A novel CLN8 mutation in late‐infantile‐onset neuronal ceroid lipofuscinosis (LINCL) reveals aspects of CLN8 neurobiological function

The late‐infantile‐onset forms of neuronal ceroid lipofuscinosis (LINCL) are the most genetically heterogeneous group among the autosomal recessive neuronal ceroid lipofuscinoses (NCLs), with causative mutations found in CLN1, CLN2, CLN5, CLN6, CLN7 (MFSD8), and CLN8 genes. Homozygous mutations in CLN8 are associated with two distinct phenotypes: progressive epilepsy and mental retardation (EPMR), first identified in Finland; and a variant of late‐infantile NCL (v‐LINCL) described in a subset of Turkish and Italian patients. The function of the protein encoded by CLN8 is currently unknown. Here we report the identification of an Italian v‐LINCL patient with a complete isodisomy of chromosome 8, leading to homozygosity of a maternally‐inherited 3‐bp deletion in CLN8 gene (c.180_182delGAA, p.Lys61del). Notably, uniparental disomy (UPD) has never been described associated with the NCLs. In addition, we provide evidence of the biological role of CLN8 characterized by expressing in different neuronal cell models the native protein, the protein carrying the mutation identified here, or three additional missense mutations previously described. Our results, validated through a gene silencing approach, indicate that CLN8 plays a role in cell proliferation during neuronal differentiation and in protection against cell death. Hum Mutat 30:1–13, 2009. © 2009 Wiley‐Liss, Inc.     

CLN3 protein is targeted to neuronal synapses but excluded from synaptic vesicles: new clues to Batten disease

Batten disease (juvenile neuronal ceroid lipofuscinosis, JNCL), the most common neurodegenerative disease of childhood, is caused by mutations in the CLN3 gene encoding a putative transmembrane protein. The function of CLN3 is currently unknown but it has been shown to localize in the endosomal/lysosomal compartments of non-neuronal cells. In addition, several other intracellular localizations have been proposed and the controversy of the reports suggests that CLN3 may have different intracellular localization in different cell types. Batten disease severely affects neuronal cells but leaves other organs clinically unaffected, and thus it is of utmost importance to approach the disease mechanism by studying the expression and localization of CLN3 in the brain and neuronal cells. We have analysed here CLN3 in the mouse brain using in situ hybridization, immunohistochemical staining and western blot analysis of subcellular fractions. As visual deterioration is the hallmark of Batten disease we have set up primary retinal cultures from the mouse and analysed both endogenous mouse CLN3 and Semliki Forest virus-mediated human CLN3 localization using immunofluorescence staining and confocal microscopy. We demonstrate that CLN3 is abundantly expressed in neuronal cells, especially in the cortex, hippocampus and cerebellum of the adult mouse brain. Furthermore, our results indicate that in neurons CLN3 is not solely a lysosomal protein. It is localized in the synaptosomes but, interestingly, is not targeted to the synaptic vesicles. The novel localization of CLN3 directs attention towards molecular alterations at the synapses. This should yield important clues about the mechanisms of neurodegeneration in Batten disease.     

The Batten disease gene product (CLN3p) is a Golgi integral membrane protein

Batten disease (juvenile neuronal ceroid lipofuscinosis) is a recessive neurodegenerative disorder of childhood. The gene, CLN3, was recently identified and found to encode a novel 438 amino acid protein of unknown function. In order to gain insight into the function of the Batten disease protein (CLN3p), we investigated its subcellular localization. Protein constructs incorporating CLN3p fused to the green fluorescence protein or an eight amino acid peptide tag were transiently expressed in fibroblasts, HeLa and COS-7 cells. A juxtanuclear, asymmetric localization pattern was observed that correlated with the Golgi apparatus in all three cell types. However, a proportion of transiently transfected cells exhibited a punctate vesicular distribution throughout the cytoplasm in addition to or without the Golgi localization. In order to account for localization patterns arising from intracellular protein transport disruption due to exaggerated overexpression in transiently transfected cells, we isolated a stably transfected cell line expressing only one copy of the CLN3 -GFP DNA construct. Fluorescence and biochemical analyses using this cell line demonstrated that CLN3p is an integral membrane protein that localizes primarily in the Golgi apparatus. The functional implications of this finding are discussed.      

Batten disease: evaluation of CLN3 mutations on protein localization and function

Juvenile neuronal ceroid lipofuscinosis (JNCL), Batten disease, is an autosomal recessive lysosomal storage disease associated with mutations in CLN3. CLN3 has no known homology to other proteins and a function has not yet been described. The predominant mutation in CLN3 is a 1.02 kb genomic deletion that accounts for nearly 85% of the disease alleles. In this mutation, truncation of the protein by a premature stop codon results in the classical phenotype. Additional missense and nonsense mutations have been described. Some missense substitutions result in a protracted phenotype, with delays in the onset of classical clinical features, whereas others lead to classical JNCL. In this study, we examined the effect of naturally occurring point mutations on the intracellular localization of CLN3 and their ability to complement the CLN3-deficient yeast, btn1-Delta. We also examined a putative farnesylation motif thought to be involved in CLN3 trafficking. All of the point mutations, like wild-type CLN3, were highly associated with lysosome-associated membrane protein II in non-neuronal cells and with synaptophysin in neuronal cell lines. In the yeast functional assay, point mutations correlating with a mild phenotype also demonstrated CLN3 activity, whereas the mutations associated with severe disease failed to restore CLN3 function completely. CLN3 with a mutation in the farnesylation motif trafficked normally but was functionally impaired. These data suggest that these clinically relevant point mutations, causative of Batten disease, do not affect protein trafficking but rather exert their effects by impairing protein function.     

Intracellular trafficking of the JNCL protein CLN3

Juvenile neuronal ceroid lipofuscinosis is a lysosomal storage disease that causes visual impairment, progressive mental deterioration, and eventually death. A predominant 1.02-kb deletion as well as other mutations have been described in the CLN3 gene. Lacking significant identity with proteins of known function and no overt targeting signals within the primary amino acid sequence, accurate predictions of the intracellular location and function could not be made. Further, recent conflicting reports identified CLN3 as either a lysosomal or a mitochondrial protein. Transfection experiments using native and epitope-tagged fusion proteins were evaluated to help delineate CLN3 localization. We confirmed by immunohistochemistry and brefeldin A treatment that NH2-terminal green fluorescence protein (GFP)-CLN3 fusion proteins were retained in the Golgi apparatus, with no colocalization with mitochondrial markers. Anti-CLN3 antibodies directed against amino acids 67-90 of CLN3 were generated and shown to be specific for a 50-kDa protein in HEK 293 cells and GFP-CLN3 in transfected cells. However, cells transfected with nontagged CLN3 or carboxyl-terminal-tagged CLN3 were not immunoreactive with anti-CLN3 antibodies, suggesting that normally, the amino terminus interacts with other molecules. Thus, tags on the NH2-terminus probably inhibited these interactions and movement of CLN3 from the Golgi to more distal compartments. Also, CLN3 tagged at the COOH-terminus with either GFP or FLAG epitopes were retained in the ER, indicating a role for the COOH-terminus in trafficking. Taken together, these data confirm that CLN3 traffics through the ER and Golgi.     

Lysosomal localization of the neuronal ceroid lipofuscinosis CLN5 protein

The Finnish variant late infantile neuronal ceroid lipofuscinosis (vLINCL) belongs to the neuronal ceroid lipofuscinosis group of common recessively inherited neurodegenerative disorders. The CLN 5 gene responsible for this brain disorder codes for a novel protein with no homology to previously reported proteins. In this study, we have investigated the biosynthesis and intracellular localization of this protein in transiently transfected BHK-21 cells using a CLN5-specific peptide antibody. Confocal immunofluorescence microscopy showed that wild-type CLN5 is predominantly targeted to lysosomes and immunoprecipitation analysis recognized a 60 kDa polypeptide. The molecular weight of this protein was reduced to 40 kDa by deglycosylation with Endo H and to 38 kDa with PNGase F. The same-sized glycosylated polypeptides were also observed in the media, suggesting that the 60 kDa glycosylated CLN5 polypeptide represents a soluble lysosomal glycoprotein, not an integral transmembrane protein as predicted earlier. The most common human vLINCL mutation blocked the lysosomal targeting of expressed polypeptides. This would imply that the pathogenesis of vLINCL would be associated with the defective lysosomal trafficking, preventing the normal biological function of the corresponding polypeptide.     

Expression studies of CLN3 protein (battenin) in fusion with the green fluorescent protein in mammalian cells in vitro

The gene for Batten disease, the CLN3 gene, encodes a novel, highly hydrophobic, multitransmembrane protein, predicted to consist of 438 amino acid residues. We have expressed a full-length CLN3 protein in fusion with green fluorescent protein in various cell lines to provide its initial biochemical characterization and subcellular localization. By using Western blotting, Percoll density gradient fractionation, and Triton X-114 extraction, we demonstrate that the product of the CLN3 gene, which we call battenin, in mammalian expression system studied is a highly glycosylated protein of lysosomal membrane. In addition our data suggest that CLN3 protein is processed proteolytically in acidic compartments of the cell. Thus, battenin represents the novel constituent of a growing family of lysosomal membrane proteins.     

Tissue expression and subcellular localization of CLN3, the Batten disease protein

Juvenile neuronal ceroid lipofuscinosis (Batten disease) is a progressive neurologic disorder which results from mutations in the CLN3 gene, which normally produces a 48-kDa polypeptide of unknown function. To help characterize the CLN3 protein, we have studied its tissue distribution and subcellular localization in human tissues using three epitope-specific polyclonal antibodies to human CLN3 by immunoblot, immunocytochemical, and immunoelectron microscopic analysis. The most abundant CLN3 protein expression was in the gray matter of the brain, where it was localized to astrocytes, capillary endothelium, and neurons. CLN3 was also evident in peripheral nerve, in pancreatic islet cells, and within the seminiferous tubules in the testis. Staining was generally diffuse within the cytoplasm with some nuclear reactivity. Subcellular localization identified the CLN3 protein within the nucleus and along cell membranes. These results were contrasted with the cellular distribution of palmitoyl-protein thioesterase (PPT), the enzyme whose deficiency is responsible for infantile neuronal ceroid lipofuscinosis (CLN1). PPT was most abundant in brain and visceral macrophages where it displayed a coarse granular staining pattern typical of lysosomal distribution. Immunoelectron microscopy confirmed that PPT immunoreactivity was limited to lysosomes.     

Neuronal ceroid lipofuscinoses and possible pathogenic mechanism

The neuronal ceroid lipofuscinoses (NCLs) consist of eight autosomal recessively inherited storage disorders characterized by lysosomal inclusions of autofluorescent lipofuscins and rapid neurodegenerative progression. The NCLs include eight forms that result from genetic deficiency on genes CLN(1) to CLN(8), respectively: four classic forms with clinical onset at varying ages-infantile (INCL), late-infantile (LINCL), juvenile (JNCL), and adult (ANCL)-and four variants of late-infantile onset-the Finnish variant LINCL (fLINCL), Portuguese variant LINCL (pLINCL), Turkish variant LINCL (tLINCL), and progressive epilepsy with mental retardation (EPMR). The genes CLN(1) and CLN(2) have been characterized to encode lysosomal hydrolytic enzymes, but CLN(3), CLN(5), and CLN(8) encode transmembranous proteins with unknown function. Although clinical and pathological abnormalities have been recognized to be similar in all eight forms, the molecular mechanism explaining NCL pathogenesis remains unclear. In this review, the molecular basis for NCLs and a possible pathogenic mechanism are discussed.     

Update of the mutation spectrum and clinical correlations of over 360 mutations in eight genes that underlie the neuronal ceroid lipofuscinoses

The neuronal ceroid lipofuscinoses (NCLs) are clinically and genetically heterogeneous neurodegenerative disorders. Most are autosomal recessively inherited. Clinical features include a variable age of onset, motor and mental decline, epilepsy, visual loss, and premature death. Mutations in eight genes (PPT1/CLN1, TPP1/CLN2, CLN3, CLN5, CLN6, MFSD8/CLN7, CLN8) have been identified and several more are predicted to exist, including two provisionally named CLN4 and CLN9. Despite excessive in vitro and in vivo studies, the precise functions of the NCL proteins and the disease mechanisms remain elusive. To date 365 NCL-causing mutations are known, with 91 novel disease-causing mutations reported. These are reviewed with an emphasis on their complex correlation to phenotypes. Different mutations within the NCL spectrum can cause variable disease severity. The NCLs exemplify both phenotypic convergence or mimicry and phenotypic divergence. For example, mutations in CLN5, CLN6, MFSD8, or CLN8 can underlie the clinically similar late infantile variant NCL disease. Phenotypic divergence is exemplified by different CLN8 mutations giving rise to two very different diseases, the mild CLN8 disease, EPMR (progressive epilepsy with mental retardation), and the more severe CLN8 disease, late infantile variant. The increase in the genetic understanding of the NCLs has led to improved diagnostic approaches, and the recent proposal of a new nomenclature.     

Topology and Membrane Anchoring of the Lysosomal Storage Disease‐Related Protein CLN5

One late infantile variant of the neurodegenerative disease neuronal ceroid lipofuscinosis (NCL) is caused by a mutation in the CLN5 gene. CLN5 encodes a lysosomal glycoprotein whose structure and function have not yet been clearly defined. In the present study, we used epitope‐tagged CLN5 to determine the topology and solubility of the CLN5 protein. Our results indicated that CLN5 is synthesized as a type II transmembrane (TM) glycoprotein with a cytoplasmic N‐terminus, one TM segment, and a large luminal C‐terminal domain containing an amphipathic helix (AH). The cytoplasmic and TM domains were rapidly removed following signal‐peptide cleavage, and the resulting mature CLN5 was tightly associated with the lumen of the membrane through the AH. CLN5 pathological mutants deprived of AH lose their membrane association, are retained in the endoplasmic reticulum, and are rapidly degraded by the proteasomal machinery. We experimentally define the topology of CLN5 and demonstrate the existence of an AH that anchors the protein to the membrane. Our work sheds light on the basic properties of CLN5 required to better understand its biological functions and involvement in NCL pathogenesis.     

Novel mutations in the CLN6 gene causing a variant late infantile neuronal ceroid lipofuscinosis

The neuronal ceroid lipofuscinoses (NCLs) are a heterogeneous group of autosomal recessive neurodegenerative diseases comprising Batten and other related diseases plus numerous variants. They are characterized by progressive neuronal cell death. The CLN6 gene was recently identified, mutations in which cause one of the variant late infantile forms of NCL (vLINCL). We describe four novel mutations in the CLN6 gene. This brings the total number of CLN6 mutations known to 11 in 38 families. This suggests that the CLN6 gene may be highly mutable. An American patient of Irish/French/Native American origin was heterozygous for a 4-bp insertion (c.267_268insAACG) in exon 3. The other allele had a point mutation (c.898T>C) in exon 7 resulting in a W300R amino acid change. Two Trinidadian siblings of Indian origin were homozygous for a mutation at the 5' donor splice site of exon 4 (IVS4+1G>T), affecting the first base of the invariant GT at the beginning of intron 4. The fourth novel mutation, a double deletion of 4 bp and 1 bp in exon 7 (c.829_832delGTCG;c.837delG), was identified in a Portuguese patient heterozygous for the I154del Portuguese CLN6 mutation. Four of the 11 mutations identified are in exon 4. Three Portuguese patients with clinical profiles similar to CLN6 patients without defects in CLN6 or other known NCL genes are described. We conclude the following: 1) the CLN6 gene may be a highly mutable gene; 2) exon 4 must code for a segment of the protein crucial for function; 3) vLINCL disease in Portugal is genetically heterogeneous; 4) the I154del accounts for 81.25% of affected CLN6 Portuguese alleles; and 5) three vLINCL Portuguese patients may have defects in a new NCL gene.     

Pathogenic mutations cause rapid degradation of lysosomal storage disease‐related membrane protein CLN6

One variant form of late infantile neuronal ceroid lipofuscinosis is an autosomal recessive inherited neurodegenerative lysosomal storage disorder caused by mutations in the CLN6gene. The function of the polytopic CLN6 membrane protein localized in the endoplasmic reticulum is unknown. Here we report on expression studies of three mutations (c.368G>A, c.460‐462delATC, c.316insC) found in CLN6 patients predicted to affect transmembrane domain 3 (p.Gly123Asp), cytoplasmic loop 2 (p.Ile154del) or result in a truncated membrane protein (p.Arg106ProfsX26), respectively. The rate of synthesis and the stability of the mutant CLN6 proteins are reduced in a mutation‐dependent manner. None of the mutations prevented the dimerization of the CLN6 polypeptides. The particularly rapid degradation of the p.Arg106ProfsX26 mutant which is identical with the mutation in the murine orthologue Cln6 gene in the nclf mouse model of the disease, can be strongly inhibited by proteasomal and partially by lysosomal protease inhibitors. Both degradative pathways seem to be sufficient to prevent the accumulation/aggregation of the mutant CLN6 polypeptides in the endoplasmic reticulum. © 2009 Wiley‐Liss, Inc.     

Variant late infantile neuronal ceroid lipofuscinosis in a subset of Turkish patients is allelic to Northern epilepsy

Childhood-onset neuronal ceroid lipofuscinoses (NCL) are a group of autosomal recessive progressive encephalopathies characterized by the accumulation of autofluorescent material in various tissues, notably in neurons. Based on clinical features, the country of origin of patients, and the molecular genetic background of the disorder, at least seven different forms are thought to exist. Northern epilepsy is a novel form of NCL so far described only in Finland, where all patients are homozygous for a missense mutation in the CLN8 gene. A variant form of late infantile NCL (vLINCL) present in Turkish patients has been considered a distinct clinical and genetic entity among the NCL, the underlying gene (CLN7) being unknown. Recently, we reported homozygosity over the Northern epilepsy CLN8 gene region on 8p23 in four out of five Turkish vLINCL families studied. However, no common mutation in CLN8 was found in these families. We have now extended the Turkish vLINCL family panel to 18 families, of which only one is nonconsanguineous. Nine families were excluded from CLN8 by lack of homozygosity. In the remaining families, four CLN8 gene mutations were identified indicating that in a subset of patients with Turkish vLINCL, the disorder is allelic to Northern epilepsy. There is no apparent genotype-phenotype correlation among the Turkish patients with CLN8 mutations, although their phenotype is distinct from that of Finnish Northern epilepsy patients. The molecular genetic background of the Turkish vLINCL families not linked to CLN8 remains to be clarified.     

Motifs within the CLN3 protein: modulation of cell growth rates and apoptosis

Juvenile Batten disease (JNCL) is an autosomal recessive disease that results from mutations in the CLN3 gene. The wild-type CLN3 gene coding sequence has 15 exons, and the translated protein consists of 438 amino acids. The most commonly observed mutation is a 1.02 kb deletion in the genomic DNA. This deletion results in a truncated protein due to the loss of amino acids 154-438, and the introduction of 28 novel amino acids at the c-terminus. We demonstrate that, compared to normal controls, CLN3-deficient immortalization of lymphoblasts homozygous for this deletion grow at a slower rate, and show increased sensitivity to etoposide-induced apoptosis, supporting the notion that CLN3 may negatively regulate apoptosis. Using immortalized JNCL lymphoblast cell lines as a model system, we assess the effects of specific CLN3 mutations on cell growth rates and protection from etoposide-induced apoptosis. Protection from etoposide-induced apoptosis occurs and the cell growth rate is restored following transfection of JNCL lymphoblasts with mutant CLN3 cDNA that includes exons 11 or 13. We show that deletion of the glycosylation sites 71NQSH74 and 310NTSL313, and also mutations within the highly conserved amino acid stretches 184WSSGTGGAGLLG195, 291VYFAE295 and 330VFASRSSL337, result in slowed growth and susceptibility to apoptosis.     

Studies of membrane association of CLN3 protein

The product of the CLN3 gene is a novel protein of unknown function. Simulations using amphiphacy algorithms have shown that structurally CLN3 may be another candidate for the family of membranous proteins. Signals controlling intracellular targeting of many membrane proteins are present as short sequences within their cytoplasmic domains. In fact, the sequence of CLN3 protein contains several such signaling sequences, which are conserved among mammals. First, at the N-terminus, potential N-myristoylation motif is present. Second, the C-terminal part of CLN3 protein contains both the dileucine motif, which is a potential lysosomal targeting signal, and the prenylation motif. There is scanty evidence of lysosomal and/or mitochondrial localization of CLN3 protein. However, the question of where the functional site of the cln3 protein exists in vivo remains unanswered. From theoretical calculations, we hypothesized that CLN3 should be an integral part of the membranous micro-environment. First, to test this hypothesis, we initiated detergent-partitioning experiments, localizing CLN3 predominantly in a pool of membranous protein. Further studies have shown that CLN3 protein integrates spontaneously with cellular membranes. Second, based on the prenylation results of CLN3 protein in vitro, we discussed the possible topological consequences of C-terminal fragment of CLN3 protein.     

CLN2/TPP1 deficiency: the novel mutation IVS7-10A>G causes intron retention and is associated with a mild disease phenotype

The classical form of late infantile neuronal ceroid lipofuscinosis (LINCL) is a childhood hereditary neurodegenerative disease usually fatal in the first decade of life. The underlying gene, CLN2, encodes the lysosomal soluble enzyme tripeptidyl-peptidase 1 (TPP1). In a Portuguese patient with juvenile form of the disease, the histochemical study revealed the presence of curvilinear inclusions typical of LINCL. In vitro TPP1 activity was deficient in patient's cells. CLN2 gene analysis revealed the transition IVS7-10A>G (g.4196A>G) in both alleles. In silico analysis suggested that A-to-G change in the A-rich region of intron 7 could cause aberrant splicing of exon 8 by creating a novel acceptor splice site. However, because the wild-type acceptor of intron 7 is weak and it was not apparently affected, the severity of this mutation could not be established through sequencing data of gDNA. Normal level of spliced CLN2/mRNA was observed in patient's fibroblasts. In the cDNA, the 9-nt retention of intronic sequence (c.886_887ins9) was observed. The mutation is predicted to result in a protein with three extra amino acids between proline 295 and glycine 296. In patient's fibroblasts the level of mutant CLN2p was reduced to about 60% but the migration pattern was similar to the wild-type protein, suggesting that it was correctly targeted to the lysosomes. Taken together, these findings suggest that the first "ag" is selected for splicing and the mutant protein must retain some residual catalytic activity, thus explaining the late onset and the delayed progression of the disease.