Ceroid Lipofuscinosis, Neuronal, 2

A number sign (#) is used with this entry because neuronal ceroid lipofuscinosis-2 is caused by homozygous or compound heterozygous mutation in the TPP1 gene (607998) on chromosome 11p15.

Description

The neuronal ceroid lipofuscinoses (NCL; CLN) are a clinically and genetically heterogeneous group of neurodegenerative disorders characterized by the intracellular accumulation of autofluorescent lipopigment storage material in different patterns ultrastructurally. The clinical course includes progressive dementia, seizures, and progressive visual failure. The lipopigment pattern seen most often in CLN2 consists of 'curvilinear' profiles (Mole et al., 2005).

For a general phenotypic description and a discussion of genetic heterogeneity of CLN, see CLN1 (256730).

Nomenclature

The CLNs were originally classified broadly according to age at onset, with CLN2 as the late infantile-onset (LINCL) form with onset between 2 and 4 years of age. With the identification of molecular defects, however, the CLNs are now classified numerically according to the underlying gene defect. CLN2 refers to CLN caused by mutation in the CLN2 gene, regardless of the age at onset.

Clinical Features

Hassin (1926) reviewed the pathology of late infantile NCL. Seitelberger et al. (1957) collected 28 cases from the world's literature. Some cases reported as LINCL may have been instances of generalized gangliosidosis (Donahue et al., 1967).

Gonatas et al. (1968) described the gross and microscopic findings in LINCL. Brain weight was severely diminished and there was neuronal loss as well as intraneuronal accumulation of an eosinophilic material. Ultrastructural examination showed curvilinear bodies of a lysosomal nature. The characteristic inclusions may be found in other tissue types. On the basis of electron microscopic findings of 'multilamellar cytosomes,' Gonatas et al. (1968) suggested that 2 cases they studied and 4 cases reported by others represented a different type of late infantile cerebral lipidosis. Elfenbein and Cantor (1969) and Richardson and Bornhofen (1968) also reported cases of LINCL with multilamellar cytosomes. By electron microscopy, Dolman and Chang (1972) found curvilinear bodies not only in the central and autonomic nervous system but also in the cells of virtually every organ examined.

In the provinces of Quebec and Newfoundland, Andermann et al. (1977) ascertained 46 cases of cerebromacular degeneration (CMD) in 30 sibships: 27 cases of late infantile CMD (Jansky-Bielschowsky), 17 cases of juvenile CMD (Spielmeyer-Vogt), and 2 cases in 1 family of the adolescent form (Kufs) (204300). Two-thirds were Newfoundlanders of Anglo-Saxon descent. Andermann et al. (1988) described further the Newfoundland aggregate of LINCL. Age of onset was 2.5 to 3.5 years with seizures, rapid mental deterioration, ataxia, dementia, and quadriparesis. Retinal blood vessels were narrowed and optic atrophy was common. Ultrastructural study showed curvilinear profiles.

Taratuto et al. (1995) reported 24 cases of LINCL that had been diagnosed in Argentina from 1985 to 1993. Age of onset ranged from 1 to 6 years (mean 3.1). The clinical findings were homogeneous, including refractory epilepsy, mental regression and deterioration, ataxia, myoclonus, and visual loss. Affected patients demonstrated abnormal electroretinography (ERG), visual evoked potentials (VEP), and electroencephalograms (EEG). Brain biopsies from 3 patients showed neuronal loss, distended neurons with granular PAS-positive material, and curvilinear inclusions on electron microscopy.

Children affected with LINCL have retinal degeneration which is most visible in the macula, and the entire retina is involved as reflected by extinction of the electroretinogram (ERG) early in the disease (Brodsky et al., 1996). The cherry red spot typical of the infantile form of Tay-Sachs disease (272800) is not observed. Usual features are a strikingly enlarged VEP and large photically driven spikes on EEG.

Diagnosis

Prenatal Diagnosis

MacLeod et al. (1985) reported on the successful prenatal diagnosis of this form of NCL. A fetus was studied by electron microscopy at 16 weeks of gestation because of an affected sib. About one-third of a subpopulation of dark, elongated amniotic fluid cells contained one or more deposits of curvilinear cytosomes bound by a single unit membrane. After delivery at term, a punch biopsy and buffy coat preparation from the baby showed similar characteristic inclusions.

Berry-Kravis et al. (2000) used mutation analysis for prenatal diagnosis. Previously, prenatal testing for LINCL had been accomplished through electron microscopic examination of uncultured amniocytes for typical curvilinear bodies. They reported success in 2 cases and described a new private mutation in the CLN2 gene in one of the families (607998.0006).

Clinical Management

Schulz et al. (2018) reported the results of a multicenter, open-label study to evaluate the effect of intraventricular infusion of cerliponase alfa every 2 weeks in children with CLN2 disease who were between the ages of 3 and 16 years. Twenty-four patients were enrolled, 23 of whom constituted the efficacy population. All patients received a 300-mg dose for at least 96 weeks after a dose-escalation phase. The primary outcome was the time until a 2-point decline in the score on the motor and language domains of the CLN2 Clinical Rating Scale (which ranges from 0 to 6, with 0 representing no function and 3 representing normal function in each of the 2 domains), which was compared with the time until a 2-point decline in 42 historical controls. The median time until a 2-point decline in the motor-language score was not reached for treated patients and was 345 days for historical controls. The mean (+/-SD) unadjusted rate of decline in the motor-language score per 48-week period was 0.27 +/- 0.35 points in treated patients and 2.12 +/- 0.98 points in 42 historical controls (mean difference, 1.85; p less than 0.001). Common adverse events included convulsions (present in all patients and also part of the disease), pyrexia, vomiting, hypersensitivity reactions, and failure of the intraventricular device. In 2 patients, infections developed in the intraventricular device that was used to administer the infusion, which required antibiotic treatment and device replacement. Fifty percent of patients had 34 device-related adverse events, of which 5 were grade 3 in 4 patients. The authors concluded that intraventricular administration of cerliponase alfa every 2 weeks at a dose of 300 mg in children with CLN2 disease resulted in a slower rate of decline in motor and language function than in historical controls. Intraventricular ERT was associated with device-related complications including grade 3 infection, leakage, and increased white cell count in cerebrospinal fluid in half of the patients.

Mapping

Yan et al. (1993) demonstrated that CLN2 is not linked to markers in the region of chromosome 16 that carries the gene (CLN3; 607042) for Batten disease (204200). From studies of 25 families segregating for late infantile NCL, Williams et al. (1993) excluded both 16p and 1p as the site of the mutation, thus demonstrating that this form of NCL is not allelic to CLN3 or CLN1.

Sharp et al. (1995) excluded linkage of the CLN5 locus (608102) on 13q21.1-q32 to classic late infantile neuronal ceroid lipofuscinosis in a subset of 17 classic LINCL non-Finnish families originating from the United Kingdom, United States, and northern Europe.

Sharp et al. (1997) analyzed 400 DNA markers in 5 consanguineous classic LINCL families. The initial search identified several regions of apparent homozygosity shared between 2 or more families. However, analysis using a denser marker map revealed heterozygosity in all locations except for a region on chromosome 11. The analysis was extended to include an additional 33 nonconsanguineous classical families. A maximum pairwise total lod score of 3.07 was obtained at theta = 0.06 (m = f) with D11S1338. The markers showing linkage were located on 11p15.

Haines et al. (1998) presented evidence indicating that the CLN2 locus is located in a minimum candidate region of 11 cM on 11p15.5, flanked by marker loci D11S4046 on the telomeric side and D11S1996 on the centromeric side.

Molecular Genetics

In 2 unrelated patients with LINCL, Sleat et al. (1997) identified mutations in the CLN2 gene (607998.0001 and 607998.0002).

Zhong et al. (1998) screened 16 LINCL probands for 4 previously described CLN2 mutations. The intronic mutation IVS5-1G-C (607998.0004) was found in 9 of the 16 patients, of whom 2 were homozygous, and accounted for 34% (11 of 32) of CLN2 chromosomes. A nonsense mutation (607998.0003) was found in 31% (5 of 16) of the patients, including 1 homozygote, and accounted for 19% (6 of 32) of the CLN2 chromosomes. Together, one or both of these mutations were seen in 11 (69%) cases. The 2 other missense mutations were not found in any of the 16 probands, and no mutation was identified in 5.

To better understand the molecular pathology of LINCL, Sleat et al. (1999) conducted a genetic survey of the CLN2 gene in 74 LINCL families. In 14 patients, CLN2 protease activities were normal and no mutations were identified, suggesting other forms of NCL. Both pathogenic alleles were identified in 57 of the other 60 LINCL families studied. In total, 24 mutations were associated with LINCL, comprising 6 splice junction mutations, 11 missense mutations, 3 nonsense mutations, 3 small deletions, and 1 single-nucleotide insertion. Two previously reported mutations were particularly common: a splice junction mutation (607998.0004), found in 38 of 115 alleles, and a stop mutation (607998.0003), found in 32 of 115 alleles.

Genotype/Phenotype Correlations

Steinfeld et al. (2002) described the natural progression of LINCL in 22 German patients with CLN2 mutations, using a scoring system that allowed quantification of the motor, visual, and verbal performances over long periods of time. Sixteen of the patients, who were grouped together in the study, were homozygous or compound heterozygous for common mutations that result in complete loss of enzymatic activity.

Bessa et al. (2008) reported a 40-year-old Portuguese man with a mild protracted form of CLN2 who was homozygous for a mutation that created a potential acceptor site in intron 7 of the TPP1 gene (IVS7AS-10A-G; 607998.0009), predicted to result in a protein with 3 extra amino acids between codons 295 and 296 and not affecting the wildtype splice site. The patient had onset at age 10 years of progressive cognitive and motor dysfunction and seizures. Western blot analysis detected a 60% reduction in overall TPP1 protein levels, suggesting that the mutant protein had decreased stability. Bessa et al. (2008) concluded that the mutant protein retained enzyme activity, which was consistent with the milder phenotype.

Population Genetics

Moore et al. (2008) observed extensive genetic heterogeneity for NCL in Newfoundland. In total, 52 patients from 34 families were identified clinically. Of the 28 families with available DNA, 18 had 5 different mutations in the CLN2 gene (see, e.g., 607998.0007). One family had a CLN3 mutation (607042.0001), another had a CLN5 mutation (608102.0005), and 5 families shared the same mutation in CLN6 (606725.0010). One family was misdiagnosed, and molecular testing was inconclusive in 2 families. Patients with CLN2 had an earlier presentation and seizure onset compared to those with CLN6 mutations. There was a slower clinical course for those with CLN5 mutations compared with CLN2 mutations. Moore et al. (2008) estimated that NCL in Newfoundland has an incidence of 1 in 7,353 live births. The incidence of CLN2 was 9.0 per 100,000, or 1 in 11,161 live births, the highest reported in the world.

Animal Model

Sleat et al. (2004) found that mice with targeted homozygous disruption of the Tpp1 gene were viable and healthy at birth, but developed progressive neurologic deterioration around 7 weeks of age. Clinical features included tremor and ataxia, and neuropathologic examination showed extensive neuronal pathology with accumulation of autofluorescent cytoplasmic storage material within the lysosomal-endosomal compartment, loss of cerebellar Purkinje cells, and widespread axonal degeneration. The life span of mutant mice was significantly decreased compared to wildtype. The findings recapitulated the features of human CLN2.

Sleat et al. (2008) generated mouse models of CLN2 with different hypomorphic Tpp1 mutations. Mice that were homozygous for R446H, which is analogous to human R447H (607998.0005), had approximately 6% residual brain activity of Tpp1. Mice that were compound heterozygous for the R446H allele and a null allele had about 3% residual Tpp1 activity and showed delayed disease onset and longer survival compared to homozygous-null mice. Homozygosity for R446H resulted in dramatic attenuation of disease, with even further expanded life span. The findings indicated that residual levels of Tpp1 can ameliorate disease, which has potential therapeutic implications for humans with the disorder.

In addition to typical retinal changes in the Dachshund model of neuronal ceroid lipofuscinosis due to a null mutation in the TPP1 gene, Whiting et al. (2015) identified a retinopathy consisting of multifocal, bullous retinal detachment lesions in 65% of the TPP1-null dogs. The lesions did not occur in littermates that were heterozygous or homozygous for the normal TPP1 allele. The retinopathy consisted of progressive multifocal serous retinal detachments. The severity of the disease-related retinal thinning was no more serious in most detached areas than in adjacent areas of the retina that remained in close apposition to the retinal pigment epithelium. The ERG a-wave amplitudes were relatively preserved in the TPP1-null dogs, whether or not they developed the multifocal retinopathy. DNA sequence analysis ruled out a mutation in the BEST1 (607854) exons and splice junctions as a cause for the retinopathy.