Mutations in OSTM1 (Grey Lethal) Define a Particularly Severe Form of Autosomal Recessive Osteopetrosis With Neural Involvement

Mutations in OSTM1 (Grey Lethal) Define a Particularly Severe Formof Autosomal Recessive Osteopetrosis With Neural Involvement

Alessandra Pangrazio,1 Pietro Luigi Poliani,2 André Megarbane,3 Gérard Lefranc,4 Edoardo Lanino,5 Maja Di Rocco,6

Francesca Rucci,1 Franco Lucchini,7 Maria Ravanini,2 Fabio Facchetti,2 Mario Abinun,8 Paolo Vezzoni,1 Anna Villa,1

and Annalisa Frattini1

ABSTRACT: We report three novel osteopetrosis patients with OSTM1 mutations and review two that havebeen previously described. Our analysis suggests that OSTM1 defines a new subset of patients with severecentral nervous system involvement. This defect is also present in the gl mouse, which could represent a goodmodel to study the role of the gene in the pathogenesis of this disease.

Introduction: Autosomal recessive osteopetrosis (ARO) is a severe hereditary bone disease whose cellularbasis is in the osteoclast, but with heterogeneous molecular defects. In addition to the TCIRG1 and the ClCN7genes, whose mutations account for ∼55% and 10% of cases, respectively, the OSTM1 gene has been describedthus far in only two ARO patients.Materials and Methods: We report here three novel ARO patients presenting with severe primary centralnervous system involvement in addition to the classical stigmata of severe bone sclerosis, growth failure,anemia, thrombocytopenia, and visual impairment with optic atrophy. In addition we analyzed the brainmorphology and histology of the grey lethal mutant mouse.Results: The analysis of the OSTM1 gene in two patients, both from Kuwait, showed homozygous twonucleotide deletion in exon 2, leading to a frameshift and premature termination. The third (Lebanese) patientshowed a single point mutation in exon 1, leading to a nonsense mutation. The clinical neurological evaluationof the two Kuwaiti patients by CT and MRI scans showed a defect in the white matter, with a specific diagnosisof severe cerebral atrophy. The gl brain showed a diffuse translucent appearance with loss of the normaldemarcation between the white and the grey matter, features consistent with myelin loss or hypomyelination.Histological and myelin staining analysis evidenced an atrophy of the corpus callosum with loss of myelinfibers, and in cortical areas, loss of the normal lamination consistent with multiple foci of cortical dysplasia.Conclusions: These findings suggest that OSTM1-dependent ARO defines a new subset of patients withsevere central nervous system involvement leading to a very poor prognosis. The fact that central nervoussystem involvement is also present in the gl mouse mutant suggests that this mouse is a good model to testpossible therapies.J Bone Miner Res 2006;21:1098–1105. Published online on May 8, 2006; doi: 10.1359/JBMR.060403

Key words: osteopetrosis, OSTM1, grey-lethal mouse model, neural defect, myelin defect

INTRODUCTION

AUTOSOMAL RECESSIVE OSTEOPETROSIS (ARO; MIM259700) is a severe hereditary bone disease whose cel-

lular basis is in the osteoclast, but whose molecular defect isheterogeneous. Approximately 55% of patients with clini-cal diagnosis of ARO show an abnormality in the TCIRG1gene (MIM 604592), coding for the a3 subunit of the vacu-olar proton pump (VPP).(1–6) VPP plays a fundamental role

in acidifying the osteoclast–bone interface, which is a pre-requisite for bone mineral resorption. This acidification isalso hampered by a defect in the ClCN7 gene (MIM602727), which is mutated in ∼10% of ARO patients.(7,8)

This gene, whose mutations are also responsible for most ofpatients with autosomal dominant osteopetrosis type II(ADOII)(9) is involved in the acidification of resorptionlacuna, although its exact role in this process has not beendefined. Indeed, ClCN7 is usually considered a Cl− channel,but recent work on other members of the CLC family hasquestioned this assumption.(10,11)The authors state that they have no conflicts of interest.

1Institute for Biomedical Technologies, CNR, Milan, Italy; 2Department of Pathology I, University of Brescia, Brescia, Italy; 3Unité deGénétique Médicale, Campus des Sciences Médicales, Beyrouth, Lebanon; 4Laboratoire d’Immunogénétique Moléculaire, UniversitéMontpellier II, Montpellier, France; 5Department of Pediatric Hematology-Oncology - BMT Unit, Gaslini Institute, Genoa, Italy; 6Unitof Rare Diseases, Gaslini Institute, Genoa, Italy; 7Biotechnologies Research Center, Catholic University of Sacro Cuore, Cremona, Italy;8Children’s BMT Unit, Newcastle General Hospital, Newcastle upon Tyne, United Kingdom.

JOURNAL OF BONE AND MINERAL RESEARCHVolume 21, Number 7, 2006Published online on May 8, 2006; doi: 10.1359/JBMR.060403© 2006 American Society for Bone and Mineral Research

1098

JO512760 1098 1105 July

A third gene found mutated in human ARO is the greylethal gene (gl), named OSTMI, (osteopetrosis associatedtransmembrane protein 1; MIM 607649). It is responsiblefor the grey-lethal mutation in mouse and has been de-scribed so far in only two ARO patients.(12–14) Careful ge-notype/phenotype correlations in ARO have started to sug-gest that, although at a first glance ARO patients lookmonomorphous, at a closer examination the specific under-lying molecular defect seems to be of some prognostic andtherapeutic relevance. TCIRG1-dependent ARO patientshave a severe homogeneous phenotype; their nervous sys-tem involvement (hydrocephalus and cranial nerve defects)is secondary to the compression because of skull deformi-ties and the hematological defect can be rescued by HLA-matched hematopoietic stem cell (HSC) transplanta-tion.(15,16) On the other hand, there is increasing evidencethat patients with recessive ClCN7 mutations (ClCN7-dependent ARO), besides the bone manifestations, show aprimary severe neurological defect (retinopathy and pro-gressive cortical atrophy in addition to the secondary neuraldefects). The prognosis of this ClCN7-dependent formseems particularly poor because of the co-existence of thisneurological defect, which leads to death despiteHSCT.(7,17) Although TCIRG1 and ClCN7 are both di-rectly involved in the acidification of resorbing lacuna, thefunction of OSTM1 gene is much less clear. The gene hasbeen identified by classical molecular genetic studies in thespontaneous murine model, named grey lethal, which showsa severe molecular defect leading to a null phenotype.(14)

Interestingly, a study performed on the OSTM1 protein hasidentified this molecule as a G� interacting protein, a pu-tative E3 ubiquitin ligase, which could regulate protein ac-tivity through degradation.(18) Despite this study, the roleof the OSTM1 gene is still unknown.

Here we report three novel patients with OSTM1 muta-tions and review the two previously described.(12–14) Thisanalysis allows us to suggest that OSTM1-dependent AROdefines a new subset of patients with severe central nervoussystem involvement leading to a very poor prognosis. Inaddition, we show that the CNS involvement, although notpreviously described, is also present in the mouse mutant,suggesting that the gl mouse is a good model to study therole of this gene and to test possible alternative therapies.

MATERIALS AND METHODS

TCIRG1, ClCN7, and OSTM1 genemutation analysis

Specimens, including DNA, frozen peripheral bloodcells, fibroblasts, Epstein Barr virus (EBV)-transformedlymphoblast cell lines from patients, as well as DNAsamples from their parents, were collected with informedconsent. TCIRG1 (accession no. AF033033) and ClCN7(accession no. AL031600) gene sequence analysis was per-formed as previously described.(1,7) OSTM1 gene (acces-sion no. Z98200) analysis was performed by PCR with thefollowing primers (in each primer pair the first is the for-ward and the second the reverse)—for exon 1: ATAC-CCCAATGCACCACTCC and CGCTGACCATCA-TTACACCCTCC; for exon 2: ACTTAGTTCCTTGCTT-

GGGGC and TCGCTATTTGCTCAGTTGCC; for exon 3:CCGTGATTAGACCTTGTGCC and GTGCTCTA-AAAACTTGGAACTGC; for exon 4: TGCCATCTGCT-TCACATACCG and ACTAGAAGGTACATTCAA-TAACACTC; for exon 5: CCTGGCAGAAGAAGTT-GTCCTC and GCCACTGCACCTAGCCCTG; and forexon 6: TGATGTTGTTTTATTTGTACTGCTTC andGATTGTTCTTGCATGTTCTG.

All the reactions were performed in 25 �l of final volumewith 0.4 U Taq polymerase, 1.5 mM MgCl2, 300 �M dNTPs,10 pmol of each oligonucleotide primer, and 20 ng of puri-fied DNA. The thermocycling conditions used for amplifi-cation consisted of an initial denaturation step at 94°C for 3minutes, followed by 34 cycles of denaturation at 94°C for30 s, annealing at 65°C (for exon 1) at 58°C (for the otherexons) for 30 s, and 72°C for 30 s.

Automated sequencing was performed directly on thePCR products purified from the gel.

The mutations and patient information were collectedinto a database (http: / /bioinf .uta . f i /base_root/mutationdatabases.php).

Mice and histological analysis

Two pairs of heterozygous (GL/Le Edardl-J) mice werefrom the Jackson Laboratory (Bar Harbor, ME, USA).Mice were maintained in accordance with Italian Ministryof Health and European Community guidelines, and theexperimental procedures were in compliance with the guid-ing principles in the Guide for the Care and Use of Labo-ratory Animals.

Neuropathological studies were performed on four rep-resentative mice: two gl and two normal mice from thesame litter. All mice were deeply anesthetized and killed onthe 13th postnatal day. Brains and spinal cords were re-moved, fixed in 4% paraformaldehyde overnight, washed inPBS, and embedded in paraffin following a routine proto-col. Tissue sections were cut at 4 �m on a microtome andstained for histological examination. Routine H&E stainingwas used to study basic histopathological changes, and spe-cial staining was used to highlight peculiar neuropatholog-ical features. Myelin was studied by both histochemicalstaining (Luxol Fast Blue) and immunohistochemicalmethod, using antibodies against myelin basic protein (ratanti-MBP, 1:50; Chemicon, Hofheim, Germany) and 2�,3�-cyclic nucleotide 3�-phosphodiesterase (mouse anti-CNPase, 1:1,000; Sigma, St Louis, MO, USA). Silver im-pregnation (Bielschowsky staining) was used to study indetail the central nervous system cell population and fibersdistribution. Gliosis and astrocytic cell populations wereassayed by immunohistochemistry against glial fibrillaracidic protein (rabbit anti-GFAP; Dako Cytomation, Glos-trup, Denmark). Immunohistochemistry was performed ac-cording to the following procedure. Briefly, sections weredewaxed and rehydratated through serial passages in xy-lene and alcohol, washed in PBS before incubation for 15minutes at room temperature in blocking solution (2% nor-mal goat serum [NGS] in PBS/1% BSA), and incubatedwith primary antibodies for 2 h at room temperature in thesame solution. Sections were washed in PBS, incubated

HUMAN OSTEOPETROSIS CAUSED BY THE GREY LETHAL GENE I 1099

with a buffer containing the correct secondary antibody(goat anti-mouse/rabbit, 1:200; Dako Cytomation; rabbitanti-rat, 1:200; Vector, Burlingame, CA, USA) for 30 min-utes, washed again in PBS, and incubated for 30 minutes inbuffer containing streptavidin-horseradish peroxidase(HRP, 1:20; BioGenex, S. Ramon, CA, USA). Signal wasrevealed with diaminobenzydine (DAB, 1:50; BioGenex),and slides were washed, counterstained with hematoxylin,and coverslipped in a xyline-based mounting medium.

RESULTS

Patients and clinical findings

As part of an international study, we collected materialfrom 160 unrelated patients with a clinical diagnosis of se-vere osteopetrosis made on the basis of the following cri-teria at presentation: early diagnosis (before 2 years of age)and radiographic evidence of markedly increased BMDwith consequent anemia, pancytopenia, hepatosplenomeg-aly, and cranial nerves defects. Clinical, radiological, andlaboratory data were collected for genotype/phenotype cor-relation studies. Based on the clinical phenotype, patientswere screened for the TCIRG1 and ClCN7 genes. All thepatients with no mutation in these genes were sequencedfor the OSTM1 gene. Including the previously reported pa-tient(12,14) (referred as patient 4 in the present paper),among 160 ARO patients, we have found four cases har-boring defects in OSTM1 gene, highlighting the low inci-dence of osteopetrosis caused by mutations in this gene(2.5%). Here we report the three novel patients presentingwith a very severe picture of ARO (patients 1, 2, and 3),whereas the molecular and clinical analysis of patient 4 hasbeen described. Interestingly, these three patients werefrom the Middle East, all from consanguineous parents,whereas patient 4 was Italian.

The clinical data of patient 1 (male), the third child bornin Kuwait to consanguineous parents (first cousins) hasbeen previously described,(19) although at that time, no mo-lecular analysis was done. At 7 weeks of age, in addition toanemia and thrombocytopenia, he had severe visual impair-ment with optic atrophy, hypertonicity, and microcephaly.

His hearing assessment revealed normal bilateral oto-acoustic emissions and brainstem responses.

However, he had no consistent visual response, his retinawas abnormal (retinal dystrophy), both discs were pale, andvisual evoked responses were absent. These findings sug-gested a neuronal storage disease, but the electroencepha-logram and the white cell lysosomal enzyme assays showed

no abnormalities. MRI and CT scans of the head/brainshowed normal-looking white matter, with marked cerebralatrophy (Fig. 1A).

Patients 2 and 3 are novel and are described here. Patient2 (male) is the third child of consanguineous parents (firstcousins) from Lebanon. His older sister died at 6 months ofage from osteopetrosis. At birth, patient 2 was of normalweight (3100 g, 25th percentile) and length (48 cm, 25thpercentile). He was admitted to the hospital at age 14 dayswith hypotonia, diarrhea, vomiting, and difficulty in movinghis hands. On physical examination, he had a squared face,mild exophthalmia, a long philtrum, and marked hepato-splenomegaly. X-rays showed diffuse increase in BMD withloss of corticomedullary differentiation, especially in skullbones, upper limbs, pelvis, and femora, in addition to bilat-eral congenital hip subluxation. Diaphyseal fractures of theupper left humerus and the right femur were present. Labo-ratory studies revealed hypochromic anemia and pancyto-penia.

Bone marrow analysis showed a defect of central origininvolving the megakaryocyte and granulocyte lineages.

He was recurrently admitted for treatment of his anemiaand thrombocytopenia that worsened with time, even withrepeated blood transfusion. At 5 months of age, he wasadmitted to the intensive care for severe diarrhea, melena,and hyperthermia. His weight was 4200 g (<5th percentile).He died a few days later because of diffuse bleeding, relatedto his low platelet count (16,000/�l).

Patient 3 was a 12-month-old female, born from healthyconsanguineous Kuwaiti parents. Two brothers and one sis-ter died at the age of 18 months, 2 months, and 13 days,respectively, with the same clinical picture.

In the first month of life, osteopetrosis was diagnosed,and she received blood transfusions weekly. At the age of10 months she was referred for bone marrow transplanta-tion (BMT).

At clinical examination, she presented with severegrowth failure, hepatosplenomegaly, gum hypertrophy,limb hypertony, hyperreflexia, clonus, axial hypotonia, ab-sent head control, and nistagmus. Fundoscopic examinationrevealed pale optic disks. Because of the presence of theneurological symptoms described above, further examina-tion of the nervous system was performed. Cranial CT scansshowed severe cerebral atrophy with ex vacuo enlargementof ventricular system and subarachnoidal spaces and nar-rowing of optic foramina. Brain MRI showed diffuse ab-normal white matter signal with atrophy of cerebral hemi-

FIG. 1. (A) MRI and CT scan of patient 1and (B) MRI of patient 3. Coronal T2-weighted MRI shows marked cerebral atro-phy with corresponding enlargement of thelateral ventricles and subarachnoid spacesand right parietal subdural hematoma sec-ondary to brain shrinkage. Notice diffuse,faint hyperintensity of the bilateral cerebralwhite matter, consistent with hypomyelin-ation. Also notice diffuse hypomyelination ofthe cerebellar white matter in the absence ofcerebellar atrophy.

PANGRAZIO ET AL.1100

spheres, corpus callosum, and brainstem (Fig. 1B); protonMR spectroscopy showed reduction of the N-acetyl aspar-tate peak. The EEG was normal. Evoked visual potentialsshowed an abnormal cortical conduction. At brainstem au-ditory potentials, brainstem conduction was normal. Elec-tromyography and motor nerve conduction abnormalitieswere consistent with neuroaxonal disease; sensitive nerveconduction was normal.

Because of the extensive neurological impairment, BMTwas judged to not be indicated, and she was discharged withpalliative steroid therapy.

OSTM1 molecular analysis

Figure 2A shows the genomic structure of the OSTM1gene and the location of the identified mutations. In pa-tients 1 and 3, a homozygous two nucleotides deletion wasfound in exon 2 at nucleotide 415–416 of the cDNA se-

quence (415_416delAG of the cDNA sequence startingfrom the ATG codon; 90762–90763delAG of the genomicsequence, accession no. Z98200; Fig. 2B). This mutation,previously described in a Kuwaiti patient reported byRamirez et al.(13) and referred to here as patient 5, leads toa frameshift starting at codon 140 and premature termina-tion after 11 extraneous amino acids.

In both cases, the mutation was present at the heterozy-gous level in the parents. Because all these three patientswere from Kuwait, we studied whether the patients were insome way related. However, we found that they repre-sented three distinct families (C Kubisch, personal com-munication and A Frattini, unpublished data), although it ispossible that they share a common ancestor. This suggeststhat a mutation in this gene could be responsible for themajority of ARO in the Kuwait region.

Analysis of patient 2 showed a single point mutation (T

FIG. 2. Mutation analysis of patients. (A) Genomic structure of OSTM1 gene; arrows indicate the mutations of patients 1, 2, and 3.The chromatograms show the mutated OSTM1 sequence of (B) patients 1 and 3 and (C) patient 2 and the predicted changes in thecoded protein. In B, the 2-bp homozygous deletion in patients 1 and 3 is shown: an AG dinucleotide (in bold) is absent in thechromatogram (compare with the sequence below the chromatogram), predicting a frameshift with premature termination after 11codons, as shown at the bottom of the panel (also in bold). (C) A homozygous T to A transversion (in bold) in patient 2 causes a stopcodon.

HUMAN OSTEOPETROSIS CAUSED BY THE GREY LETHAL GENE I 1101

Fig 2 live 4/C

to A transversion) at nucleotide 36 of the cDNA sequence(corresponding to nt 80433 of genomic sequence), leadingto a nonsense mutation (Cys12X; Fig. 2C). Both parentswere heterozygous for this mutation.

Haplotype analysis of OSTM1-dependentARO patients

Because of the fact that two of our patients (patients 1and 3) and a third reported in the literature (patient 5) wereof Kuwaiti origin, we performed analysis of 22 singlenucleotide polymorphisms (SNPs) in the OSTM1 regionranging from 2.3 Mb upstream to 1 Mb downstream of thegene to study whether all three patients shared the samehaplotype.

As shown in Table 1, all the SNPs were identical in all thethree patients except for the two last downstream SNPsmapping at about 1 Mb, which were different in patient 5,who inherited a recombinant allele from both his parents.This might suggest a common origin for the mutation-bearing allele, with a recombination event occurring in theancestors of patient 5, 1 Mb downstream the OSTM1 gene.

Nature of the brain defect inOSTM1-related pathologies

An involvement of the nervous system not secondary tothe skull defect was apparent in all the patients and wassevere enough to cause early death or lead to the decisionnot to perform BMT. Post mortem data were available inonly one case,(12) revealing a decreased myelinization andabnormality of the white matter, in agreement with theresults of a CT scan. Detailed clinical neurological evalua-tion was available for patients 1 and 3, and both hadmarked neurological impairment and cortical cerebral at-rophy. A defect in the white matter was also found in pa-tient 3. Because of these findings, we analyzed the gl mutantmouse to study whether similar findings were present.

On gross examination, the brains from gl mice displayedsmaller size (1 cm of major axis), with irregular globose

hemispheres compared with normal (wt) mice (1.2 cm ofmajor axis; Fig. 3A). In contrast to wt mice, on cut surface,gl mouse brains showed a diffuse translucent appearancewith loss of the normal demarcation between the white andgrey matter (Figs. 3B–3E), features consistent with myelinloss or hypomyelination. In one of the animals, the cerebel-lar shape was found to be irregular. This finding did notshow any corresponding histological alteration and mightbe related to secondary involvement of cerebellar paren-chyma because of a skull defect rather then to primarycerebellar atrophy. Histological analysis showed a mild at-rophy of the corpus callosum in gl with loss of myelin fiberscompared with wt mice showing normal thickness withlarge number of myelin fibers crossing the hemispheres(Fig. 3F), as highlighted by myelin staining. Interestingly,many sections throughout the whole brain from gl miceshowed cortical areas with loss of normal lamination (Fig.3H) constituted of misaligned pleomorphic dysplastic neu-rons (Fig. 3I) and consistent with multiple foci of corticaldysplasia. Glial fibrillar acidic protein (GFAP) stainingshowed only mild gliosis with normal distribution of glialcells. Other brain regions and spinal cord didn’t show anyapparent alterations.

DISCUSSION

Dissection of the heterogeneity of hereditary diseases isincreasingly showing to be of clinical relevance. Mutationsin the same gene have been associated to different diseaseswith the pathology depending, at least in part, on the par-ticular protein domain affected by the mutation. Con-versely, relatively homogeneous clinical picture has beenfound to be caused by mutations in different genes. Humanosteopetroses are a clear example of this phenomenon, be-cause careful correlation between clinical findings and mo-lecular analysis contributed to understand the natural his-tory of these diseases and to a better formulation ofprognosis that could eventually direct therapeutic choices.

TABLE 1. SNP ANALYSIS OF KUWAITI PATIENTS AND THEIR PARENTS*

Patient

Chromosomal SNP in to the OSTM1 locus†

2.3 Mb upstream(rs2223972,rs12524594,rs2057154)

1.0 Mb upstream(rs9398192, rs9487019,rs932223, rs9374070)

7 kb upstream(rs9384672,rs2157503,rs7449720)

OSTM1 intron 3(rs723861, rs723862,rs1989193, rs718174)

3� UTR(rs9320250)

1.0 Mbdownstream(rs9373914,rs1762716)

Patient 1 C/C C/C A/A C/C A/A T/T G/G C/C A/A A/A A/A T/T G/G C/C A/A A/A C/CPatient 3 C/C C/C A/A C/C A/A T/T G/G C/C A/A A/A A/A T/T G/G C/C A/A A/A C/CPatient 5 C/C C/C A/A C/C T/T G/G C/C A/A A/A A/A T/T G/G C/C A/A T/T G/GFather of

patient 3 C/C C/T A/A C/C A/G T/T G/G C/C A/A A/A A/A T/T G/G C/C A/A A/A C/CMother of

patient 3 C/T C/C A/G C/C A/A T/T G/G C/G A/G A/G A/G T/C G/A C/T A/G A/A C/CFather of

patient 5 C/C C/C A/A C/T T/C G/A C/C A/A A/A A/A T/T G/G C/C A/A A/T C/GMother of

patient 5 C/C C/T A/A C/C A/A T/T G/G C/G A/G A/G A/G T/C G/A C/T A/G A/T C/G

* The following SNPs were tested and found identical in all samples: rs2057152, rs9384704, rs17069239, rs15856, and rs12525270.† Accession number: Z98200.

PANGRAZIO ET AL.1102

This can be exemplified by the recent analysis of theClCN7 gene. Molecular defects in ClCN7 gene are associ-ated to a wide range of clinical presentations, ranging fromasymptomatic ADOII to a very severe ARO.(7–9,20–24)

However, despite the great variability which is sometimespresent even within the same family,(7,9,22) the clinical pic-ture of individual patients is related, to some extent, tospecific mutations. Only a limited number of ClCN7 allelesat the heterozygous state give rise to ADOII, whereas oth-ers, such as those present in the parents of ClCN7-dependent ARO patients, are asymptomatic. Moreover,the presence of two mutated alleles underlying a very se-vere ARO picture (ClCN7-dependent ARO) discriminatethese patients from the classical TCIRG1-dependent be-cause the former has been shown to bear a poorer progno-sis, related in part to the associated primary neurologicaland retinal defect. As a consequence of the different defectsinvolved, it has been suggested that BMT would not beuseful in ClCN7-ARO,(7,17) because the neurological defectcannot be rescued as previously shown.(16)

OSTM1 is the third gene mutated in human ARO. Threenovel OSTM1-dependent cases are described here, in ad-dition to the two already reported cases.(12,13) Despite thesmall number of patients, two sets of considerations suggestthat OSTM1-dependent ARO can represent a distinct sub-set of patients with a peculiar clinical picture and prognosis.First, all the cases showed a severe central nervous systeminvolvement, whose predominant symptoms are marked ce-rebral atrophy and a decreased myelinization, documentedby autopsy and/or imaging (patients 1, 3, and 4). Thesefindings initially suggested a clinical diagnosis of a lysosom-

al storage disease. However, electroencephalography, elec-tron microscopy of skin biopsy, and white cell lysosomalenzyme assays excluded this clinical hypothesis. In parallel,the analysis of the murine model revealed, besides the bonedefect, severe neuropathological alterations. Careful mor-phological and histological evaluation of the brain in glmice revealed atrophy of the corpus callosum with myelinloss or hypomyelination. A high number of cortical dyspla-sia foci were present in the cortex of gl mice, suggesting apossible defect of neural origin. Because the real cause ofdeath in these mice is not completely clear, further studiesshould be performed to correlate them to the neurologicaldeficit in OSTM1 patients and better explain the neuro-pathological features.

All the OSTM1-ARO patients described thus far show avery severe clinical presentation and poor prognosis. In par-ticular, patients 1, 3, and 5 presented with such severe neu-rological signs that BMT was not offered as a therapeuticprocedure.(13,19) Patients 2 and 4 died very early.(14) In ad-dition, one sister of patient 2, three siblings of patient 3, anda brother of patient 5 who underwent BMT all died early. Itis likely that this severe prognosis is mainly caused by theinvolvement of the nervous system. It is not always easy todistinguish between primary and secondary involvement ofthe nervous system in osteopetrosis. Interestingly, osteope-trotic patients exhibiting specific features reminiscent of alysosomal storage disease with or without the agenesis ofthe corpus callosum have been described(25–28) (OMIM600329).

Taken together, although the underlying mechanisms ofthe nervous involvement need to be further studied, it is

FIG. 3. Neuropathological features of glmutant versus wildtype (wt) mouse brain.(A) On gross examination, brains from wtmice show normal morphology with well-formed hemispheres and regular size (1.2 cmof major axis for a 13-day-old mouse) in con-trast with brains from a gl mouse that showglobose hemispheres and smaller size (1 cmof major axis for a 13-day-old mouse). Thecut surface of (B and D) wt mouse brainshows good demarcation between white andgrey matter in contrast with (C and E) glmice that show a translucid appearance withmodest white-grey matter demarcation. My-elin staining using an antibody recognizingthe myelin basic protein (MBP) shows nor-mal staining of the corpus callosum in wtmice (F; original magnification, ×4) withdense number of myelin fibers crossing thehemispheres. In contrast, gl mouse (G; origi-nal magnification, ×4) shows a thinner corpuscallosum with loss of myelin fibers. Interest-ingly, on regular H&E staining, the sectionshows many diffuse foci of cortical dysplasiawith loss of normal lamination (H; originalmagnification, ×4; arrows indicate two differ-ent foci of cortical dysplasia) and irregulardistribution of misaligned and pleomorphicneurons (I; original magnification, ×20).

HUMAN OSTEOPETROSIS CAUSED BY THE GREY LETHAL GENE I 1103

Fig 3 live 4/C

clear that OSTM1 patients show abnormalities in the cen-tral nervous system, similar to what can be observed in glmice. Defective myelination, hypoplasia of corpus callo-sum, and cerebral atrophy are shared by both human pa-tients and gl mice. These observations might explain at leastin part their poor prognosis and suggest the hypothesis thatthese patients do not benefit from HSC therapeutic ap-proach.

ACKNOWLEDGMENTS

The authors thank Dr C Kubisch for providing DNAsamples from the family described in Ramirez et al. (2004)and Drs MT Sfar and J. Alroy for providing samples ofosteopetrotic patients. The authors thank Drs Noëlle Sou-raty and Peter Noun for help in the diagnosis of patient 2.The authors thank Dr Paola Forabosco for SNP analysisand for useful discussion. The technical assistance of LuciaSusani and Stefano Mantero is acknowledged. This studywas supported by grants from FIRB to PV and AV(RBNE019J9W) and from Grant 2006 by FondazioneCariplo to AF. This is manuscript 2 of the N.O.B.E.L. Proj-ect funded by Fondazione CARIPLO to AV and PV.

REFERENCES

1. Susani L, Pangrazio A, Sobacchi C, Taranta A, Mortier G,Savarirayan R, Villa A, Orchard P, Vezzoni P, Albertini A,Frattini A, Pagani F 2004 TCIRG1-dependent recessive osteo-petrosis: Mutation analysis, functional identification of thesplicing defects and in vitro rescue by U1snRNA. Hum Mutat24:225–235.

2. Scimeca J-C, Quincey D, Parrinello H, Romatet D, Gros-george J, Gaudray P, Philip N, Fischer A, Carle GF 2003 Novelmutations in the TCIRG1 gene encoding the a3 subunit of thevacuolar proton pump in patients affected by infantile malig-nant osteopetrosis. Hum Mutat 21:151–157.

3. Taranta A, Migliaccio S, Recchia I, Caniglia M, Luciani M, DeRossi G, Dionisi-Vici C, Pinto RM, Francalanci P, Boldrini R,Lanino E, Dini G, Morreale G, Ralston SH, Villa A, VezzoniP, Del Principe D, Cassiani F, Palumbo G, Teti A 2003 Geno-type-phenotype relationship in human ATP6i-dependent auto-somal recessive osteopetrosis. Am J Pathol 162:57–68.

4. Sobacchi C, Frattini A, Orchard P, Porras O, Tezcan I, Ando-lina M, Babul-Hirji R, Baric I, Canham N, Chitayat D, Dupuis-Girod S, Ellis I, Etzioni A, Fasth A, Fisher A, Gerritsen B,Gulino V, Horwitz E, Klamroth V, Lanino E, Mirolo M, MusioA, Matthijs G, Nonomaya S, Notarangelo LD, Ochs HD, Su-perti Furga A, Valiaho J, van Hove JL, Vihinen M, Vujic D,Vezzoni P, Villa A 2001 The mutational spectrum of humanmalignant autosomal recessive osteopetrosis. Hum Mol Genet10:1767–1773.

5. Frattini A, Orchard PJ, Sobacchi C, Giliani S, Abinun M,Mattsson JP, Keeling DJ, Andersson A-K, Wallbrandt P,Zecca L, Notarangelo LD, Vezzoni P, Villa A 2000 Defects inthe TCIRG1 subunit of the vacuolar proton pump are respon-sible for a subset of human autosomal recessive osteopetrosis.Nat Genet 25:343–346.

6. Kornak U, Schulz A, Friedrich W, Uhlhaas S, Kremens B, VoitT, Hasan C, Bode U, Jentsch TJ, Kubisch C 2000 Mutations inthe a3 subunit of the vacuolar H(+)-ATPase cause infantilemalignant osteopetrosis. Hum Mol Genet 9:2059–2063.

7. Frattini A, Pangrazio A, Susani L, Sobacchi C, Mirolo M, Abi-nun M, Andolina M, Flanagan A, Horwitz EM, Mihci E, No-tarangelo LD, Ramenghi U, Teti A, Van Hove J, Vujic D,Young T, Albertini A, Orchard PJ, Vezzoni P, Villa A 2003

Chloride channel ClCN7 mutations are responsible for severerecessive, dominant, and intermediate osteopetrosis. J BoneMiner Res 18:1740–1747.

8. Kornak U, Kasper D, Bosl MR, Kaiser E, Schweizer M, SchulzA, Friedrich W, Delling G, Jentsch TJ 2001 Loss of the ClC-7chloride channel leads to osteopetrosis in mice and man. Cell104:205–215.

9. Cleiren E, Benichou O, Van Hul E, Gram J, Bollerslev J,Singer FR, Beaverson K, Aledo A, Whyte MP, Yoneyama T,deVernejoul M-C, Van Hul W 2001 Albers-Schonberg disease(autosomal dominant osteopetrosis, type II) results from mu-tations in the ClCN7 chloride channel gene. Hum Mol Genet10:2861–2867.

10. Picollo A, Pusch M 2005 Chloride/proton antiporter activity ofmammalian CLC proteins ClC-4 and ClC-5. Nature 436:420–423.

11. Scheel O, Zdebik AA, Lourdel S, Jentsch TJ 2005 Voltage-dependent electrogenic chloride/proton exchange by endo-somal CLC proteins. Nature 436:424–427.

12. Quarello P, Forni M, Barberis L, Defilippi C, Campagnoli MF,Silvestro L, Frattini A, Chalhoub N, Vacher J, Ramenghi U2004 Severe malignant osteopetrosis caused by a GL gene mu-tation. J Bone Miner Res 19:1194–1199.

13. Ramirez A, Faupel J, Goebel I, Stiller A, Beyer S, Stockle C,Hasan C, Bode U, Kornak U, Kubisch C 2004 Identification ofa novel mutation in the coding region of the grey-lethal geneOSTM1 in human malignant infantile osteopetrosis. Hum Mu-tat 23:471–476.

14. Chalhoub N, Benachenhou N, Rajapurohitam V, Pata M, Fer-ron M, Frattini A, Villa A, Vacher J 2003 Grey-lethal mutationinduces severe malignant autosomal recessive osteopetrosis inmouse and human. Nat Med 9:399–406.

15. Driessen GJ, Gerritsen EJ, Fischer A, Fasth A, Hop WC, VeysP, Porta F, Cant A, Steward CG, Vossen JM, Uckan D, Frie-drich W 2003 Long-term outcome of haematopoietic stem celltransplantation in autosomal recessive osteopetrosis: AnEBMT report. Bone Marrow Transplant 32:657–663.

16. Gerritsen EJ, Vossen JM, Fasth A, Friedrich W, Morgan G,Padmos A, Vellodi A, Porras O, O’Meara A, Porta F 1994Bone marrow transplantation for autosomal recessive osteo-petrosis. A report from the Working Party on Inborn Errors ofthe European Bone Marrow Transplantation Group. J Pediatr125:896–902.

17. Kasper D, Planells-Cases R, Fuhrmann JC, Scheel O, Zeitz O,Ruether K, Schmitt A, Poet M, Steinfeld R, Schweizer M,Kornak U, Jentsch TJ 2005 Loss of the chloride channel ClC-7leads to lysosomal storage disease and neurodegeneration.EMBO J 24:1079–1091.

18. Fischer T, De Vries L, Meerloo T, Farquhar MG 2003 Promo-tion of G alpha i3 subunit down-regulation by GIPN, a putativeE3 ubiquitin ligase that interacts with RGS-GAIP. Proc NatlAcad Sci USA 100:8270–8275.

19. Abinun M, Newson T, Rowe PW, Flood TJ, Cant AJ 1999Importance of neurological assessment before bone marrowtransplantation for osteopetrosis. Arch Dis Child 80:273–274.

20. Del Fattore A, Peruzzi B, Rucci N, Recchia I, Cappariello A,Longo M, Fortunati D, Ballanti P, Iacobini M, Luciani M,Devito R, Pinto R, Caniglia M, Lanino E, Messina C, Cesaro S,Letizia C, Bianchini G, Fryssira H, Grabowski P, Shaw N,Bishop N, Hughes D, Kapur R, Datta H, Taranta A, Fornari R,Migliaccio S, Teti A 2006 Clinical, genetic and cellular analysisof forty-nine osteopetrotic patients: Implications for diagnosisand treatment. J Med Genet 43:315–325.

21. Letizia C, Taranta A, Migliaccio S, Caliumi C, Diacinti D,Delfini E, D’Erasmo E, Iacobini M, Roggini M, Albagha OM,Ralston SH, Teti A 2004 Type II benign osteopetrosis (Albers-Schonberg disease) caused by a novel mutation in CLCN7 pre-senting with unusual clinical manifestations. Calcif Tissue Int74:42–46.

22. Campos-Xavier AB, Saraiva JM, Ribeiro LM, Munnich A,Cormier-Daire V 2003 Chloride channel 7 (CLCN7) gene mu-tations in intermediate autosomal recessive osteopetrosis.Hum Genet 112:186–189.

PANGRAZIO ET AL.1104

23. Waguespack SG, Koller DL, White KE, Fishburn T, Carn G,Buckwalter KA, Johnson M, Kocisko M, Evans WE, ForoudT, Econs MJ 2003 Chloride channel (ClCN7) gene mutationsand autosomal dominant osteopetrosis, type II. J Bone MinerRes 18:1513–1518.

24. Campos-Xavier AB, Casanova JL, Doumaz Y, Feingold J,Munnich A, Cormier-Daire V 2005 Intrafamilial phenotypicvariability of osteopetrosis due to chloride channel 7 (CLCN7)mutations. Am J Med Genet 133:216–218.

25. Steward CG 2003 Neurological aspects of osteopetrosis. Neu-ropathol Appl Neurobiol 29:87–97.

26. Ben Hamouda H, Sfar MN, Braham R, Ben Salah M, Ayadi A,Soua H, Hamza H, Sfar MT 2001 Association of severe auto-somal recessive osteopepetrosis and Dandy-Walker syndromewith agenesis of the corpus callosum. Acta Orthop Belg67:528–532.

27. Rees H, Ang L-C, Casey R, George DH 1995 Association of

infantile neuroaxonal dystrophy and osteopetrosis: A rare au-tosomal recessive disorder. Pediatr Neurosurg 22:321–327.

28. Alroy J, Castagnaro M, Skutelsky E, Lomakina I 1994 Lectinhistochemistry of infantile lysosomal storage disease associatedwith osteopetrosis. Acta Neuropathol (Berl) 87:594–597.

Address reprint requests to:Annalisa Frattini, PhD

Institute for Biomedical TechnologiesCNR via fratelli Cervi 93

20090 Segrate, Milan, ItalyE-mail: [email protected]

Received in original form December 19, 2005; revised form March3, 2006; accepted April 7, 2006.

HUMAN OSTEOPETROSIS CAUSED BY THE GREY LETHAL GENE I 1105