observed adhesion of larger crystal agglomerates toantisense ClC-5-transfected murine collecting duct cellscompared to controls. These calcium phosphate and calciumoxalate crystals potentially act as a nidus for subsequentstone formation. In the same cell model, internalization oflectin as well as calcium oxalate and calcium phosphatecrystals was also impaired [56], and most recently, theseauthors could demonstrate in their cell line amarked increasein plasmamembrane annexinA2, a crystal-bindingmolecule[57]. Although the precise mechanism underlying thelithogenic properties of ClC-5-deficient collecting ductcells is not yet fully understood, it offers an attractiveexplanation for the presence of nephrocalcinosis and neph-rolithiasis in patients with CLCN5 mutations who do nothave hypercalciuria.

In summary, the clinical spectrum of CLCN5 mutationscomprises a considerable number of patients with anincomplete phenotype, with incidental patients having onlyLMW proteinuria. Normal calcium excretion does notexclude an “X-linked hypercalciuric nephrolithiasis” syn-drome and should not prevent the clinician from consider-ing a CLCN5 mutation in the differential diagnosis. Thepresence of nephrocalcinosis/nephrolithiasis in patientswith persistently normal calcium excretion supports therecent concept of disturbed clearance of microcrystals inCLCN5 mutations.

Acknowledgements The authors express their sincere appreciationto the patients and their families for their participation in this study.Pia Uerdingen is acknowledged for excellent technical assistance andSigrid Bruinsma for secretarial assistance with the preparation of themanuscript.

References

1. Dent CE, Friedman M (1964) Hypercalciuric rickets associatedwith renal tubular damage. Arch Dis Child 39:240–249

2. Wrong OM, Norden AGW, Feest TG (1994) Dent’s disease; afamilial proximal renal tubular syndrome with low-molecular-weight proteinuria, hypercalciuria, nephrocalcinosis, metabolicbone disease, progressive renal failure and a marked malepredominance. QJM 87:473–493

3. Fisher SE, Bakel IV, Lloyd SE, Pearce SH, Thakker RV, CraigIW (1995) Cloning and characterization of CLCN5, the humankidney chloride channel gene implicated in Dent disease (an X-linked hereditary nephrolithiasis). Genomics 29:598–606

4. Scheel O, Zdebik AA, Lourdel S, Jentsch TJ (2005) Voltage-dependent electrogenic chloride/proton exchange by endosomalCLC proteins. Nature 436:424–427

5. Picollo A, Pusch M (2005) Chloride/proton antiporter activityof mammalian CLC proteins ClC-4 and ClC-5. Nature436:420–423

6. Thakker RV (2000) Pathogenesis of Dent’s disease and relatedsyndromes of X-linked nephrolithiasis. Kidney Int 57:787–793

7. Bonnardeaux A, Bichet DG (2000) Inherited disorders of therenal tubule. In: Brenner BM (ed) The kidney, 6th edn.Saunders, Philadelphia, pp 1656–1659

8. Lloyd SE, Pearce SHS, Fisher SE, Steinmeyer K, SchwappachB, Scheinman SJ, Harding B, Bolino A, Devoto M, Goodyer P,Rigden SPA, Wrong O, Jentsch TJ, Craig IW, Thakker RV(1996) A common molecular basis for three inherited kidneystone diseases. Nature 379:445–449

9. Igarashi T, Hayakawa H, Shigara H, Kawato H, Yan K,Kawaguchi H, Yamanaka T, Tsuchida S, Akagi K (1995)Hypercalciuria and nephrocalcinosis in patients with idiopathiclow-molecular-weight proteinuria in Japan: is the diseaseidentical to Dent’s disease in the United Kingdom? Nephron69:242–247

10. Igarashi T, Günther W, Sekine T, Inatomi J, Shiraga H,Takahashi S, Suzuki J, Tsuru N, Yanagihara T, Shimazu M,Jentsch TJ, Thakker RV (1998) Functional characterization ofrenal chloride channel, CLCN5, mutations associated withDent’s Japan disease. Kidney Int 54:1850–1856

11. Nakazato H, Hattori S, Furuse A, Kawano T, Karashima S,Tsuruta M, Yoshimuta J, Endo F, Matsuda I (1997) Mutation inthe CLCN5 gene in Japanese patients with familial idiopathiclow molecular weight proteinuria. Kidney Int 52:895–900

12. Scheinman SJ (1998) X-linked hypercalciuric nephrolithiasis:clinical syndromes and chloride channel mutations. Kidney Int53:3–17

13. Devuyst O, Christie PT, Courtoy PJ, Beauwens R, Thakker RV(1999) Intra-renal and subcellular distribution of the humanchloride channel, CLC-5, reveals a pathophysiological basis forDent’s disease. Hum Mol Genet 8:247–257

14. Günther W, Piwon N, Jentsch TJ (2003) The ClC-5 chloridechannel knock-out mouse—an animal model for Dent’s disease.Pflugers Arch 445:456–462

15. Piwon N, Günther W, Schwake M (2000) ClC-5 Cl-channeldisruption impairs endocytosis in a mouse model for Dent’sdisease. Nature 408:369–373

16. Ludwig M, Beck A, Wickert L, Bolkenius U, Tittel B, HinkelK, Bidlingmaier F (1998) Female pseudohermaphroditismassociated with a novel homozygous G-to-A (V370-to-M)substitution in the P-450 aromatase gene. J Pediatr EndocrinolMetab 11:657–664

17. Ludwig M, Doroszewicz J, Seyberth HW, Bökenkamp A,Balluch B, Nuutinen M, Utsch B, Waldegger S (2005)Functional evaluation of Dent’s disease-causing mutations:implications for ClC-5 channel trafficking and internalization.Hum Genet 117:228–237

18. LudwigM,Waldegger S, NuutinenM, Bökenkamp A, ReissingerA, Steckelbroeck S,Utsch B (2003) Four additionalCLCN5 exonsencode a widely expressed novel long CLC-5 isoform but fail toexplain Dent’s phenotype in patients without mutations in theshort variant. Kidney Blood Press Res 26:176–184

19. Besbas N, Ozaltin F, Jeck N, Seyberth H, Ludwig M (2005)CLCN5 mutation (R347X) associated with hypokalaemicmetabolic alkalosis in a Turkish child: an unusual presentationof Dent’s disease. Nephrol Dial Transplant 20:1476–1479

20. Hoopes RR Jr, Hueber PA, Reid RJ Jr, Braden GL, GoodyerPR, Melnyk AR, Midgley JP, Moel DI, Neu AM, VanWhy SK,Scheinman SJ (1998) CLCN5 chloride channel mutation in sixnew North American families with X-linked nephrolithiasis.Kidney Int 54:698–705

21. Hoopes RR Jr, Raja KM, Koich A, Hueber P, Reid R, Knohl SJ,Scheinman SJ (2004) Evidence for genetic heterogeneity inDent’s disease. Kidney Int 65:1615–1620

22. Oudet C, Martin-Coignard D, Pannetier S, Praud E, ChampionG, Hanauer A (1997) A second family with XLRH displays themutation S244L in the CLCN5 gene. Hum Genet 99:781–784

23. Kelleher CL, Buckalew VM, Frederickson ED, Rhodes DJ,Conner DA, Seidman JG, Seidman CE (1998) CLCN5 mutationSer244Leu is associated with X-linked renal failure without X-linked recessive hypophosphatemic rickets. Kidney Int 53:31–37

1249

24. Friedrich T, Breiderhoff T, Jentsch TJ (1999) Mutationalanalysis demonstrates that ClC-4 and ClC-5 directly mediateplasma membrane currents. J Biol Chem 274:896–902

25. Akuta N, Lloyd SE, Igarashi T, Shiraga H, Matsuyama T,Yokoro S, Cox JPD, Thakker RV (1997) Mutations of CLCN5in Japanese children with idiopathic low molecular weightproteinuria, hypercalciuria and nephrocalcinosis. Kidney Int52:911–916

26. Morimoto T, Uchida S, Sakamoto H, Kondo Y, Hanamizu H,Fukui M, Tomino Y, Nagano N, Sasaki S, Marumo F (1998)Mutations in CLCN5 chloride channel in Japanese patients withlow molecular weight proteinuria. J Am Soc Nephrol 9:811–818

27. Takemura T, Hino S, Ikeda M, Okada M, Igarashi T, Inatomi J,Yoshioka K (2001) Identification of two novel mutations in theCLCN5 gene in Japanese patients with familial idiopathic lowmolecular weight proteinuria (Japanese Dent’s disease). Am JKidney Dis 37:138–143

28. Cheong HI, Lee JW, Zheng SH, Lee JH, Kang JH, Kang HG, HaIS, Lee SJ, Choi Y (2005) Phenotype and genotype of Dent’sdisease in three Korean boys. Pediatr Nephrol 20:455–459

29. National Kidney Foundation (2002) K/DOQI clinical practiceguidelines for chronic kidney disease: evaluation, classificationand stratification. Am J Kidney Dis 39:S1–S266

30. Lloyd SE, Pearce SHS, Günther W, Kawaguchi H, Igarashi T,Jentsch TJ, Thakker RV (1997) Idiopathic low molecularweight proteinuria associated with hypercalciuric nephrocalci-nosis in Japanese children is due to mutations of the renalchloride channel (CLCN5). J Clin Invest 99:967–974

31. Lloyd SE, Günther W, Pearce SHS, Thomson A, Bianchi ML,Bosio M, Craig IW, Fisher SE, Scheinman SJ, Wrong O, JentschTJ, Thakker RV (1997) Characterisation of renal chloridechannel, CLCN5, mutations in hypercalciuric nephrolithiasis(kidney stones) disorders. Hum Mol Genet 6:1233–1239

32. Langlois V, Bernard C, Scheinman SJ, Thakker RV, Cox JPD,Goodyer PR (1998) Clinical features of X-linked nephrolith-iasis in childhood. Pediatr Nephrol 12:625–629

33. Nakazato H, Yoshimuta J, Karashima S, Matsumoto S, Endo F,Matsuda I, Hattori S (1999) Chloride channel CLCN5mutations in Japanese children with familial idiopathic lowmolecular weight proteinuria. Kidney Int 55:63–70

34. Bosio M, Bianchi ML, Lloyd SE, Thakker RV (1999) Afamilial syndrome due to Arg648Stop mutation in the X-linkedrenal chloride channel gene. Pediatr Nephrol 13:278–283

35. Cox JPD, Yamamoto K, Christie PT, Wooding C, Feest T,Flinter FA, Goodyer PR, Leumann E, Neuhaus T, Reid C,Williams PF, Wrong O, Thakker RV (1999) Renal chloridechannel, CLCN5, mutation in Dent’s disease. J Bone Miner Res14:1536–1542

36. Yamamoto K, Cox JPDT, Friedrich T, Christie PT, Bald M,Houtman PN, Lapsley MJ, Patzer L, Tsimaratos M, Van’t HoffWG, Yamaoka K, Jentsch TJ, Thakker RV (2000) Character-ization of renal chloride channel (CLCN5) mutations in Dent’sdisease. J Am Soc Nephrol 11:1460–1468

37. Igarashi T, Inatomi J, Ohara T, Kuwahara T, Shimadzu M,Thakker RV (2000) Clinical and genetic studies of CLCN5mutations in Japanese families with Dent’s disease. Kidney Int58:520–527

38. Carballo-Trujillo I, Garcia-Nieto V, Moya-Angeler FJ, Antón-Gamero M, Loris C, Méndez-Alvarez S, Claverie-Martin F(2003) Novel truncating mutations in the CLC-5 chloridechannel gene in patients with Dent’s disease. Nephrol DialTransplant 18:717–723

39. Claverie-Martin F, González-Acosta H, Flores C, Antón-Gamero M, García-Nieto V (2003) De novo insertion of anAlu sequence in the coding region of the CLCN5 gene results inDent’s disease. Hum Genet 113:480–485

40. Matsuyama T, Awazu M, Oikawa T, Inatomi J, Sekine T,Igarashi T (2004) Molecular and clinical studies of Dent’sdisease in Japan: biochemical examination and renal ultraso-nography do not predict carrier state. Clin Nephrol 61:231–237

41. Brakemeier S, Si H, Gollasch M, Höffler D, Buhl M, Köhler R,Hoyer J, Eichler I (2004) Dent’s disease: identification of anovel mutation in the renal chloride channel CLCN5. ClinNephrol 62:387–390

42. Yanagida H, Ikeoka M, Kuwajima H, Wada N, Tabata N,Sugimoto K, Okada M, Takemura T (2004) A boy withJapanese Dent’s disease exhibiting abnormal calcium metabo-lism and osseous disorder of the spine: defective megalinexpression at the brushborder of renal proximal tubules. ClinNephrol 62:306–312

43. Luyckx VA, Leclercq B, Dowland LK, Yu ASL (1999) Diet-dependent hypercalciuria in transgenic mice with reduced CLC-5 chloride channel expression. Proc Natl Acad Sci USA96:12174–12179

44. Wang SS, Devuyst O, Courtoy PJ, Wang XT, Wang H, Wang Y,Thakker RV, Guggino S, Guggino WB (2000) Mice lackingrenal chloride channel, CLC-5, are a model for Dent’s disease, anephrolithiasis disorder associated with defective receptor-mediated endocytosis. Hum Mol Genet 9:2937–2945

45. Silva IV, Cebotaru V, Wang H, Wang XT, Wang SS, Guo G,Devuyst O, Thakker RV, Guggino WB, Guggino SE (2003)The ClC-5 knockout mouse model of Dent’s disease has renalhypercalciuria and increased bone turnover. J Bone Miner Res18:615–623

46. Reinhart SC, Norden AGW, Lapsley M, Thakker RV, Pang J,Moses AM, Frymoyer PA, Favus MJ, Hoepner JA, ScheinmanSJ (1995) Characterization of carrier females and affectedmales with X-linked recessive nephrolithiasis. J Am SocNephrol 5:1451–1461

47. Foreman JW (2004) Cystinosis and Fanconi syndrome. In: AvnerED, HarmonWE, Niaudet P (eds) Pediatric nephrology, 5th edn.Lippincott Williams & Wilkins, Philadelphia, pp 789–806

48. Yu ASL (2001) Role of ClC-5 in the pathogenesis ofhypercalciuria: recent insights from transgenic mouse models.Curr Opin Nephrol Hypertens 10:415–420

49. Hoopes RR Jr, Shrimpton AE, Knohl SJ, Hueber P, Hoppe B,Matyus J, Simckes A, Tasic V, Toenshoff B, Suchy SF,Nussbaum RL, Scheinman SJ (2005) Dent disease withmutations in OCRL1. Am J Hum Genet 76:260–267

50. Barbano G, Ghiggeri GM, Fierro S, Tosetto E, Anglani F,Perfumo F (2005) Isolated low molecular weight proteinuria aspresenting syndrome of Dent’s disease in children: relationshipbetween calciuria, calcium (Ca) and phosphorus (P) intake,PTH and bone homeostasis (abstract). Pediatr Nephrol 20:C62

51. Raja KM, Schurman S, D’Mello RG, Blowey D, Goodyer P,Van Why S, Ploutz-Snyder RJ, Asplin J, Scheinman SJ (2002)Responsiveness of hypercalciuria to thiazide in Dent’s disease.J Am Soc Nephrol 13:2938–2944

52. Cebotaru V, Kaul S, Devuyst O, Cai H, Racusen L, GugginoWB, Guggino SE (2005) High citrate diet delays progression ofrenal insufficiency in the ClC-5 knockout mouse model ofDent’s disease. Kidney Int 68:642–652

53. Sayer JA, Simmons NL (2002) Urinary stone formation: Dent’sdisease moves understanding forward. Exp Nephrol 10:176–181

54. Lieske JC, Norris R, Swift H, Toback FG (1997) Adhesion,internalization and metabolism of calcium oxalate monohydratecrystals by renal epithelial cells. Kidney Int 52:1291–1301

55. Sayer JA, Carr G, Pearce SH, Goodship TH, Simmons NL(2003) Disordered calcium crystal handling in antisense CLC-5-treated collecting duct cells. Biochem Biophys Res Commun300:305–310

56. Sayer JA, Carr G, Simmons NL (2004) Calcium phosphate andcalcium oxalate crystal handling is dependent upon CLC-5expression in mouse collecting duct cells. Biochim BiophysActa 1689:83–90

57. Carr G, Simmons NL, Sayer JA (2006) Disruption of clc-5leads to redistribution of annexin A2 and promotes calciumcrystal agglomeration in collecting duct epithelial cells. CellMol Life Sci 63:367–377

1250