Loci for classical and a variant late infantile neuronal ceroid lipofuscinosis map to chromosomes 11p15 and 15q21-23
Loci for classical and a variant late infantile neuronal ceroid lipofuscinosis map to chromosomes 11p15 and 15q21-23J. D.Sharp1,*,+, R. B.Wheeler1,*,+, B. D.Lake2, M.Savukoski3, I. E.Järvelä3, L.Peltonen3, R. M.Gardiner1 and R. E.Williams1
1Department of Paediatrics, University College London Medical School, The Rayne Institute, University Street,London WC1E 6JJ,UK,2Department of Histopathology, Great Ormond Street Hospital for Children,London WC1E 3JH,UK and3Laboratory of Molecular Genetics, National Public Health Institute,Helsinki,Finland
Received November 28, 1996;Revised and Accepted January 14, 1997
The childhood neuronal ceroid lipofuscinoses (NCLs) are a group of autosomal recessive neurodegenerative disorders characterised by progressive visual failure, neurodegeneration, epilepsy and the accumulation of an autofluorescent lipopigment in neurones and other cells. Three main subtypes have been identified according to age of onset, clinical features and ultrastructural morphology. These are infantile NCL (INCL;CLN1), classical late infantile NCL (LINCL;CLN2) and juvenile NCL (JNCL;CLN3). Several atypical forms of late infantile NCL (LINCL) have also been described including a Finnish variant LINCL (CLN5). TheCLN2 gene has been excluded from theCLN1,CLN3 andCLN5 loci. A genome search was initiated using a homozygosity mapping strategy in five classical LINCL and two variant LINCL consanguineous families. A common region of homozygosity was identified on chromosome 11p15 in two of the classical families. Analysis of a further 33 classical LINCL families supported linkage in this region (Zmax = 3.07 at[Theta] = 0.06 atD11S1338). A common region of homozygosity was also observed on chromosome 15q21-23 in the two variant LINCL families. Extension of the analysis to include a further seven families of identical ultrastructural phenotype established linkage to this region (Zmax = 6.00 at[Theta] = 0.00 atD15S1020).
The childhood neuronal ceroid lipofuscinoses (NCLs) are a group of autosomal recessive disorders characterised by progressive visual failure, neurodegeneration, epilepsy and the accumulation of an autofluorescent lipopigment in neurones and other cells. They have been classified into three main subtypes according to age of onset, clinical features and ultrastructural morphology. These are infantile NCL (INCL; Haltia-Santavuori disease;CLN1; MIM256730) characterised ultrastructurally by granular osmiophilic deposits (GRODs); classical late infantile NCL (LINCL; Jansky-Bielschowsky disease;CLN2; MIM204500) characterised by curvilinear bodies (CVB); and juvenile NCL (JNCL; Batten's disease;CLN3; MIM304200) characterised by fingerprint profiles (FPP). INCL occurs predominantly in Finland, andCLN1maps to 1p32 (1 ). Concurrent with the positional cloning ofCLN1, a palmitoyl-protein thioesterase (PPT) was mapped by fluorescencein situ hybridisation (FISH) to 1p32. PPT is a glycoprotein involved in lipid modification and was confirmed as the INCL gene by mutation analysis (2 ). The same point mutation accounts for 95% of Finnish disease chromosomes.
TheCLN3 gene maps to 16p12 (3 ) and codes for a 438 amino acid protein with unknown function (4 ). A 1 kb deletion accounts for 81% of disease chromosomes. Several NCL families have presented with a juvenile age of onset, but ultrastructural morphology revealed GRODs instead of FPPs. These families have been excluded from 16p12 by linkage analysis (5 ), and a subset have been found to map to theCLN1region (O'Raweet al., unpublished results).
Cases with onset of symptoms within the late infantile age range exhibit the widest phenotypic variability, and a number of atypical forms or variants have been documented. Lake and Cavanagh (6 ) described a so-called early juvenile NCL subtype which had a clinical course typical of LINCL but with a later age of onset and unusual ultrastructural morphology (FPP). In addition, a variant form of LINCL found only in Finland has also been described(7 ,8 ). This variant exhibits an intermediate phenotype between LINCL and JNCL. Finnish variant LINCL (CLN5; MIM256731) has been mapped to chromosome 13q22 (9 ,10 ).
TheCLN2 gene has been excluded from theCLN1,CLN3 andCLN5 loci (9 ,11 ). This study describes the localisation of genes for both classical LINCL on chromosome 11p15 and a variant LINCL on 15q21-23 using a homozygosity mapping approach (12 ). An initial genome search forCLN2 was carried out in 24 classical LINCL families with >300 microsatellite markers. No evidence for linkage was detected (13 ,14 ). Therefore, a second search was carried out in a subset of five classical and two variant consanguineous LINCL families using homozygosity mapping.
Two-point lod scores between (a)CLN2 and 11p15 markers and (b) betweenCLN6 and 15q21 markers (total lod scores for markersD15S988 andD15S216 are for eight families only)
Over 400 markers were analysed in the five consanguineous classical LINCL families. This initial search identified several apparent regions of homozygosity shared between two or more families. However, analysis using a denser marker map revealed heterozygosity in all locations except for a region on chromosome 11. Two of the five families, 167 and 182, shared a 26 cM region of homozygosity on chromosome 11 betweenD11S1760 andD11S929 (Fig.1 a). In this region, families 87 and 178 were homozygous atD11S4181 andD11S1329, but lack of parental heterozygosity makes identity by state indistinguishable from identity by descent (Fig.1 b). Family 270 was also homozygous atD11S1329, but this marker was uninformative and haplotype analysis indicated that the family was recombinant betweenD11S1760 andD11S929 (Fig.1 c). Multipoint linkage analysis using the programme HOMOZ gave positive individual lod scores for families 167 and 182 and negative lod scores for families 87, 178 and 270. Whilst statistically significant evidence for linkage with either homogeneity or heterogeneity was not obtained in this small group of families, haplotype analysis was consistent with the existence of a locus in this region in at least two families. The analysis was therefore extended to include an additional 33 non-consanguineous classical families. A maximum pairwise total lod score of 3.07 was obtained at [Theta] = 0.06 (m = f) withD11S1338 (Table1 a). Analysis of the data with HOMOG did not provide significant evidence in favour of linkage with heterogeneity rather than homogeneity. Recombination events in the consanguineous families 167 and 182 placeCLN2 in the 28 cM region betweenD11S4181 andD11S929 (Fig.1 a).
A genome search was also carried out using two variant LINCL consanguineous families. Both originated from the Indian sub-continent. The affected individuals demonstrated a similar clinical course to the classical families but their histology included storage bodies typical of both LINCL and JNCL, i.e. CVB typical of LINCL and FPP more usually associated with JNCL. Both these families were recombinant at theCLN3 andCLN5 loci and were, therefore, distinct from both juvenile and Finnish variant LINCL (Minna Savukoski, unpublished results). No evidence for linkage was found in these families with markers on chromosome 11p15. Common regions of homozygosity were identified and analysed as previously. A 38 cM region of homozygosity in family 233 was present on chromosome 15q. Within this region, homozygosity was shared with family 318 between markersD15S970 andD15S983, a genetic distance of 12 cM in the region 15q21-23 (Fig.2 ). Analysis of one unaffected individual from family 318 (IV-I) indicated a recombination event in the maternal meiosis betweenD15S993 andD15S1020 (Fig.2 ). This narrowed the critical region down to 12 cM betweenD15S993 andD15S216. Analysis using HOMOZ gave a maximum lod score of 4.25 betweenD15S125 andD15S988. Subsequently, seven markers across this region were analysed in seven additional variant LINCL families. All these families demonstrated identical histology (CVB and FPP) but had a clinical course representing either classical LINCL or early juvenile NCL. A maximum total lod score of 6.0 was obtained at markerD15S1020 ([Theta] = 0.00). Two-point lod scores for all nine families with 15q21-23 markers are shown in Table1 b. One of the additional families was homozygous across the linked region and has since been confirmed to be consanguineous. The locus for variant LINCL on 15q21-23 has been designatedCLN6.
These data complete the mapping of loci which account for the main childhood subtypes of NCL. They illustrate the power of using a homozygosity mapping approach in a rare autosomal recessive disorder. An initial genome search, carried out in 24 classical LINCL families, did not have sufficient power to provide statistically significant results. A maximum lod score of only 0.91 ([Theta] = 0.10) was obtained for a marker (D11S860) which lies within the region of chromosome 11 to whichCLN2 maps (unpublished results). In contrast, although only two of the five consanguineous classical LINCL families in this study show definitive evidence for linkage to chromosome 11, a two-point lod score of 2.18 was obtained atD11S1329.
A founder mutation in strong linkage disequilibrium with close markers has been demonstrated in all the other childhood NCLs (15 -17 ). Whilst no evidence for this has been observed on either chromosome 11 or 15, it is anticipated that linkage disequilibrium may be apparent with additional markers closer to the disease locus. No strong candidate genes are present in these regions of chromosome 11p15 or 15q21-23. The gene for cathepsin D (CTSD, CPSD) maps to 11p15. A role for cathepsin in the NCLs has been suggested previously (18 ) but biochemical studies have revealed normal enzyme levels and activity in LINCL patients (19 ). An autosomal recessive mouse model of NCL (nclf) with FPP has been reported recently (20 ). This maps to the middle of mouse chromosome 9 in a region which is syntenic with human chromosome 15q. This may represent a homologue of variant LINCL and its localisation may help in the fine mapping ofCLN6.
The hallmark of the variant LINCL families linked to chromosome 15 is a mixed histology which in some cases is the only feature distinguishing them from classical LINCL. This further illustrates that ultrastructural features appear to be a better guide to locus homogeneity than phenotypes based on the clinical course alone. Some juvenile NCL patients with GRODs rather than FPP, for example, map to theCLN1region on chromosome 1 rather than theCLN3 region on chromosome 16 (O'Raweet al., unpublished results). These observations suggest that ultrastructural morphology is a better guide to classification of the NCL sub-types than is clinical phenotype.
Identification of these two loci demonstrate that at least five different genes underlie the autosomal recessive childhood varieties of NCL. Two NCL genes have already been isolated,CLN1 andCLN3, but have no apparent function in common. The geneCLN1 codes for a protein PPT which is involved in the lipid modification of proteins, whereasCLN3 codes for a highly hydrophobic protein with no homology to any existing protein of known function. It is likely that elucidation of whatever molecular process is common to this group of disorders awaits the molecular cloning ofCLN2, CLN5andCLN6.
Thirty eight families from 10 countries segregating for classical LINCL were studied. Five families were consanguineous, with six affected and one unaffected individuals. Thirty three additional families gave a total of 42 affected and 36 healthy siblings. All affected children developed symptoms between 2 and 4 years of age. The disease was characterised by ataxia, loss of skills, myoclonic seizures followed by visual failure and characteristic CVB seen on electron microscopy (EM) of various tissue samples.
A further nine variant LINCL families were studied originating from six different countries. Six of these families showed classical LINCL disease progression but with mixed histology (CVB and FPP). Of these, two were consanguineous and originated from the Indian sub-continent. These six families gave a total of 10 affected and eight healthy siblings. In a further three families, the phenotype was characterised by mixed CVB and FPP with clinical symptoms corresponding to early juvenile NCL. These three families gave a total of six affected and five healthy siblings.
Blood samples were obtained from all family members with informed consent. DNA was extracted using standard methods. Affected members of the consanguineous families were typed using fluorescently labelled microsatellite markers (heterozygosities >70%), distributed at 10-20 cM intervals, chosen from the set made available through the MRC (21 ) and the Généthon catalogue (22 ). Amplification was carried out in 15 µl reactions containing 25 ng of genomic DNA, 20 pmol of each primer, 1.0-3.0 mM MgCl2, 200 µM dNTPs, 10 mM Tris-HCl (pH 8.3), 50 mM KCl and 2 U of Red Hot DNA polymerase (Advanced Biotechnologies). Reactions were performed in 96-well microtitre plates using a Hybaid OmniGene thermal cycler and consisted of an initial denaturation at 94oC for 2.5 min, followed by 25 cycles of denaturation at 94oC for 35 s, annealing at 45-60oC for 30 s (dependent on the primers used) and elongation at 72oC for 15 s. The final step consisted of an additional elongation step at 72oC for 2 min. Following amplification, 5 µl of each reaction were pooled dependent on size and fluorescent label. An aliquot equivalent to 0.2 µl of each marker was removed, and 0.5 µl of an internal size standard (Genescan 500-TAMRA) and 4 µl of blue formamide were added. Samples were denatured at 94oC for 3 min and then separated by electrophoresis on 6% (w/v) denaturing polyacrylamide gels using a model 373A Applied Biosystems automatic sequencer. Genescan 672 and Genotyper version 1.1 software was used to size the PCR products and analyse the data.
Regions of homozygosity were identified and characterised further by typing the rest of the family members and using additional markers selected from the Généthon map (22 ). Informative microsatellite markers were typed in the total family resource. The GAS program (Alan Young, Oxford University) was used to convert allele sizes from mobility units to discrete alleles. Linkage analysis was carried out using LINKAGE version 5.2 (23 ). Consanguineous loops were broken at mothers. Autosomal recessive inheritance with complete penetrance and a disease allele frequency of 0.002 was assumed. Pairwise lod scores were calculated using the LODSCORE and MLINK options. Multipoint linkage analysis for the consanguineous families was calculated using HOMOZ (24 ). HOMOG (25 ) was used to formally test for locus heterogeneity in the classical resource.
This project would not have been possible without the helpful co-operation of the families, their physicians and the Batten's Cell Bank at Indiana University School of Medicine (supported by NS30171). Thanks go to Professor Pirkko Santavuori for clinical advice and to Mr Keith Parker for invaluable technical assistance with the ABI sequencer. We particularly wish to acknowledge the unfailing support and encouragement of Dr J. Alfred Rider and the Children's Brain Disease Foundation (USA). The financial support of the following is gratefully acknowledged: The Medical Research Council (UK), The Wellcome Trust (UK), United States Public Health Service, National Institutes of Health (USA), Action Research (UK), European Commision Concerted Action (contract BMH-4-CT95-0563), The Research Trust for Metabolic Diseases in Childhood (UK), The Sigrid Juselius Foundation, The Academy of Finland.
1 Järvelä, I. (1991) Infantile neuronal ceroid lipofuscinosis (CLN1): linkage disequilibrium in the Finnish population and evidence that variant late infantile form (variant CLN2) represents a nonallelic locus. Genomics,10, 333-337.MEDLINE Abstract
2 Vesa, J., Hellsten, E., Verkruyse, L.A., Camp, L.A., Rapola, J., Santavuori, P., Hofmann, S.L. and Peltonen, L. (1995) Mutations in the palmitoyl protein thioesterase gene causing infantile neuronal ceroid lipofuscinosis. Nature,376, 584-376.MEDLINE Abstract
3 Gardiner, R.M. (1992) Mapping the gene for juvenile onset neuronal ceroid lipfuscinosis to chromosome 16 by linkage analysis. Am. J. Med. Genet.,42, 539-541.MEDLINE Abstract
4 The International Batten Disease Consortium (1995) Isolation of a novel gene underlying Batten disease. Cell,82, 949-957.
5 Mitchison, H.M., O'Rawe, A.M., Lerner, T.J., Tascher, P.E.M., Schlumf, K., D'Arigo, K., Vos, N.d., Gormally, E., Phillips, H.A., Thompson, A.D., Haines, J.L., Hart, Y.M., Anderman, E., Callen, D.F., Breuning, M.H., Gardiner, R.M. and Mole, S.E. (1995) Refined localization of the Batten disease gene (CLN3) by haplotype and linkage disequilibrium mapping to D16S288-D16S383 and exclusion from this region of a variant form of Batten disease with granular osmoiphilic deposits. Am. J. Med. Genet.,57, 312-315.MEDLINE Abstract
6 Lake, B.D. and Cavanagh, N.P.C. (1978) Early-juvenile Batten's disease-a recognisable subgroup distinct from other forms of Batten's disease. J. Neurol. Sci.,36, 265-271.MEDLINE Abstract
7 Santavuori, P., Rapola, J., Saino, K. and Raitta, C. (1982) A variant of Jansky-Bielschowsky disease. Neuropediatrics,13, 135-141.MEDLINE Abstract
8 Santavuori, P., Rapola, J., Nuutila, A., Raininko, R., Lappi, M., Launes, J., Herva, R. and Sainio, K. (1991) The spectrum of Jansky-Bielowsky disease. Neuropaediatrics,22, 92-96.
9 Savukoski, M., Kestila, M., Williams, R., Jarvela, I., Sharp, J., Harris, J., Santavuori, P., Gardiner, M. and Peltonen, L. (1994) Defined chromosomal assignment of CLN5 demonstrates that at least four genetic loci are involved in the pathogenesis of human ceroid lipofuscinoses. Am. J. Hum. Genet.,55, 695-701.MEDLINE Abstract
10 Klockars, T., Savukoski, M., Isosomppi, J., Laan, M., Jarvela, I., Petrukhin, K., Palotie, A. and Peltonen, L. (1996) Efficient construction of a physical map by fiber-fish of the CLN5 region: refined assingnment and long-range contig covering the critical region on 13q22. Genomics,35, 71-78.MEDLINE Abstract
11 Williams, R., Vesa, J., Jarvlea, I., McKay, T., Mitchison, H., Hellsten, E., Thompson, A., Callen, D., Sutherland, G., Luna-Battadano, D., Stallings, R., Peltonen, L. and Gardiner, M. (1993) Genetic heterogeneity in neuronal ceroid lipofuscinosis (NCL): evidence that the late infantile subtype (Jansky-Bielschowsky disease; CLN2) is not an allelic form of the juvenile or infantile subtypes. Am. J. Hum. Genet.,53, 931-935.MEDLINE Abstract
12 Lander, E.S. and Bostein, D. (1987) Homozygosity mapping: a way to map recessive traits with the DNA of inbred children. Science,236, 1567-1570.MEDLINE Abstract
13 Sharp, J.D., Wheeler, R.B., Savukoski, M., Kestila, M., Jarvela, I.E., Peltonen, L., Eiberg, H., Gardiner, R.M. and Williams, R. (1995) Exclusion mapping of classical late infantile neuronal ceroid lipofuscinosis (Jansky-Bielschowsky disease, CLN2). Eur. J. Hum. Genet.,3, 326-328.MEDLINE Abstract
14 Wheeler, R.B., Sharp, J.D., Savukoski, M., Jarvela, I., Peltonen, L., Gardiner, R.M. and Williams, R.E. (1997) Linkage analysis of late infantile neuronal ceroid lipofuscinosis (Jansky-Bielchowsky disease). Neuropaediatrics,in press.
15 Hellsten, E., Vesa, J., Speer, M.C., Mäkelä, T.P., Järvelä, I., Alitaro, K., Ott, J. and Peltonen, L. (1993) Refined assignment of the infantile neuronal ceroid lipofuscinosis (INCL,CLN1) locus at 1p32: incorporation of linkage disequilibrium in multipoint analysis. Genomics,16, 720-725.MEDLINE Abstract
16 Mitchison, H.M., Thompson, A.D., Mulley, J.C., Kozman, H.M., Richards, R.I., Callen, D.F., Stallings, R.L., Doggett, N.A., Attwood, J., Mackay, T.R., Sutherland, G.R. and Gardiner, R.M. (1993) Fine genetic mapping of the Batten disease locus (CLN3) by haplotype analysis and demonstration of allelic association with chromosome 16p microsatellite loci. Genomics,16, 455-460.MEDLINE Abstract
17 Varilo, T., Savukoski, M., Norio, R., Santavuori, P., Peltonen, L. and Järvelä, I. (1996) The age of human mutation: genealogical and linkage disequilibrium analysis of the CLN5 mutation in the Finnish population. Am. J. Hum. Genet.,58, 506-512.MEDLINE Abstract
18 Ivy, G., Schottler, F., Wenzel, J., Baudry, M. and Lynch, G. (1984) Inhibitors of lysosomal enzymes: accumulation of lipofuscin-like dense bodies in the brain. Science,226, 985-987.MEDLINE Abstract
19 Ezaki, J., Wolfe, L.S. and Kominami, E. (1996) Ex vivo and in vitro studies on specific delay of degradation of mitochondrial ATP synthase subunit C in late infantile neuronal ceroid lipofuscinosis. The Sixth International Congress on Neuronal Ceroid Lipofuscinosis Gustavelund, Finland.
20 Donahue, L.R., Faust, J.R. and Bronsen, R.T. (1996) A new mouse model of Batten disease, neuronal ceroid lipofuscinosis (nclf), maps to chromosome 9. 10th International Mouse Genome Conference Paris, France.
22 Dib, C., S Faure, Fizames, C., Samson, D., Drouot, N., Vignal, A., Millasseau, P., Marc, S., Hazan, J., Seboun, E., Lathrop, M., Gyapay, G., Morrissette, J. and Weissenbach, J. (1996) A comprehensive genetic map of the human genome based on 5,264 microsatellites. Nature,380, 152-154.MEDLINE Abstract
23 Lathrop, G.M., Lalouel, J.M., Julier, C. and Ott, J. (1984) Strategies for multilocus linkage analysis in humans. Proc. Natl Acad. Sci. USA,81, 3443-3446.MEDLINE Abstract
24 Kruglyak, L., Daly, M.D. and Lander, E.S. (1995) Rapid multipoint linkage analysis of recessive traits in nuclear families, including homozygosity mapping. Am. J. Hum. Genet.,56, 519-527.MEDLINE Abstract
25 Ott, J. (1986) Linkage probability and its approximate confidence interval under possible heterogeneity. Genet. Epidemiol. Suppl. 1, 251-257.
*To whom correspondence should be addressed
+These two authors contributed equally to this work
This page is maintained by OUP admin. Last updated Thu Mar 13 17:29:21 GMT 1997. Part of the OUP Journals World Wide Web service.Copyright Oxford University Press, 1996
