Molecular Medicine

, Volume 17, Issue 11–12, pp 1253–1261 | Cite as

Analysis of Potential Biomarkers and Modifier Genes Affecting the Clinical Course of CLN3 Disease

  • Anne-Hélène Lebrun
  • Parisa Moll-Khosrawi
  • Sandra Pohl
  • Georgia Makrypidi
  • Stephan Storch
  • Dirk Kilian
  • Thomas Streichert
  • Benjamin Otto
  • Sara E. Mole
  • Kurt Ullrich
  • Susan Cotman
  • Alfried Kohlschütter
  • Thomas Braulke
  • Angela SchulzEmail author
Open Access
Research Article


Mutations in the CLN3 gene lead to juvenile neuronal ceroid lipofuscinosis, a pediatric neurodegenerative disorder characterized by visual loss, epilepsy and psychomotor deterioration. Although most CLN3 patients carry the same 1-kb deletion in the CLN3 gene, their disease phenotype can be variable. The aims of this study were to (i) study the clinical phenotype in CLN3 patients with identical genotype, (ii) identify genes that are dysregulated in CLN3 disease regardless of the clinical course that could be useful as biomarkers, and (iii) find modifier genes that affect the progression rate of the disease. A total of 25 CLN3 patients homozygous for the 1-kb deletion were classified into groups with rapid, average or slow disease progression using an established clinical scoring system. Genome-wide expression profiling was performed in eight CLN3 patients with different disease progression and matched controls. The study showed high phenotype variability in CLN3 patients. Five genes were dysregulated in all CLN3 patients and present candidate biomarkers of the disease. Of those, dual specificity phosphatase 2 (DUSP2) was also validated in acutely CLN3-depleted cell models and in CbCln3Δex7/8 cerebellar precursor cells. A total of 13 genes were upregulated in patients with rapid disease progression and downregulated in patients with slow disease progression; one gene showed dysregulation in the opposite way. Among these potential modifier genes, guanine nucleotide exchange factor 1 for small GTPases of the Ras family (RAPGEF1) and transcription factor Spi-B (SPIB) were validated in an acutely CLN3-depleted cell model. These findings indicate that differential perturbations of distinct signaling pathways might alter disease progression and provide insight into the molecular alterations underlying neuronal dysfunction in CLN3 disease and neurodegeneration in general.


The neuronal ceroid lipofuscinoses (NCLs) are the most common inherited neurodegenerative disorders mainly affecting children. Typical symptoms comprise progressive visual loss, psychomotor deterioration and epilepsy. NCLs are incurable lysosomal storage diseases, and lysosomal accumulation of autofluorescent storage material, called ceroid lipofuscin, can be detected in almost all tissues (1). The most dramatically affected organ is the brain, which undergoes massive neuronal cell atrophy and cell loss (2).

One of the most frequently occurring NCL forms in Northern European countries is caused by mutations in the CLN3 gene (CLN3 disease, also called juvenile NCL or Batten disease). CLN3 disease starts at the age of 4–6 years with a progressive retinopathy leading to blindness. After several years, dementia, epilepsy and motor disturbances ensue. The diagnostic hallmark of this NCL type are conspicuous vacuoles in the cytoplasm of lymphocytes (3).

CLN3 encodes a polytopic membrane protein of 438 amino acids (4). To date, >40 different mutations in CLN3 have been described (summarized in a mutation database: CLN3 disease is recessive, and 85% of CLN3 patients carry a 1-kb deletion that causes the loss of exons 7 and 8 (5). Although the CLN3 protein is well conserved from yeast to humans, its function is a matter of debate. There are reports that CLN3 is involved in lysosomal acidification, arginine import, autophagy and apoptosis (6).

To understand the pathomechanisms underlying CLN3 disease, any natural variability in the clinical course of CLN3 disease first has to be assessed. We applied a clinical scoring system to 25 CLN3 patients who were homozygous for the 1-kb deletion and classified them into patients with slow, average and rapid disease progression. To investigate whether the clinical variability can be correlated with changes in gene expression, we performed comparative genome-wide microarray analyses in lymphocytes of eight selected CLN3 patients with different rates of disease progression. We identified both genes that are dysregulated in CLN3 disease regardless of the clinical course that might be useful as biomarkers of the disease and genes for which expression correlated with clinical variability, possibly indicating differential expression of modifier genes. The general significance of selected genes of both sets was confirmed in a cell model system in which CLN3 expression was depleted by small interfering RNA (siRNA).

Materials and Methods

Distribution of CLN3 Patients

A total of 25 patients, 11 male and 14 female, participated in this study. Among the patients were three pairs of siblings. The mean age of the patients (corresponding to the mean time period of assessment) was 18.4 ± 6.2 years (mean ± standard deviation [SD], range 12–34 years).

Clinical Scoring

The patients were analyzed according to a previously established clinical scoring system (7). Clinical scoring and calculation of an Index of Relative Severity (Table 1) are described in the supplementary material.
Table 1

Distribution of CLN3 patients according to slow, average or rapid disease progression defined by the Index of Relative Severity.

Clinical course

Index of Relative Severity

Number of patients





−0.5 to +0.5





Patients and Sample Isolation

Eight patients diagnosed with CLN3 disease and classified as having rapid (n = 2), average (n = 4) and slow disease progression (n = 2), and all homozygous for the 1-kb deletion in the CLN3 gene, were recruited from the NCL specialty clinic at the University Medical Center Hamburg-Eppendorf (Table 2, patients 1–8). These eight patients did not receive anticonvulsive medication; therefore, they were particularly suitable for providing material to be used for microarray analysis. Two of the patients with average disease progression were siblings from one family, and one patient with average and one patient with rapid disease progression were siblings from another family. Seven age- and gender-matched controls were included in the microarray study.
Table 2

Clinical course, age and gender distribution of the studied CLN3 patients who did not receive anticonvulsants.

Clinical course


Age (years)


Index of Relative Severity









































aSiblings family 1.

bSiblings family 2.

In addition, lymphocytes from three patients with CLN3 disease, 14, 13 and 14 years old (P14M, male; P13F, female; P14F, female), who received anticonvulsive medication and who were all homozygous for the 1-kb deletion in the CLN3 gene, and age- and gender-matched controls (C14M, C13F and C14F) were used for RNA isolation for real-time polymerase chain reaction (RT-PCR) to validate dual specificity phosphatase 2 (DUSP2) upregulation. Patient P14M was classified as having average disease progression, patient P13F as slow disease progression, and patient P14F as rapid disease progression with Indexes of Relative Severity of 0.3 (P14M), 0.8 (P13F), and −1.0 (P14F). Patient P14M received valproate, patient P13F topiramate and patient P14F valproate and lamotrigine as anticonvulsive medications.

The patients’ or parents’ consent for participation in this study was obtained according to the Declaration of Helsinki (1991). The study was approved by the medical ethics committee of the Ärztekammer Hamburg. Lymphocytes were prepared from fresh patient blood samples by Ficoll-gradient centrifugation (Biocoll®; Biochrom, Berlin, Germany) according to the manufacturer’s protocol and used for RNA isolation (RNeasy® Micro Kit; Qiagen, Hilden, Germany).

Microarray Analysis

Procedures for RNA extraction, cDNA synthesis, labeling and hybridization were carried out according to the manufacturer’s protocol (Affymetrix, High Wycombe, UK), followed by microarray data analyses, as described in the supplementary material.

Cell Culture and RNA Preparation

HeLa cells were provided by K. von Figura (University of Göttingen, Göttingen, Germany) and were cultured in Dulbecco’s minimal essential medium containing 10% fetal bovine serum and antibiotics. SH-SY5Y cells were provided by E. Kramer, Center for Molecular Neurobiology Hamburg (ZMNH) (Hamburg, Germany) and were cultured in Dulbecco’s minimal essential medium containing 20% fetal bovine serum, antibiotics and 1 × GlutaMAX™. CbCln3Δex7/8 cerebellar precursor cells were cultured as previously described (8). RNA was extracted from the cells using an RNeasy® Mini Kit (Qiagen) according to the manufacturer’s protocol.


For RT-PCR, TaqMan® Gene Expression Assays (Applied Biosystems, Darmstadt, Germany) including predesigned probes and primer sets were used. For details, see the supplementary material.

Other Methods

cDNA synthesis and posttranscriptional gene silencing with siRNA for human CLN3 in HeLa cells and SH-SY5Y cells was performed as previously described (9).

All supplementary materials are available online at .


Clinical Course and Classification of CLN3 Disease Progression

A total of 25 patients, 11 male and 14 female, participated in this study. The mean age of the patients (corresponding to the mean time period of assessment) was 18.4 ± 6.2 years (mean ± SD; range 12–34 years). Scoring of vision, intellect, language and motor function was performed, both retrospectively and prospectively, at 6-month intervals starting at birth.

First clinical signs in these studied CLN3 patients were observed at age 5.4 ± 1.4 years (mean ± SD; range 1.5–7 years), which for the majority of patients (n = 17) was visual loss. However, in a few patients, other symptoms preceded the visual deterioration such as behavioral problems (n = 5), motor problems (n = 1), language problems (n = 1) and seizures (n = 1). Interestingly, the age of occurrence of these atypical first clinical signs was comparable to the age of onset of visual loss in the remaining 17 patients: 5.8 ± 1.1 years (mean ± SD; range 3.5–7 years).

Figure 1 summarizes the clinical courses of all 25 patients over time for each problem category, indicating the median and the range between the 10th and 90th percentile. By calculating an Index of Relative Severity (7), we classified 5, 9 and 11 CLN3 patients into groups with rapid, average and slow disease progression, respectively (see Table 1), reflecting a high phenotype variability in CLN3 patients despite an identical genetic defect.
Figure 1

Clinical scoring of 25 CLN3 patients homozygous for the 1-kb deletion in the CLN3 gene. Scoring for each problem category was made as follows: normal function (3); problem slight, but readily recognized (2); problem severe (1); total loss of function (0). The range between the 10th and 90th percentile is represented by the shaded area. The black line represents the median.

Figure 2 compares the disease progression of two patients (1 and 2) with rapid versus two patients (3 and 4) with slow disease progression (see Table 2). None of the four patients received anticonvulsants. Patients 1 and 2 had scorings for motor function, language and intellect that decreased rapidly over time, following the 10th percentile for intellect and language and being briefly beyond the 10th percentile for motor function. For patient 1 (solid red line), regression of visual function was similar to the average disease course until age 10 years and led to complete blindness 1 year later. In contrast, in patient 2, regression of visual function appeared to be slow, with scoring mostly following the 90th percentile.
Figure 2

Comparison of clinical scoring of CLN3 patients with rapid (n = 2) versus slow disease progression (n = 2). Clinical scoring is shown in red for patients with rapid disease progression (patient 1, solid line; patient 2, broken line) and in blue for patients with slow disease progression (patient 3, solid line; patient 4, broken line). The shaded area represents the range between the 10th and 90th percentile calculated from clinical scoring of 25 patients who are homozygous for the 1-kb deletion in the CLN3 gene (including the two patients with rapid and slow disease progression).

Patients 3 and 4 had scorings for motor function, language and intellect mainly on or above the 90th percentile. The regression of visual function was similar to (patient 3, solid blue line), or started even earlier (patient 4, broken blue line) than, the two patients with rapid progression. Patient 3 was diagnosed with sinus bradycardia and received a pacemaker implant at age 30 years. Six months after this procedure, language and motor function each improved to score 1, which remains to date. Even though scoring of epilepsy has not been used to classify these patients, it should be noted that this patient has not had any detectable seizure to date and has a normal electroencephalogram reading.

For patient 4, scoring for intellect, language and motor function mainly remained on or above the 90th percentile until age 14.5 years and then deteriorated slowly. Like patient 3, this patient was diagnosed with sinus bradycardia and received a pacemaker implant at age 26 years. Since then, neither motor nor intellect or language deteriorated further, remaining constant to date at age 34 years.

Dysregulated Genes in CLN3 Patients

In addition to the CLN3 patients 1–4, four patients with average progression of the disease (patients 5–8) were selected whose clinical course was not influenced by anticonvulsants (see Table 2). To gain insight into molecular processes possibly related to the different disease courses in CLN3 patients, we performed RNA microarray analyses in freshly isolated lymphocytes. These cells are easily accessible, and since no culturing is required, they are not affected by subsequent handling. The entire set of the original microarray data obtained is accessible at the Gene Expression Omnibus database (

A total of 89 genes were dysregulated in patients with average disease progression compared with healthy controls (Supplementary Figure 1). Using gene ontology tools to categorize the dysregulated genes (10), the encoded proteins were predicted to play a wide variety of cellular roles (Supplementary Table 1). In patients with rapid and slow progression, 643 and 839 genes were dysregulated, respectively (Supplementary Figure 1, Supplementary Tables 2A and 3A). Of these dysregulated genes, 89 and 121 genes, respectively, encode proteins involved in functions of the nervous system, lysosomes and proteolysis, protein transport and autophagy, apoptosis and visual perception (Supplementary Tables 2B and 3B).
Table 3

List of potential modifier genes with opposite dysregulation in patients with rapid versus slow disease progression.


Fold change

Gene symbol

Gene name




Kruppel-like factor 6




Myristoylated alanine-rich protein kinase C substrate




Hypothetical protein LOC283663




B lymphoid tyrosine kinase




Fc fragment of IgG, low affinity IIb, receptor (CD32)/Fc fragment of IgG, low affinity IIc, receptor for (CD32)




Spi-B transcription factor (Spi-1/PU. 1 related)




Membrane-spanning 4-domains, subfamily A, member 1




Rap guanine nucleotide exchange factor (GEF) 1




Zinc finger protein 718




β-Site APP-cleaving enzyme 2




Hypothetical protein LOC283663




G protein-coupled receptor 109B




Chronic lymphocytic leukemia upregulated 1




Protein SOLO



In patients with rapid or slow disease progression, 105 genes were dysregulated that were not changed in patients with average disease progression. Of these, 13 genes were upregulated in patients with rapid disease progression and downregulated in patients with slow disease progression. One gene was identified that was upregulated in patients with slow disease progression and downregulated in patients with rapid disease progression (Table 3).

A total of five genes were significantly dysregulated in all CLN3 patients (Table 4). Of these, regulator of G protein signaling 1 (RGS1), DUSP2 and poly(ADP-ribose) polymerase family member 15 (PARP15) were significantly upregulated in all three CLN3 patient groups. Two genes, DNA-directed RNA polymerase II polypeptide J-related (POLR2J2) and CDC42 small effector 2 (CDC42SE2), were significantly downregulated in all three CLN3 patient groups.
Table 4

List of genes that were significantly dysregulated in all CLN3 patients.


Fold change

Gene symbol

Gene name





DNA-directed RNA polymerase II polypeptide J-related





CDC42 small effector 2





Regulator of G protein signaling 1





Dual specificity phosphatase 2





Poly(ADP-ribose) polymerase family member 15




To verify the significance of the biomarker candidates, the relative mRNA expression of DUSP2 was determined in lymphocytes derived from three CLN3 patients under anticonvulsive treatment with average, slow and rapid disease progression (P14M, P13F and P14F, respectively) and age- and gender-matched controls (Figure 3A). The mRNA expression level of DUSP2 was upregulated 2.2- to 5.6-fold in extracts from all three patients cells, indicating that DUSP2 upregulation is not influenced by anticonvulsive medication.
Figure 3

Relative mRNA expression of DUSP2. (A) The relative mRNA expression of DUSP2 in lymphocytes derived from CLN3 patients receiving anticonvulsive medication compared with age- and gender-matched controls was measured by RT-PCR and normalized to RPS18 expression. The patients were 14, 13 and 14 years old (P14M, male; P13F, female; P14F, female) and were all homozygous for the 1-kb deletion in the CLN3 gene. They were classified as having average (P14M), slow (P13F) and rapid disease progression (P14F). (B) The relative mRNA expression of Dusp2 in CbCln3Δex7/8 cerebellar precursor cells compared with wild-type cells was measured by RT-PCR and normalized to Actb expression. Data are means ± SD of three independent RT-PCR experiments. *P < 0.05 compared with controls; **P < 0.01 compared with controls; ***P < 0.001 compared with controls.

In addition, we tested DUSP2 mRNA expression in CbCln3Δex7/8 cerebellar precursor cells derived from Cln3Δex7/8 knock-in mice that represent precise genetic replicas of the mutation that causes disease in most CLN3 patients: the 1-kb deletion of exons 7 and 8 (8). DUSP2 mRNA expression was upregulated 3.5- to 12.3fold in CbCln3Δex7/8 cerebellar precursor cells compared with controls (Figure 3B). These data show that upregulation of DUSP2 mRNA expression is not restricted to CLN3 patient lymphocytes but is also detected in murine neuronal cells that harbor the exact same genetic defect like the CLN3 patients studied here.

Expression of Candidate Genes in Acutely CLN3-Depleted Cells

To further prove the general significance of our results, we explored whether acute loss of CLN3 in two different cell models affects expression of the same genes that are altered in lymphocytes from patients. mRNA expression of several candidate genes was analyzed in HeLa and SH-SY5Y neuronal cell lines treated with CLN3 siRNA. Quantitative RT-PCR revealed a down-regulation of CLN3 mRNA expression of at least 97% in HeLa cells and 62% in SH-SY5Y cells under the conditions used (Figures 4A, B) (11). The relative mRNA expression of DUSP2, for which expression was 2.4-fold upregulated in all CLN3 patients (see Table 4), was increased in both cell types ranging from 1.4- to 2.8-fold in CLN3-depleted HeLa cells and 1.4- to 2.9-fold in SH-SY5Y cells compared with controls (see Figures 4A, B). The CDC42SE2 mRNA level, which was decreased (1.8- to 3.2-fold) in CLN3 patient cells, showed a slight reduction ranging from 0.6- to 0.9-fold in CLN3-depleted HeLa cells compared with controls (Supplementary Figure 2). The alterations in the amounts of mRNA of three other genes (POLR2J2, RGS1 and PARP15) were similar in CLN3-depleted HeLa and control cells.
Figure 4

Relative mRNA expression of candidate genes in siRNA-mediated acutely CLN3-depleted cells. HeLa cells or SH-SY5Y cells were transfected with CLN3 siRNA for 72 h in total. The relative mRNA expressions of candidate genes were measured by RT-PCR and normalized to ACTB expression. mRNA levels of scrambled siRNA-treated cells (white bar) were set as 1 and used as control. (A) Relative mRNA expression of CLN3, and DUSP2 in CLN3 siRNA-treated HeLa cells. (B) Relative mRNA expression of CLN3, and DUSP2 in CLN3 siRNA-treated SH-SY5Y cells. (C) Relative mRNA expression of CLN3, RAPGEF1 and SPIB in CLN3 siRNA-treated HeLa cells. Data are mean ± SD of three (B) or five (A, C) independent transfection experiments. **P< 0.01 compared with controls; ***P< 0.001 compared with controls.

Next, mRNA expression of 5 of the 14 genes that appear to modify the progression rate of CLN3 disease were analyzed in HeLa cells after acute siRNA-mediated CLN3 depletion (Figure 4C). Four of the genes tested were upregulated in CLN3 patients with rapid disease progression and downregulated in patients with slow disease progression (RAPGEF1 [guanine nucleotide exchange factor 1 for small GTPases of the Ras family], SPIB (transcription factor Spi-B), MARCKS [myristoylated alanine-rich protein kinase C substrate], BACE2 [β-site APP {β-amyloid precursor protein} cleaving enzyme 2]), and one was dysregulated in the opposite way (FLJ10357). Quantitative RT-PCR of RAPGEF1 revealed a significant increase of mRNA expression ranging from 1.7- to 2.1-fold in CLN3-depleted cells compared with controls. SPIB mRNA level was also significantly increased (range 1.7- to 3.6-fold). mRNA levels of MARCKS, BACE2 and FLJ10357 were not significantly altered compared with controls.


High Phenotype Variability in CLN3 Patients with Identical Genotype

This study shows that the natural variability in CLN3 disease caused by the same 1-kb deletion in the CLN3 gene is wide and independent of genotype. We performed a detailed analysis of the natural history of 25 CLN3 patients homozygous for the 1-kb deletion over a mean period of about 18 years, using an established scoring system and an Index of Relative Disease Severity (7). Of these patients, 36%, 44% and 20% exhibited an average, rapid and slow progression of the disease, respectively.

Previous studies also described a variable phenotype, but for these patients, genetic diagnosis was not possible, since the CLN3 gene had not been identified at that time (11,12). In contrast, a study by Adams et al. (13) analyzing only the behavioral phenotype of a heterogeneous cohort of CLN3 patients did not find differences between these CLN3 patients. Loss of vision was previously described to occur within a narrow age range in juvenile NCL patients (7). We confirm this for our patients, including those whose clinical progression in other categories (intellect, language and motor function) varied widely.

Interestingly, both patients with slow disease progression (patients 3 and 4) received pacemaker implants because of sinus bradycardia at age 30 and 26 years, respectively. Scoring of these two patients remained stable (patient 4) or even improved (patient 3) after treatment of the cardiac problem by a pacemaker implant. Other groups also reported on few CLN3 patients with cardiac pathology (14,15). By electrocardiography, significant rhythm abnormalities, the most frequent being sinus bradycardia and sinus arrhythmia, were found in 16 of our 25 CLN3 patients. Ten of the 25 CLN3 patients examined had abnormal results in echocardiographic examinations, with left ventricular hypertrophy being the most common finding. The mean age of CLN3 patients with pathologic cardiac findings was 17.7 ± 7.3 years (mean ± SD, range 11–34 years) (A Schulz, unpublished results). These results might point to additional system involvement in NCL. Whether cardiac involvement in CLN3 patients is a common phenomenon and how it might influence the course of the disease needs to be studied in more detail. However, these observations suggest that regular cardiologic examinations are warranted in CLN3 patients.

The observation of a wide clinical variability in CLN3 patients with the same mutation (apart from the conformity of the visual loss) is intriguing. To exclude any effect of anticonvulsants on the general progression of the disease (16), we selected eight patients who did not receive anticonvulsants over the entire assessment period. Scoring of these patients classified them into four individuals with average disease progression and two each with slow and rapid disease progression.

The present and previous studies suggest that genetic variations in addition to the defect in the CLN3 gene are responsible for the rapid or slow progression rate in CLN3 patients. To identify genetic variations responsible for this clinical variability, we performed genome-wide microarray analyses in lymphocytes derived from patients 1 to 8 with strikingly different disease progression. Lymphocytes are ideal as the source of the required RNA because (i) they contain large vacuoles with lysosomal storage material characteristic for CLN3 disease, indicating a specific disease phenotype effect in this tissue (17,18), and (ii) the rapid isolation of lymphocytes within a time frame of 5 min after taking the blood sample prevents or reduces any secondary effects on gene expression patterns that might be induced by cell culture conditions and repeated passaging if cells such as fibroblasts from a skin biopsy were used.

Identification of a Potential New Biomarker for CLN3 Disease

Five genes were found to be dysregulated in all three patient groups. The expression of POLR2J2 and CDC42SE2 were decreased, whereas RGS1, DUSP2 and PARP15 were increased, in all CLN3 patients compared with controls. Of these, DUSP2 was studied in more detail because of its (i) high dysregulation and (ii) potential relevance for brain function. DUSP2 dephosphorylates both phosphotyrosine and phosphoserine/threonine residues simultaneously within the mitogen-activated protein kinases ERK (extracellular signal regulated kinase), p38 and JNK (c-jun N-terminal kinase) and thereby acts as an antagonist of associated signaling cascades (19). It has been reported that overexpression of DUSP2 increases susceptibility to p53-mediated cell death induced by oxidative stress (20). In mice, seizures can induce DUSP2 expression in microglial cells and dying neurons of the hippocampus (21). Although DUSP2 was found to be upregulated in cells of CLN3 patients not receiving anticonvulsive medication, a biomarker should also be dysregulated in patients under medication. Therefore, DUSP2 mRNA expression was also tested in CLN3 patients receiving anti-convulsive medication, where it was again found to be significantly upregulated regardless of which group of disease progression these patients belonged. Moreover, DUSP2 mRNA expression was slightly higher in patients under anticonvulsive medication compared with untreated patients, suggesting that seizures do not significantly influence DUSP2 upregulation in CLN3 patient lymphocytes.

Of note, expression of DUSP2 already serves as a biomarker of therapeutic responsiveness, since it is associated with poor outcome in ovarian cancer patients (22). DUSP2 upregulation in CbCln3Δex7/8 cerebellar precursor cells harboring the same genetic defect like most CLN3 patients shows that both neuronal and nonneuronal regulation of DUSP2 is affected in CLN3 disease. This result is also supported by the result that acute and short-term induced CLN3 depletion leads to DUSP2 upregulation, not only in HeLa cells, but also in SH-SY5Y neuronal cells.

DUSP2 expression was increased in lymphocytes of all CLN3 patient groups without and with medication as well as in murine neuronal cells harboring the patients genetic CLN3 defect and in both acutely CLN3-depleted nonneuronal and neuronal cells. This DUSP2 expression increase makes DUSP2 suitable as a potential biomarker of CLN3 disease.

The mRNA expression of another member of the dual-specificity phosphatases group, DUSP7, was increased 1.7-fold in embryonic, primary neuron cultures of Cln3-deficient mice in comparison with control neuronal cells (23). These data support our findings and suggest members of the DUSP family as likely candidates linked to CLN3 deficiency. Detailed analysis of the specificity of DUSP2 expression in CLN3 disease and its regulatory function may contribute to our understanding of the molecular mechanisms of inflammatory responses and cell survival pathways in CLN3 disease. In addition, the identification of biomarkers for CLN3 disease will support the identification of therapeutic targets for experimental therapies and will provide valuable tools to monitor the effect of such new treatment strategies.

Previously, we found that ACP2 encoding the lysosomal acid phosphatase was significantly upregulated at the mRNA and protein level in CLN3-deficient mouse brain and patient fibroblasts (9). The basal mRNA expression of ACP2 in lymphocytes, however, appears to be too low to allow detection of alterations between CLN3 patients and controls.

Identification of Potential Modifier Genes for CLN3 Disease

We also identified genes that were dysregulated in opposite directions in CLN3 patients with rapid and slow progression of the disease and that may therefore represent disease-modifier genes. Thirteen genes were upregulated in patients with rapid disease progression and downregulated in patients with slow disease progression. One gene only was dysregulated in the opposite direction, suggesting that the course of the disease is determined more by genes that aggravate rather than by genes that reduce deterioration. Expression of two of the genes upregulated in patients with rapid disease progression and downregulated in patients with slow disease progression was evaluated in CLN3 siRNA-treated HeLa cells: RAPGEF1, which is relevant for brain function, and SPIB, which is a lymphocyte-specific gene.

RAPGEF1 protein plays a role in multiple signal transduction pathways that regulate growth, differentiation and neuronal migration processes during brain development (23,25,26). RapGef1-deficient mouse embryos exhibit a cortical neuron migration defect resulting in a failure to split the preplate into marginal zone and subplate and a failure to form a cortical plate (27). Quantitative RT-PCR of RAPGEF1 in CLN3-depleted cells confirmed the significant increase in mRNA expression in CLN3 patients with rapid progression.

Expression analysis of SPIB, which is involved in the regulation of development and differentiation of T- and B-lymphocytes (28,29), was of interest for two reasons. First, changes in the expression of lymphocyte-specific genes such as SPIB are not surprising because pathological storage products are found both in T- and B-lymphocytes of CLN3 patients (30,31). However, it remains to be examined whether the differentiation of B-lymphocytes to plasma cells and T-cell-dependent immune responses are differentially impaired in CLN3 patients with rapid and slow progression of disease. Second, in CLN3-depleted cells, SPIB showed the highest and most significantly increased mRNA level of the genes tested.

Together, these findings suggest that CLN3 depletion causes increases in the expression of genes that were also found to be dysregulated in patients with rapid progression but decreased in patients with slow progression. Additional dysregulated genes listed in Table 4 have yet to be evaluated to demonstrate that CLN3 depletion in HeLa cells truly represents a useful cell model to follow molecular alterations in CLN3 patients with rapid rather than slow progression of the disease. However, of note, the expression of FCGR2B, which is increased in CLN3 patients with rapid disease progression and decreased in patients with slow disease progression, is also elevated in brain tissues of mouse models of the more severe infantile CLN1 disease and variant late infantile CLN5 disease, suggesting that some of these genes may be general markers for NCL (32).

Several comparative gene expression analyses have been performed in whole brain, cerebellum and eyes of presymptomatic Cln3-deficient mice with targeted deletion of exons 1–6 or in embryonic cultured neuronal cells of these mice (33, 34, 35). More than 800 genes were found to be altered. The physiological relevance of these genes in the pathogenesis of neurodegeneration or for the function of CLN3 has not yet been proven, nor can they explain differences in the progression rates of human CLN3 disease.

In contrast to studies performed in NCL mouse models, which are not limited by the number of samples, our small sample number is due to the rareness of the CLN3 disease and especially that we included patients who were without any medication. Because of the progressive course of a disease with multiple neurologic symptoms including epilepsy, few patients who are unmedicated exist at any one time. Therefore, we used this to guide our approach for statistical analysis. To compensate for the small sample size, a sophisticated preanalytical selection of control samples was performed to reduce the number of degrees of freedom by gender and age and in turn the biological background noise. Considering the rareness of the disease, the challenge of finding unmedicated patients and matching these with defined age and gender in controls, this study represents a significant step toward the identification of potential biomarker and modifier genes in juvenile CLN3 disease.

To conclude, our study showed a high phenotype variability in CLN3 patients with an identical genotype. Among five genes identified by microarray analysis as dysregulated in all CLN3 patients regardless of the progression of the disease, DUSP2 is a potential and much-needed biomarker of the disease. Furthermore, 13 genes were found that appear to modify the progression of the disease. Among these potential modifier genes, RAPGEF1 and SPIB were validated in an acutely CLN3-depleted cell model. These findings indicate that differential perturbations of distinct signaling pathways might alter the progression. Future longitudinal studies are required to evaluate the regulation of these identified biomarker or modifier genes during the different stages of the disease. Finally, this project provides valuable insights into the molecular alterations that underlie neuronal dysfunction in juvenile CLN3 disease, and the genes identified in the work reported here should become a focus for future studies.


The authors declare that they have no competing interests as defined by Molecular Medicine, or other interests that might be perceived to influence the results and discussion reported in this paper.



We thank the affected and control families for participating in this study and Johannes Brand for technical assistance. This work was supported by the Deutsche Forschungsgemeinschaft (grant SCHU1597/2–1) and the parent organizations Nächstenliebe e.V. and NCL-Gruppe Deutschland e.V.

Supplementary material

10020_2011_17111253_MOESM1_ESM.pdf (1.8 mb)
Supplementary material, approximately 1833 KB.


  1. 1.
    Jalanko A, Braulke T. (2009) Neuronal ceroid lipofuscinoses. Biochim. Biophys. Acta. 1793:697–709.CrossRefGoogle Scholar
  2. 2.
    Haltia M. (2003) The neuronal ceroid-lipofuscinoses. J. Neuropathol. Exp. Neurol. 62:1–13.CrossRefGoogle Scholar
  3. 3.
    Williams RE, et al. (2006) Diagnosis of the neuronal ceroid lipofuscinoses: an update. Biochim. Biophys. Acta. 1762:865–72.CrossRefGoogle Scholar
  4. 4.
    International Batten Disease Consortium. (1995) Isolation of a novel gene underlying Batten disease, CLN3. Cell. 82:949–57.CrossRefGoogle Scholar
  5. 5.
    Munroe PB, et al. (1997) Spectrum of mutations in the Batten disease gene, CLN3. Am. J. Hum. Genet. 61:310–6.CrossRefGoogle Scholar
  6. 6.
    Kyttälä A, Lahtinen U, Braulke T, Hofmann SL. (2006) Functional biology of the neuronal ceroid lipofuscinoses (NCL) proteins. Biochim. Biophys. Acta. 1762:920–33.CrossRefGoogle Scholar
  7. 7.
    Kohlschütter A, Laabs R, Albani M. (1988) Juvenile neuronal ceroid lipofuscinosis (JNCL): quantitative description of its clinical variability. Acta. Paediatr. Scand. 77:867–72.CrossRefGoogle Scholar
  8. 8.
    Fossale E, et al. (2004) Membrane trafficking and mitochondrial abnormalities precede subunit c deposition in a cerebellar cell model of juvenile neuronal ceroid lipofuscinosis. BMC Neurosci. 5:57.CrossRefGoogle Scholar
  9. 9.
    Pohl S, et al. (2007) Increased expression of lysosomal acid phosphatase in CLN3-defective cells and mouse brain tissue. J. Neurochem. 103:2177–88.CrossRefGoogle Scholar
  10. 10.
    Ashburner M, et al. (2000) Gene ontology: tool for the unification of biology. The Gene Ontology Consortium. Nat. Genet. 25:25–9.CrossRefGoogle Scholar
  11. 11.
    Lou HC, Kristensen K. (1973) A clinical and psychological investigation into juvenile amaurotic idiocy in Denmark. Dev. Med. Child Neurol. 15:313–23.CrossRefGoogle Scholar
  12. 12.
    Sorensen JB, Parnas J. (1979) A clinical study of 44 patients with juvenile amaurotic family idiocy. Acta. Psychiatr. Scand. 59:449–61.CrossRefGoogle Scholar
  13. 13.
    Adams HR, et al. (2010) Genotype does not predict severity of behavioural phenotype in juvenile neuronal ceroid lipofuscinosis (Batten disease). Dev. Med. Child. Neurol. 52:637–53.CrossRefGoogle Scholar
  14. 14.
    Hofman IL, van der Wal AC, Dingemans KP, Becker AE. (2001) Cardiac pathology in neuronal ceroid lipofuscinoses: a clinicopathologic correlation in three patients. Eur. J. Paediatr. Neurol. 5:213–7.CrossRefGoogle Scholar
  15. 15.
    Tomiyasu H, et al. (2000) An autopsy case of juvenile neuronal ceroid-lipofuscinosis with dilated cardiomyopathy. Rinsho. Shinkeigaku. 4:350–7.Google Scholar
  16. 16.
    Kohlschütter A, Schulz A. (2009) Towards understanding the neuronal ceroid lipofuscinoses. Brain Dev. 31:499–502.CrossRefGoogle Scholar
  17. 17.
    Anderson G, Smith VV, Malone M, Sebire NJ. (2005) Blood film examination for vacuolated lymphocytes in the diagnosis of metabolic disorders; retrospective experience of more than 2,500 cases from a single centre. J. Clin. Pathol. 58:1305–10.CrossRefGoogle Scholar
  18. 18.
    Mole SE, Williams RE, Goebel HH. (2005) Correlations between genotype, ultrastructural morphology and clinical phenotype in the neuronal ceroid lipofuscinoses. Neurogenetics. 6:107–26.CrossRefGoogle Scholar
  19. 19.
    Patterson KI, Brummer T, O±Brien PM, Daly RJ. (2009) Dual-specificity phosphatases: critical regulators with diverse cellular targets. Biochem. J. 418:475–89.CrossRefGoogle Scholar
  20. 20.
    Yin Y, Liu YX, Jin YJ, Hall EJ, Barrett JC. (2003) PAC1 phosphatase is a transcription target of p53 in signalling apoptosis and growth suppression. Nature. 422:527–31.CrossRefGoogle Scholar
  21. 21.
    Boschert U, Muda M, Camps M, Dickinson R, Arkinstall S. (1997) Induction of the dual specificity phosphatase PAC1 in rat brain following seizure activity. Neuroreport. 8:3077–80.CrossRefGoogle Scholar
  22. 22.
    Givant-Horwitz V, et al. (2004). The PAC-1 dual specificity phosphatase predicts poor outcome in serous ovarian carcinoma. Gynecol. Oncol. 93:517–32.CrossRefGoogle Scholar
  23. 23.
    Luiro K, et al. (2006) Batten disease (JNCL) is linked to disturbances in mitochondrial, cytoskeletal, and synaptic compartments. J. Neurosci. Res. 84:1124–38.CrossRefGoogle Scholar
  24. 24.
    Knudsen BS, Feller SM, Hanafusa H. (1994) Four proline-rich sequences of the guanine-nucleotide exchange factor C3G bind with unique specificity to the first Src homology 3 domain of Crk. J Biol. Chem. 269:32781–7.PubMedGoogle Scholar
  25. 25.
    Tanaka S, et al. (1994) C3G, a guanine nucleotide-releasing protein expressed ubiquitously, binds to the Src homology 3 domains of CRK and GRB2/ASH proteins. Proc. Natl. Acad. Sci. U. S. A. 91:3443–7.CrossRefGoogle Scholar
  26. 26.
    D±Arcangelo G, et al. (1995) A protein related to extracellular matrix proteins deleted in the mouse mutant reeler. Nature. 374:719–23.CrossRefGoogle Scholar
  27. 27.
    Voss AK, et al. (2008) C3G regulates cortical neuron migration, preplate splitting and radial glial cell attachment. Development. 135:2139–49.CrossRefGoogle Scholar
  28. 28.
    Schotte R, Nagasawa M, Weijer K, Spits H, Blom B. (2004) The ETS transcription factor Spi-B is required for human plasmacytoid dendritic cell development. J. Exp. Med. 200:1503–9.CrossRefGoogle Scholar
  29. 29.
    Anderson MK. (2006) At the crossroads: diverse roles of early thymocyte transcriptional regulators. Immunol. Rev. 209:191–211.CrossRefGoogle Scholar
  30. 30.
    Aula P, Rapola J, Andersson LC. (1975) Distribution of cytoplasmic vacuoles in blood T and B lymphocytes in two lysosomal disorders. Virchows Arch. B Cell Pathol. 18:263–71.PubMedGoogle Scholar
  31. 31.
    Bruck W, Goebel HH, Dienes P. (1991) B and T lymphocytes are affected in lysosomal disorders: an immunoelectron microscopic study. Neuropathol. Appl. Neurobiol. 17:219–22.CrossRefGoogle Scholar
  32. 32.
    von Schantz C, et al. (2008) Brain gene expression profiles of Cln1 and Cln5 deficient mice unravels common molecular pathways underlying neuronal degeneration in NCL diseases. BMC Genomics. 9:146.CrossRefGoogle Scholar
  33. 33.
    Elshatory Y, et al. (2003) Early changes in gene expression in two models of Batten disease. FEBS Lett. 538:207–12.CrossRefGoogle Scholar
  34. 34.
    Chattopadhyay S, et al. (2004) Altered gene expression in the eye of a mouse model for batten disease. Invest. Ophthalmol. Vis. Sci. 45:2893–905.CrossRefGoogle Scholar
  35. 35.
    Brooks AI, Chattopadhyay S, Mitchison HM, Nussbaum RL, Pearce DA. (2003) Functional categorization of gene expression changes in the cerebellum of a Cln3-knockout mouse model for Batten disease. Mol. Genet. Metab. 78:17–30.CrossRefGoogle Scholar

Copyright information

© The Feinstein Institute for Medical Research 2011

Authors and Affiliations

  • Anne-Hélène Lebrun
    • 1
  • Parisa Moll-Khosrawi
    • 1
  • Sandra Pohl
    • 1
  • Georgia Makrypidi
    • 1
  • Stephan Storch
    • 1
  • Dirk Kilian
    • 1
  • Thomas Streichert
    • 2
  • Benjamin Otto
    • 2
  • Sara E. Mole
    • 3
  • Kurt Ullrich
    • 1
  • Susan Cotman
    • 4
  • Alfried Kohlschütter
    • 1
  • Thomas Braulke
    • 1
  • Angela Schulz
    • 1
    Email author
  1. 1.Children’s HospitalUniversity Medical Center Hamburg-Eppendorf Martinistrasse 52HamburgGermany
  2. 2.Array Service CenterUniversity Medical Center Hamburg-EppendorfHamburgGermany
  3. 3.Medical Research Council Laboratory for Molecular Cell Biology, Molecular Medicine Unit, UCL Institute of Child Health and Department of Genetics, Evolution and EnvironmentUniversity College LondonLondonUK
  4. 4.Center for Human Genetic ResearchMassachusetts General HospitalBostonUSA

Personalised recommendations