Hypomagnesemia with secondary hypocalcemia is caused by ...

letter

© 2002 Nature Publishing Group http://genetics.nature.com

Hypomagnesemia with secondary hypocalcemia is caused by mutations in TRPM6, a new member of the TRPM gene family Karl P. Schlingmann1*, Stefanie Weber1*, Melanie Peters1, Lene Niemann Nejsum2, Helga Vitzthum3, Karin Klingel4, Markus Kratz5, Elie Haddad6, Ellinor Ristoff7, Dganit Dinour8, Maria Syrrou9, Søren Nielsen2, Martin Sassen1, Siegfried Waldegger1, Hannsjörg W. Seyberth1 & Martin Konrad1 * These authors contributed equally to this work.

Published online: 28 May 2002, DOI: 10.1038/ng889

Magnesium is an essential ion involved in many biochemical and physiological processes. Homeostasis of magnesium levels is tightly regulated and depends on the balance between intestinal absorption and renal excretion. However, little is known about specific proteins mediating transepithelial magnesium transport. Using a positional candidate gene approach, we identified mutations in TRPM6 (also known as CHAK2), encoding TRPM6, in autosomal-recessive hypomagnesemia with secondary hypocalcemia (HSH, OMIM 602014)1,2, previously mapped to chromosome 9q22 (ref. 3). The TRPM6 protein is a new member of the long transient receptor potential channel (TRPM) family4 and is highly similar to TRPM7 (also known as TRP-PLIK), a bifunctional protein that combines calcium- and magnesium-permeable cation channel properties with protein kinase activity5−7. TRPM6 is expressed in intestinal epithelia and kidney tubules. These findings indicate that TRPM6 is crucial for magnesium homeostasis and implicate a TRPM family member in human disease.

The disease HSH is characterized by very low magnesium and low calcium serum levels. Affected individuals show neurologic symptoms of hypomagnesemic hypocalcemia, including seizures and muscle spasms, during infancy. Untreated, the disorder may be fatal or may result in neurological damage. Hypocalcemia is secondary to parathyroid failure resulting from

magnesium deficiency8. Normocalcemia and relief of clinical symptoms can be assured by oral administration of high doses of magnesium2. Pathophysiology of HSH is largely unknown, but physiological studies indicate a primary defect in intestinal magnesium transport9. In some individuals, an additional renal magnesium leak, caused by altered magnesium reabsorption in the distal convoluted tubule (DCT), was suspected10,11. A gene locus (HOMG1) for HSH was previously mapped to chromosome 9q22 (ref. 3) and further refined to a critical interval between markers D9S1115 and D9S175 (ref. 12). We studied five families with typical HSH (Table 1). Extended disease haplotypes in the three consanguinous families were compatible with linkage, but did not allow further narrowing of the critical interval (Fig. 1). Flanking markers D9S1115 and D9S175 define a physical interval of approximately 1.7 Mb (Fig. 2a). A search for candidate genes within this region revealed five genes (Fig. 2a) and various expressed sequence tag (EST) clones (see Methods). Focusing on sequences expressed in intestine, we identified two overlapping human ESTs with identity to parts of TRPM6, recently cloned from kidney cDNA4. The gene TRPM6 encodes a putative ion channel that is highly similar to the transient receptor potential (TRP) channel family. TRP ion channels are characterized by six transmembrane segments, a conserved pore-forming region and a Pro-Pro-Pro motif fol-

Table 1 • Clinical characteristics and TRPM6 mutations in individuals with HSH Age at onset

First symptoms

Initial Mg2+ (mM)

FE Mg2+ (%)

Origin

Parental consanguinity

Nucleotide exchange

Exon

F1.1

9 wk

seizures

0.21

4.8

Turkey

yes

1769C→G

15

Ser590X

F2.1

3 wk

seizures

0.08

2.8

Turkey

yes

2667+1G→A

5′ss19

exon skipping

F3.1

4 mo

seizures

0.10

4.0

Sweden

no

[3537–1G→A] + [422C→T]

3′ss25 4

exon skipping Ser141Leu

F4.1

4 wk

seizures

0.41

high

Israel

no

[1280delA] + [3779−91del]

10 25

His427fsX429 Glu1260fsX1283

F5.1

5 wk

seizures

0.17

4.5

Albania

yes

2207delG

16

Arg736fsX737

F5.2

5 wk

seizures

0.22

2.6

Albania

yes

2207delG

16

Arg736fsX737

Patient

Consequence

FE Mg2+, fractional excretion of magnesium; fs, frameshift; del, deletion. Nucleotide positions correspond to the coding sequence as deposited in GenBank.

1Department of Pediatrics, Philipps University of Marburg, Deutschhausstrasse 12, D-35037 Marburg, Germany. 2Water and Salt Research Center, Institute of Anatomy, University of Aarhus, Denmark. 3Department of Physiology I, University of Regensburg, Germany. 4Department of Molecular Pathology, University Hospital, Tübingen, Germany. 5University Children’s Hospital, Mannheim, Germany. 6Robert Debré Hospital, Paris, France. 7Department of Pediatrics, Karolinska Institute, Huddinge University, Stockholm, Sweden. 8Department of Nephrology, The Chaim Sheba Medical Center, Ramat-Gan, Israel. 9Laboratory of General Biology-Genetics, University of Ioannina, Greece. Correspondence should be addressed to M.K. (e-mail: [email protected]).

166

nature genetics • volume 31 • june 2002

letter


Fig. 1 Haplotype analysis of families F1, F2 and F5. Filled symbols indicate affected individuals. A slash mark indicates deceased individuals, and double lines indicate consanguinity. Genotyping data and schematic segregating haplotype bars for chromosome 9q22 are shown below the symbol for each individual. Black bars indicate the haplotype segregating with the gene underlying HSH; white bars show the wildtype haplotypes.

F1

8 2 2 7 2 2 1

D9S1799 D9S1115 8580-1540 23938-470 23938-194 D9S175 D9S284

7 4 1 5 4 1 1

2 5 2 2 4 1 1

2 5 2 2 4 1 1


a

7 1 2 3 2 1 1

3 1 2 4 2 1 2

9 1 2 3 1 3 1

6 3 2 3 1 5 1

7 7 1 1 2 2 3 3 2 2 1 1 1 1 F5.1

6 6 3 3 2 2 3 3 1 1 5 5 1 1 F2.1

7 1 2 3 2 1 1

4 2 2 5 2 4 2

7 7 1 1 2 2 3 3 2 2 1 1 1 1 F5.2

represented in the genomic sequence database (TAG instead of ATG), we carried out 5′ RACE–PCR on cDNA from human small intestine and kidney. PCR fragments from both tissues contained an upstream in-frame methionine within a suitable Kozak sequence. Comparison with genomic clones indicated an additional exon (termed exon 1A), located 29 kb upstream of exon 1 (Fig. 2d), encoding the 5′ untranslated region and ten additional amino acids. Thus, the full-length TRPM6 encodes a protein of 2,022 amino acids encoded by 39 exons. Single-strand conformational polymorphism (SSCP) analysis and subsequent direct sequencing revealed seven different

94

70

5 40 11 15 S1 0D9 858

9

79

S1

D9

6 3 2 3 1 5 1

1 1 2 3 1 1 1

F1.2

lowing the last transmem2 2 brane segment13. The TRPM6 5 5 2 2 protein belongs to the TRPM 2 2 4 4 subfamily, whose members 1 1 1 1 share a long amino-terminal F1.1 domain of unknown function. This protein shows highest similarity with TRPM7 (52% overall identity; Fig. 2b), recently identified as a magnesium- and calcium-permeable ion channel5. In addition to their ion channel domain, TRPM6 and TRPM7 have a carboxy-terminal region with sequence similarity to the atypical α-kinase family (Fig. 2c)7. The high similarity of TRPM6 to TRPM7, along with its partial identity with ESTs cloned from intestine, prompted further examination of TRPM6 in HSH. Aligning TRPM6 cDNA with genomic clones revealed 38 exons spanning 134 kb (Fig. 2d). As the translation start site indicated by the TRPM6 mRNA sequence originally reported was not

Fig. 2 Characterization of TRPM6. a, Physical map of the HOMG1 critical region (filled horizontal bar) between previously identified polymorphic markers D9S1115 and D9S175. Human genome sequence segments (white horizontal boxes) are derived from the NCBI database. The position of polymorphic markers is marked by vertical bars. Arrows below indicate the location and orientation of known and putative genes annotated by the NCBI RefSeq project. cen, centromeric; tel, telomeric. b, Phylogenetic tree of the TRPM subfamily (human fulllength protein sequences) showing the close vicinity of TRPM6 to TRPM7. c, Proposed model of TRPM6, which shows the typical properties of the TRPM family: a long conserved intracellular N terminus and six transmembrane domains with a pore-forming loop between the last two segments. The intracellular C terminus of TRPM6 is highly homologous to the α-kinase family (adapted from ref. 23). d, Genomic organization and mRNA structure of TRPM6. The gene consists of 39 exons (white boxes) spanning approximately 163 kb of genomic DNA including an additional exon (1A) and the 5′ untranslated region (UTR) located approximately 29 kb upstream of the original start site were identified. Remaining exon numbers are derived from alignment of TRPM6 mRNA with genomic sequence, which showed the absence of the published start methionine in the genomic sequence. Functional domains were deduced from the TRPM7 model as previously described5. The mutations detected in HSH, along with the corresponding exon number and their functional consequences for the protein, are indicated. fs, frameshift.

F5

F2

75 84 S1 S2 D9

-1

-4

38

38

D9

9 23

9 23

cen

tel NT_008580

NT_023953

NT_023938 RORB

FLJ10110

OSTF1

FLJ20559

TRPM6

b

NT_029358

c extracellular

TRPM7 TRPM6 TRPM1

intracellular

NH 2

TRPM2

COOH

TRPM6

TRPM4

α kinase domain

TRPM5

d 5' UTR . . . aaag ATG AAA GAA CAA CCT GTC TTG GAG CGC TTG CAG gtaagctc

.....

29 kb

.....

ctttttag TCC CAG AAA . . . genomic sequence

Q V E M K E P L R L Q K Q S 5' UTR . . . aaag ATG AAA GAA CAA CCT GTC TTG GAG CGC TTG CAG TCC CAG AAA ATG TCC CAG AAA M

loss of ss

TRPM6 1A

1

15 16

10

4

S

Q

..... .....

5' RACE fragment AF350881

K

loss of ss

19

25

38

3' UTR 3'

5' fs

1

fs

Leu429X Ser590X Leu737X

Ser141Leu

TRPM6

fs

2

3

conserved N terminus

Leu1283X

α

4 transmembrane

coiled coil

kinase

TRP domain

167

products in intestine and kidney (Fig. 3b,c). We also analyzed segmental localization of TRPM6 in microdissected nephrons (Fig. 3d), which showed the strongest signal in the DCT, the main site of active transcellular magnesium reabsorption along the nephron15. We obtained weak signals in proximal tubule and collecting duct. Expression of TRPM6 was also assessed by in situ hybridization in various human tissues. We observed TRPM6 mRNA in colon epithelial cells (Fig. 4a), duodenum (Fig. 4b), jejunum and ileum. We obtained no specific signals using sense probes (Fig. 4c). We also detected TRPM6 mRNA in single distal renal tubule cells (Fig. 4d). No signals were observed in stomach, lung and heart (data not shown). Together with the disease phenotype, these findings indicate that TRPM6 is essential in epithelial magnesium absorption. In vertebrates, intestinal magnesium absorption is related to luminal magnesium in a curvilinear manner (Fig. 5)16. The hyperbolic curve reflects a saturable active transcellular transport at low intraluminal magnesium concentrations mediated by an electrogenic apical entry and an active basolateral transport. The linear function at higher concentrations reflects a passive paracellular magnesium absorption. Alternatively, F3 this absorption pattern might result from a progressive exon 4 intron 24 exon 25 ‘tightening’ of the junction complex induced by increasA T C T C A G T C A C T A G G G T T ing luminal magnesium17. The observation that individuals with HSH achieve normal serum magnesium levels by high oral magnesium intakealthough they show impaired intestinal magnesium absorptionsupports the theory of two independent pathways. Our father mother results suggest that TRPM6 represents a molecular comA C T AN G G T T A T C T N A G T C ponent of the active transcellular pathway and that, in HSH, high oral magnesium doses are sufficient to overcome the phenotype of mutant TRPM6 by increasing paracellular magnesium absorption. As we also identified TRPM6 expression in kidney, an A T C T N A G T C A C T A N G G T T additional renal phenotype in HSH could be expected. Urinary magnesium excretion rates in individuals with HSH have been examined10,11. With respect to their low serum magnesium levels, the individuals studied here A T C T CA G T C a c t ag G G T T showed inappropriately high fractional magnesium T a excretion rates (Table 1), indicating an additional role of Ile140 Ser141Leu Val142 Val1180

control parent G G A G N T G A G

A A G T N A A A A

patient A A G T G A A A A

G G A G A T G A G

A A G TC A A A A G

G G AG g t g a g a

+

_

jej

2O

H RT

du

b

Glu889

od en u un m um ile um co lon liv er

Lys589 Ser590X Lys591

+

_

+

_

+

_

+

_

c +

sm

G G A G G T G A G

A A G T C A A A A

y

F2 exon 19 intron 19

ne

F1 exon 15

kid

a

_

TRPM6

TRPM6

415 bp _

415 bp _

β _actin

G3PDH mucosa

dT

L/a TL mT AL cT AL DC T CT /C CD OM CD IM CD

tissue

d TRPM6 415 bp _

β _actin microdissected rat nephron segments

168

.i co ntes lon tin e H 2O

mutations in TRPM6 (Table 1 and Fig. 2d). Examples of sequence analysis are given in Fig. 3a. We detected a homozygous in-frame stop codon at Ser590, which truncates the protein prior to the first transmembrane domain in family F1. In family F2, the affected individual bears a homozygous donor splice-site mutation adjacent to exon 19. One heterozygous mutation in family F3 affects the acceptor splice site of exon 25. Both dinucleotide splice-site sequences (GT, AG) are highly conserved and crucial for mRNA splicing14. The second mutation in family F3 is a nonconservative amino-acid exchange of a highly conserved serine residue (Ser141Leu) within the first N-terminal region. The affected individual of family F4 is compound heterozygous with respect to two frameshift mutations both truncating substantial parts of the TRPM6 protein (Fig. 2d and Table 1). In family F5, we identified a deletion of 1 bp that results in protein truncation prior to the transmembrane domains (Table 1). All mutations cosegregate with the phenotype and were not detected in 102 control chromosomes. To confirm intestinal expression of TRPM6, we carried out RT–PCR on various rat tissues. We obtained PCR amplification

glo m PC T PS T dT L


letter

Fig. 3 Sequence analysis and tissue distribution of TRPM6. a, Mutation analysis of TRPM6 by direct sequencing. Genomic sequence analysis in a control individual (top), one parent (middle) and the affected individual (bottom) from families F1−F3. Vertical lines divide exonic from intronic sequence. Mutated nucleotides and resulting amino-acid changes are shown under the affected individual’s sequence. Bold letters indicate the mutated nucleotides. Individual F1.1 bears a homozygous stop mutation (X) at aa S590; individual F2.1 is homozygous with respect to a G→A exchange disrupting the donor splice site after exon 19. Individual F3.1 is compound heterozygous with respect to an amino-acid substitution (Ser141Leu) and an acceptor splice-site mutation at position −1 of exon 25. Consensus splice-site sequences are indicated in red. b, TRPM6 mRNA expression in rat intestine and kidney. We detected a TRPM6 fragment of 415 bp along the entire intestinal tract and kidney. No signal was detected in liver. The presence of cDNA in all samples was confirmed by detection of β-actin mRNA. c, Detection of TRPM6 mRNA in mucosal cells from rat intestine. TRPM6 expression in intestinal epithelia could be confirmed after mechanical removal of submucosal layer and muscle tissue in rat small (sm) intestine and colon. d, TRPM6 RT−PCR on microdissected rat nephron segments. We detected the strongest signal in DCT. Weak TRPM6 mRNA was present in proximal convoluted tubule (PCT), connecting tubule/cortical collecting duct (CT/CCD), outer medullary collecting duct (OMCD) and inner medullary collecting duct (IMCD). No TRPM6 expression was detected in glomeruli (glom), proximal straight tubule (PST), descending thin limb (dTL), ascending TL (aTL), medullary thick ascending limb (mTAL) and cortical TAL (cTAL). Parallel amplifications of marker genes served as controls for segment specificity of the preparation (data not shown).


letter

impaired renal magnesium reabsorption in HSH11. This is confirmed by the characterization of a considerable renal leak of magnesium in HSH patients in an accompanying report18. This study highlights new aspects of the molecular basis of magnesium transport in humans. Mutations in CLDN16 were previously identified19 in familial hypomagnesemia with hypercalciuria and nephrocalcinosis (OMIM 248250). The gene CLDN16, a newly discovered member of the claudin multigene family, mediates paracellular magnesium reabsorption through tight junctions in renal tubular epithelia. In isolated renal magnesium loss (OMIM 154020), a dominant-negative mutation was reported in FXYD2 that results in a trafficking defect of the γsubunit of the Na+/K+-ATPase20. The identification of TRPM6 mutations in HSH suggests that TRPM6 is the first component of intestinal magnesium absorption identified at the molecular level. Moreover, the involvement of TRPM6 in magnesium absorption in intestine and kidney raises the question of whether its homolog TRPM7, which is expressed in many tissues, is also involved in cellular magnesium homeostasis. These findings emphasize the growing functional diversity of TRPM family members, which have previously been shown to be important in cellular processes such as cell survival, progression of tumors and intracellular signaling5,21–23.

Methods Subjects. Diagnosis of HSH in the five families studied was based on a characteristic history of neurologic symptoms, including tetany, muscle spasms and seizures due to hypomagnesemic hypocalcemia during the first six months of life. Families F4 and F5 have been described previously10,24. The most prominent biochemical findings were hypomagnesemia (< 0.4 mmol L−1), hypocalcemia and suppressed parathyroid hormone levels. In all affected individuals, hypomagnesemia could be controlled by high oral magnesium supplements, which is typical of HSH and contrasts the phenotype seen in individuals with exclusive renal magnesium loss. The first-grade cousin of individual F1.1 also has a clinical history typical of HSH; however, we could not obtain DNA from this individual. The study was carried out with the approval of the Ethics Committee of the Philipps University of Marburg, Germany. We obtained informed consent from all subjects. Genotyping. We isolated genomic DNA from peripheral blood by standard methods. We carried out PCR and haplotype analysis as described previously25. Microsatellite markers were generated by dinucleotide repeat searches in genomic clones mapping to the critical region or were derived from public databases. Physical map, candidate genes. We obtained genomic data for the construction of a physical map of the critical region by a search in genomic databases. We compared cDNA with genomic sequences using the BLAST program. Start codon prediction was carried out with the ATGpr program.

Fig. 5 Intestinal magnesium absorption. The curvilinear absorption of magnesium in human intestine (black line) is proposed to result from two transport mechanisms: (i) a transcellular transport, saturable at high luminal magnesium concentrations, which is of functional importance at low luminal magnesium concentrations (dotted line) and (ii) a paracellular passive transport (dashed line) linearly rising with elevated luminal magnesium concentrations. The observation that the defect in intestinal absorption in individuals with HSH may be compensated by high oral magnesium intake suggests that the paracellular pathway is intact. This, in turn, indicates that TRPM6 is involved in the transcellular active magnesium absorption pathway.


a

b

c

d

Five genes mapped to the critical interval (TRPM6, RORB, OSTF1 and the genes encoding FLJ20559 and FLJ10110). The gene RORB is an orphan nuclear receptor; RORB knockout mice show a phenotype (atypical behavior, blindness) completely different from that of HSH26. The gene OSTF1 encodes an intracellular protein produced by osteoclasts that indirectly induces osteoclast formation and bone resorption27. FLJ20559 is a hypothetical protein highly similar to uridine kinases, which are known to be involved in the salvage pathway of pyrimidine synthesis28; thus, the gene encoding this protein is probably not involved in HSH. FLJ10110 is a hypothetical protein of unknown function that seems to have a broad tissue distribution; we excluded mutations of the gene encoding this protein as the cause of HSH by direct sequencing in families F1–F3. RACE analysis. To obtain the complete TRPM6 cDNA, we carried out 5′ RACE of human small intestine and kidney marathon-ready cDNA (Clontech). Primer sequences are available upon request. Fragments were cloned into the pCR-II-TOPO vector (Invitrogen) and sequenced from both strands with M13 vector primers using an ABI Prism 310 sequencer (Applera). SSCP analysis and direct sequencing. We screened for TRPM6 mutations by SSCP analysis. Based on the sequence of the human gene, we used overlapping sets of primers to amplify the coding sequences of genomic DNA by PCR. We designed TRPM6 primers with Primer3 software; sequences are available upon request. Amplified products were separated on polyacrylamide gels by electrophoresis, and exons with conformational variants were directly sequenced using corresponding sequencing primers as described previously25. RNA extraction and RT−PCR. We extracted RNA by conventional methods. Reverse transcription of total RNA and mRNA was done using the Superscript First Strand Synthesis System for RT−PCR (Invitrogen). We carried out reverse-transcription negative control reactions on half of the extracted mRNA. PCR was carried out for 30 cycles. Primer sequences are available upon request. We included controls to confirm the presence of cDNA and carried out a PCR negative-control reaction.

net Mg 2+ absorption


Fig. 4 In situ hybridization analysis of TRPM6 expression. a,b, TRPM6 mRNA expression in the surface and crypt epithelial cells of the colon (a) and in villous epithelial cells of the duodenum (b). c, Control hybridization with a sense probe on consecutive tissue sections. d, In the kidney, TRPM6 mRNA expression is restricted to a few distal nephron segments, probably representing distal convoluted tubules.

d

bine

com

ular

cell

para

transcellular

Mg2+ intake

169

letter Microdissection of rat nephron segments. We obtained nephron segments by using a modified collagenase digestion protocol as previously described29. At least 11 mm of each segment were pooled for RNA extraction. After cDNA synthesis (as described above), we carried out PCR on cDNA samples corresponding to an initial tubule length of 1 mm derived from at least three different sets of nephron segments from different animals. Parallel amplifications of marker genes (SLC12A1, SLC12A3, AQP1, AQP4, PTHR1, SLC5A2) served as controls for segment-specificity of the preparation.

2.

3.

4.

5.


6.

In situ hybridization. Human tissue specimens from stomach, duodenum, jejunum, ileum, colon, lung and kidney were fixed in 4% paraformaldehyde, 0.1 M sodium phosphate buffer (pH 7.2) for 4 h and embedded in paraffin. Tissue sections (4 µm) were dewaxed and hybridized essentially as described30. The hybridization mixture contained either the 35S-labeled RNA antisense or sense control TRPM6 probe (500 ng ml−1) in buffer (10 mM Tris HCl, pH 7.4; 50% (vol/vol) deionized formamide; 600 mM NaCl; 1 mM EDTA; 0.02% polyvinylpyrrolidone; 0.02% Ficoll; 0.05% bovine serum albumin; 10% dextran sulfate; 10 mM dithiothreitol; denatured sonicated salmon sperm DNA at 200 µg ml−1 per rabbit liver tRNA at 100 µg ml−1). We carried out hybridization with RNA probes at 42 °C for 18 h. Slides were washed as described30, and then for 1 h at 55 °C in 2 × standard saline citrate. Nonhybridized single-stranded RNA probes were digested by RNAse A (20 µg ml−1) in 10 mM Tris HCl, pH 8.0; 0.5 M NaCl for 30 min at 37 °C. Tissue slide preparations were autoradiographed and stained with hematoxylin and eosin. Accession numbers. Genomic clones mapping to the critical interval: NT_008580, NT_023953, NT_023938, NT_029358; TRPM6, AF350881; EST clones representing parts of TRPM6: colon, AK000094; small intestine, AK026281. The sequence of the full-length human TRPM6 cDNA sequence was submitted to GenBank (AF 448232).

7.

8. 9.

10. 11. 12.

13. 14.

15. 16. 17. 18.

19.

Acknowledgments

We thank the patients and their families for participating in this study, U. Pechmann and P. Barth for excellent technical assistance, C. Antignac, R. Preisig-Müller, C. Derst and N. Jeck for helpful discussions and C. Loirat, D. Lotan, W. Scheurlen, A. Siamopoulou, S. Alfandaki, G. Celsi and A. Kernell for providing clinical data. S.W., H.W.S. and M.K. were supported by the Deutsche Forschungsgemeinschaft. S.W. was supported by the KempkesStiftung, University of Marburg. L.N.N. and S.N. were supported by the Danish National Research Foundation. Competing interests statement

20. 21. 22. 23. 24. 25. 26.

27.

The authors declare that they have no competing financial interests. 28.

Received 28 November 2001; accepted 4 April 2002.

29.

1.

30.

Paunier, L., Radde, I.C., Kooh, S.W., Conen, P.E. & Fraser, D. Primary hypomagnesemia with secondary hypocalcemia in an infant. Pediatrics 41, 385–402 (1968).

170

Shalev, H., Phillip, M., Galil, A., Carmi, R. & Landau, D. Clinical presentation and outcome in primary familial hypomagnesaemia. Arch. Dis. Child. 78, 127–130 (1998). Walder, R.Y. et al. Familial hypomagnesemia maps to chromosome 9q, not to the X chromosome: genetic linkage mapping and analysis of a balanced translocation breakpoint. Hum. Mol. Genet. 6, 1491–1497 (1997). Ryazanova, L.V., Pavur, K.S., Petrov, A.N., Dorovkov, M.V. & Ryazanov, A.G. Novel type of signaling molecules: protein kinases covalently linked with ion channels. Mol. Biol. 35, 271–283 (2001). Nadler, M.J. et al. LTRPC7 is a Mg-ATP-regulated divalent cation channel required for cell viability. Nature 411, 590–595 (2001). Runnels, L.W., Yue, L. & Clapham, D.E. TRP-PLIK, a bifunctional protein with kinase and ion channel activities. Science 291, 1043–1047 (2001). Ryazanov, A.G. et al. Identification of a new class of protein kinases represented by eukaryotic elongation factor-2 kinase. Proc. Natl Acad. Sci. USA 94, 4884–4889 (1997). Anast, C.S., Mohs, J.M., Kaplan, S.L. & Burns, T.W. Evidence for parathyroid failure in magnesium deficiency. Science 177, 606–608 (1972). Milla, P.J., Aggett, P.J., Wolff, O.H. & Harries, J.T. Studies in primary hypomagnesaemia: evidence for defective carrier-mediated small intestinal transport of magnesium. Gut 20, 1028–1033 (1979). Matzkin, H., Lotan, D. & Boichis, H. Primary hypomagnesemia with a probable double magnesium transport defect. Nephron 52, 83–86 (1989). Cole, D.E. & Quamme, G.A. Inherited disorders of renal magnesium handling. J. Am. Soc. Nephrol. 11, 1937–1947 (2000). Walder, R.Y. et al. Hypomagnesemia with secondary hypocalcemia (HSH): narrowing the disease region on chromosome 9. Am. J. Hum. Genet. 65, 451 (1999). Harteneck, C., Plant, T.D. & Schultz, G. From worm to man: three subfamilies of TRP channels. Trends Neurosci. 23, 159–166 (2000). Shapiro, M.B. & Senapathy, P. RNA splice junctions of different classes of eukaryotes: sequence statistics and functional implications in gene expression. Nucleic Acids Res. 15, 7155–7174 (1987). Quamme, G.A. Renal magnesium handling: new insights in understanding old problems. Kidney Int. 52, 1180–1195 (1997). Fine, K.D., Santa Ana, C.A., Porter, J.L. & Fordtran, J.S. Intestinal absorption of magnesium from food and supplements. J. Clin. Invest. 88, 396–402 (1991). Kayne, L.H. & Lee, D.B. Intestinal magnesium absorption. Miner. Electrolyte Metab. 19, 210–217 (1993). Walder, R.Y. et al. Mutation of TRPM6 causes familial hypomagnesemia with secondary hypocalcemia. Nature Genet. 31 (2002); advance online publication, 28 May 2002 (DOI:10.1038/ng901). Simon, D.B. et al. Paracellin-1, a renal tight junction protein required for paracellular Mg2+ resorption. Science 285, 103–106 (1999). Meij, I.C. et al. Dominant isolated renal magnesium loss is caused by misrouting of the Na(+),K(+)-ATPase γ-subunit. Nature Genet. 26, 265–266 (2000). Duncan, L.M. et al. Melastatin expression and prognosis in cutaneous malignant melanoma. J. Clin. Oncol. 19, 568–576 (2001). Perraud, A.L. et al. ADP-ribose gating of the calcium-permeable LTRPC2 channel revealed by Nudix motif homology. Nature 411, 595–599 (2001). Cahalan, M.D. Cell biology. Channels as enzymes. Nature 411, 542–543 (2001). Challa, A., Papaefstathiou, I., Lapatsanis, D. & Tsolas, O. Primary idiopathic hypomagnesemia in two female siblings. Acta Paediatr. 84, 1075–1078 (1995). Konrad, M. et al. Mutations in the chloride channel gene CLCNKB as a cause of classic Bartter syndrome. J. Am. Soc. Nephrol. 11, 1449–1459 (2000). Andre, E. et al. Disruption of retinoid-related orphan receptor β changes circadian behavior, causes retinal degeneration and leads to vacillans phenotype in mice. EMBO J. 17, 3867–3877 (1998). Reddy, S. et al. Isolation and characterization of a cDNA clone encoding a novel peptide (OSF) that enhances osteoclast formation and bone resorption. J. Cell Physiol. 177, 636–645 (1998). Koizumi, K. et al. Cloning and expression of uridine/cytidine kinase cDNA from human fibrosarcoma cells. Int. J. Mol. Med. 8, 273–278 (2001). Weber, S. et al. Primary gene structure and expression studies of rodent paracellin-1. J. Am. Soc. Nephrol. 12, 2664–2672 (2001). Klingel, K. et al. Ongoing enterovirus-induced myocarditis is associated with persistent heart muscle infection: quantitative analysis of virus replication, tissue damage, and inflammation. Proc. Natl Acad. Sci. USA 89, 314–318 (1992).
