NCLs (neuronal ceroid lipofuscinoses), a group of inherited neurodegenerative lysosomal storage diseases that predominantly affect children, are the result of autosomal recessive mutations within one of the nine cln genes. The wild-type cln gene products are composed of membrane and soluble proteins that localize to the lysosome or the ER (endoplasmic reticulum). However, the destiny of the Cln variants has not been fully characterized. To explore a possible link between ER quality control and processing of Cln mutants, we investigated the fate of two NCL-related Cln6 mutants found in patient samples (Cln6G123D and Cln6M241T) in neuronal-derived human cells. The point mutations are predicted to be in the putative transmembrane domains and most probably generate misfolded membrane proteins that are subjected to ER quality control. Consistent with this paradigm, both mutants underwent rapid proteasome-mediated degradation and complexed with components of the ER extraction apparatus, Derlin-1 and p97. In addition, knockdown of SEL1L [sel-1 suppressor of lin-12-like (Caenorhabditis elegans)], a member of an E3 ubiquitin ligase complex involved in ER protein extraction, rescued significant amounts of Cln6G123D and Cln6M241T polypeptides. The results implicate ER quality control in the instability of the Cln variants that probably contributes to the development of NCL.

INTRODUCTION

NCLs (neuronal ceroid lipofuscinoses), also referred to as Batten disease, are the most common paediatric neurodegenerative diseases having an incidence rate of 1 in 12500 in the U.S. population [1]. Their significant clinical feature is progressive encephalopathy confined to macular cerebral degeneration [2]. The symptoms of NCL encephalopathy include epileptic seizures, dementia, progressive psychomotor decline, visual retrocession and blindness, and ultimately leads to premature death [3]. The histological hallmark of NCLs is the accumulation of proteolipid pigment deposits in the lysosomes of many cell types [4,5]. Patients suffering from Batten disease show intracellular inclusions predominately in neurons caused by lysosomal accumulation of ceroid and lipofuscin with mitochondrial ATP synthase subunit C [6].

Most NCLs are inherited as autosomal recessive mutations within six identified and three unidentified genes, denoted cln1–9. Defective Cln proteins result in lysosomal dysfunction and accumulation of lysosomal storage material in the brain of affected children, leading to the development of several pathogenic processes and neurodegeneration. Current genetic classification of NCLs varies in clinical phenotypes from more benign to extremely severe, depending on the specific genetic defect of the encoded protein. They are divided into four main groups: infantile NCL, late infantile NCL, juvenile JNCL and adult NCL, and four other less frequent entities. The Cln polypeptides are both soluble and membrane proteins located in either endosomal/lysosomal compartments (Cln1, palmitoyl–protein thioesterase; Cln2, tripeptidyl peptidase I; Cln3 and Cln5) or the ER (endoplasmic reticulum) (Cln6 and Cln8) [7]. Cln4, Cln7 and Cln9 proteins have not yet been identified. Interestingly, the processing and fate of aberrant Cln proteins is not clear. In fact, hereditary mutations within cellular genes can result in clinically significant pathologies, such as cystic fibrosis, hereditary emphysema [8], nephrogenic diabetes insipidus [9] and oculocutaneous albinism [10]. Since ER quality control has been implicated in these diseases, it is likely that it plays a role in the processing of Cln variants and the development of NCL.

Mutations within the cln6 gene cause both the classical late-infantile and juvenile forms of NCL [11]. The present study explores the role of ER quality control in the early processing events of ER-localized Cln6 mutant proteins [1214]. Experiments revealed that Cln6 mutant proteins were degraded in a proteasome-dependent manner. In addition, knockdown of SEL1L [sel-1 suppressor of lin-12-like (Caenorhabditis elegans)], a protein involved in the disposal of aberrant ER proteins, rescued the degradation of the Cln6 mutants. The results supports the theory that ER quality control may contribute to the onset of NCL type 6 and offers insight towards potential therapy against NCL.

EXPERIMENTAL

Cell lines and antibodies

Human U373-MG astrocytoma cells expressing Cln6 polypeptides (U373Cln6wt, U373Cln6G123D and U373Cln6M241T cells) (where wt is wild-type) were generated and maintained as described previously [15]. Anti-PDI (protein disulfide-isomerase) [16], anti-SEL1L [17] and anti-calnexin {AF8 [18], a gift from Professor M. Brenner (Department of Medicine, Brigham and Women's Hospital, Harvard Medical School, Boston, MA, U.S.A.} antibodies were utilized as described previously. Anti-GAPDH (glyceraldehyde-3-phosphate dehydrogenase), anti-p97, anti-Derlin-1 and anti-BiP (immunoglobulin heavy-chain-binding protein) antibodies were purchased from Chemicon, Sigma and Stressgen respectively.

cDNA constructs

The Cln6 chimaeras were cloned from the Cln6 DNA template (A.T.C.C. Image ID 3878776) with an N-terminus HA (haemagglutinin) epitope tag (AYPYDVPDYA) into the retroviral vector pLpCX (Clontech). Cln6G123D and Cln6M241T mutant cDNAs were generated from PCR fragments by site-directed mutagenesis [15].

Immunoprecipitation

Immunoprecipitations were performed as described previously [15]. Briefly, cells (106) were lysed in 0.5% NP-40 (Nonidet P40) lysis mix, followed by incubation with the respective antibody and protein A–agarose beads. The immunoprecipitates were resolved using SDS/PAGE (12.5% gels) and subjected to immunoblot analysis using the respective immunoglobulin.

Pulse–chase analysis

Pulse–chase experiments were performed as described previously [15]. Cells were labelled with [35S]methionine and chased in non-radioactive methionine (25 mM). Cln6 proteins were recovered from NP-40 cell lysates using an anti-HA antibody and resolved using SDS/PAGE (12.5% gels). The polyacrylamide gel was dried, exposed to autoradiographic film and the polypeptides were quantified by densitometry analysis using an Alpha Imager 3400 densitometer.

Immunofluorescence microscopy

Cells were fixed with 1:1 (v/v) methanol/acetone solution and incubated with blocking solution [1% BSA and 0.5% cold water fish gelatin (Sigma) in PBS (pH 7.2)], followed by the respective antibody. The cells were then incubated with the respective FITC-conjugated anti-mouse and Texas Red-conjugated anti-rabbit immunoglobulins (Molecular Probes). Fluorescence microscopy was carried out using an Olympus IX70/IX-FLA inverted fluorescence microscope and a Sony DKC-5000 digital camera. Images were created using Adobe Photoshop software.

RESULTS

Cln6 mutant molecules are stabilized by inclusion of proteasome inhibition

Cln6 is a 311 amino acid non-glycosylated ER membrane protein that is predicted to contain either five or six transmembrane domains [12,14]. To investigate the processing of Cln6 mutants implicated in the development of NCL, an HA epitope tag was introduced at the N-terminus of Cln6wt and Cln6 variants (Cln6G123D and Cln6M241T) with missense mutations in the predicted transmembrane domains [19]. These mutants were chosen for study because they were identified in NCL patients [1214]. These polar residues within the membrane bilayer most probably cause the mutant to exist in a misfolded conformation, generating a potential target for the ER quality-control apparatus. HA-epitope tagged versions of Cln6wt, Cln6G123D and Cln6M241T were transduced into U373 cells (U373Cln6wt, U373Cln6G123D and U373Cln6M241T cells) and the stability of the Cln6 proteins was examined by pulse–chase analysis (Figure 1A). U373Cln6wt, U373Cln6G123D and U373Cln6M241T cells were metabolically labelled for 15 min and chased for up to 45 min. Equivalent levels of Cln6wt polypeptides were recovered throughout the chase period (Figure 1A, lanes 1–4), a result that is consistent with protein stability. In contrast, levels of Cln6G123D polypeptides decreased rapidly during the chase period (Figure 1A, lanes 5–8) with a half-life of ∼20 min (Figure 1B). Interestingly, the migration difference of Cln6G123D compared with Cln6wt or Cln6M241T is probably due to a conformational change in the overall structure that is specific to the mutation. A similar result was observed in mutants of the multi-spanning membrane protein Sec61p and the soluble protein p97 [20,21]. Even more striking was the instability of Cln6M241T polypeptides (Figure 2A, lanes 9–12). Only a minute amount of Cln6M241T polypeptides were recovered from the 0 min chase period and even less from the 15 min chase point, suggesting a half-life of less than 15 min (Figure 2A, lane 10). These results demonstrate that Cln6 mutants are degraded with fast kinetics.

Cln6G123D and Cln6M241T are degraded with rapid kinetics

Figure 1
Cln6G123D and Cln6M241T are degraded with rapid kinetics

(A) U373Cln6wt, U373Cln6G123D and U373Cln6M241T cells were labelled with [35S]methionine for 15 min and chased for up to 45 min. Cln6 polypeptides were immunoprecipitated using an anti-HA antibody and resolved by SDS/PAGE. The gel was exposed to autoradiographic film for 3 days. (B) The levels of Cln6 proteins were quantified by densitometry and plotted as a percentage of the 0 min chase point (100%). Molecular mass is shown at the left-hand side in (A).

Figure 1
Cln6G123D and Cln6M241T are degraded with rapid kinetics

(A) U373Cln6wt, U373Cln6G123D and U373Cln6M241T cells were labelled with [35S]methionine for 15 min and chased for up to 45 min. Cln6 polypeptides were immunoprecipitated using an anti-HA antibody and resolved by SDS/PAGE. The gel was exposed to autoradiographic film for 3 days. (B) The levels of Cln6 proteins were quantified by densitometry and plotted as a percentage of the 0 min chase point (100%). Molecular mass is shown at the left-hand side in (A).

Cln6G123D and Cln6M241T proteins are stabilized in the presence of proteasome and dislocation inhibitors

Figure 2
Cln6G123D and Cln6M241T proteins are stabilized in the presence of proteasome and dislocation inhibitors

(A) Cln6 polypeptides were recovered from U373Cln6wt (lanes 1, 2, 5, 6 and 10), U373Cln6G123D (lanes 3 and 4) and U373Cln6M241T (lanes 7 and 8) cells treated with (+) or without (−) the proteasome inhibitor ZL3VS. The anti-HA immunoprecipitates and total cell lysates from U373Cln6wt and U373 cells (lane 11) were subjected to immunoblotting using an anti-HA antibody. (B) Total cell lysates from U373Cln6wt (lanes 1, 2, 7, 8, 13 and 14), U373Cln6G123D (lanes 3, 4, 9, 10, 15 and 16) and U373Cln6M241T (lanes 5, 6, 11, 12, 17 and 18) cells treated with (+) or without (−) eeyarestatin (lanes 2, 4 and 6), lactacystin (lanes 8, 10 and 12) and ZL3VS (lanes 14, 16 and 18) were subjected to anti-HA (lanes 1–19) and anti-GAPDH (lanes 20–38) immunoblots. U373 cell lysates (lane 15) were used as a negative control. The Cln6 proteins, immunoglobulin light chain (IgG LC) and GAPDH are indicated. Molecular masses are shown at the left-hand side.

Figure 2
Cln6G123D and Cln6M241T proteins are stabilized in the presence of proteasome and dislocation inhibitors

(A) Cln6 polypeptides were recovered from U373Cln6wt (lanes 1, 2, 5, 6 and 10), U373Cln6G123D (lanes 3 and 4) and U373Cln6M241T (lanes 7 and 8) cells treated with (+) or without (−) the proteasome inhibitor ZL3VS. The anti-HA immunoprecipitates and total cell lysates from U373Cln6wt and U373 cells (lane 11) were subjected to immunoblotting using an anti-HA antibody. (B) Total cell lysates from U373Cln6wt (lanes 1, 2, 7, 8, 13 and 14), U373Cln6G123D (lanes 3, 4, 9, 10, 15 and 16) and U373Cln6M241T (lanes 5, 6, 11, 12, 17 and 18) cells treated with (+) or without (−) eeyarestatin (lanes 2, 4 and 6), lactacystin (lanes 8, 10 and 12) and ZL3VS (lanes 14, 16 and 18) were subjected to anti-HA (lanes 1–19) and anti-GAPDH (lanes 20–38) immunoblots. U373 cell lysates (lane 15) were used as a negative control. The Cln6 proteins, immunoglobulin light chain (IgG LC) and GAPDH are indicated. Molecular masses are shown at the left-hand side.

To examine whether the Cln6 mutants are degraded by the proteasome, Cln6 protein levels were examined from U373Cln6wt, U373Cln6G123D and U373Cln6M241T cells treated with or without the proteasome inhibitor ZL3VS (carboxybenzyl-leucyl-leucyl-leucyl vinyl sulfone) (10 μM) for 5 h [22]. Cln6 proteins recovered from the respective cells were analysed using an immunoprecipitation and immunoblot method (Figure 2A). U373 cells transduced with empty vector were used as a negative control (Figure 2A, lanes 9 and 11). Equivalent levels of Cln6wt were recovered independently of inhibitor treatment (Figure 2A, lanes 1, 2, 5 and 6), whereas a minimal number of polypeptides was recovered from U373Cln6G123D and U373Cln6M241T cells (Figure 2A, lanes 3 and 7). However, significant amounts of Cln6G123D and Cln6M241T polypeptides were recovered from ZL3VS-treated cells (Figure 2A, lanes 4 and 8). Because Cln6G123D is degraded with slower kinetics than Cln6M241T, we observed increased levels of Cln6G123D compared with Cln6M241T from proteasome inhibitor-treated cells (Figure 2A, lane 4 compared with lane 8). Consistent with these results, higher molecular-mass species, most likely ubiquitinated Cln6G123D and Cln6M241T (*), were recovered from proteasome inhibitor-treated cells (see Supplementary Figure S1 at http://www.bioscirep.org/bsr/029/bsr0290173add.htm). Collectively, these results verify that Cln6G123D and Cln6M241T are extracted from the ER to the cytosol and degraded by the proteasome.

To further confirm that the instability of Cln6G123D and Cln6M241T polypeptides was due to their transport out of the ER, we examined the effect of eeyarestatin, an inhibitor of the dislocation reaction [23]. U373, U373Cln6wt, U373Cln6G123D and U373Cln6M241T cells were treated with or without eeyarestatin (5 μg/ml) or the proteasome inhibitors lactacystin [24] (10 μM) or ZL3VS (5 μM) for 12 h. Total cell lysates were subjected to immunoblot analysis using an anti-HA antibody (Figure 2B). An immunoblot performed using an anti-GAPDH antibody verified equal protein loading (Figure 2B, lanes 20–38). As expected, treatment with the proteasome inhibitors lactacystin and ZL3VS caused a significant increase in both Cln6G123D and Cln6M241T polypeptides (Figure 2B, lanes 10, 12, 16 and 18 respectively). Moreover, eeyarestatin treatment caused increased levels of Cln6G123D and Cln6M241T polypeptides (Figure 2B, lanes 4 and 6), albeit to a lower level than the proteasome inhibitors. This is probably due to the inability of eeyarestatin to completely attenuate dislocation of ER substrates [23]. Eeyarestatin treatment caused a greater accumulation of Cln6G123D than Cln6M241T, probably due to differences in their degradation kinetics (Figure 1B). In summary, Cln6G123D and Cln6M241T polypeptides were stabilized upon inclusion of proteasome or dislocation inhibitors.

Stabilized Cln6 mutant proteins are localized to the ER

The localization of the stabilized Cln mutant Cln6G123D was examined using fluorescence microscopy (Figure 3). U373Cln6wt and U373Cln6G123D cells treated with or without ZL3VS (2.5 μM) for 12 h were probed for Cln6 or PDI polypeptides using anti-HA (Figure 3, panels 1–2 and 7–8) and anti-PDI (Figure 3, panels 3–4 and 9–10) antibodies, respectively. The merged images represent the co-localized polypeptides (Figure 3, panels 5–6 and 11–12). The intensity of the Cln6wt fluorescence signal was independent of proteasome-inhibitor treatment (Figure 3, panels 1–2) and co-localized with the ER resident protein PDI (Figure 3, panels 5–6). In contrast, a significant increase in the fluorescence signal was observed in U373Cln6G123D cells treated with proteasome inhibitor (Figure 3, compare panels 7 and 8). The ER localization of Cln6G123D polypeptides upon proteasome inhibition was confirmed by their co-localization with PDI (Figure 3, panel 12). In addition, a subcellular fractionation experiment [25] revealed that the majority of Cln6G123D and Cln6M241T mutant polypeptides was observed from the membrane fraction upon proteasome inhibitor treatment (see Supplementary Figure S2 at http://www.bioscirep.org/bsr/029/bsr0290173add.htm). These results demonstrate that stabilized Cln6 mutants are localized to the ER.

Stabilized Cln6 mutants localize to the ER

Figure 3
Stabilized Cln6 mutants localize to the ER

U373Cln6wt (panels 1–6) and U373Cln6G123D (panels 7–12) cells treated with (+) or without (−) the proteasome inhibitor ZL3VS were probed with anti-HA (panels 1, 2, 7 and 8) and anti-PDI (panels 3, 4, 9 and 10) antibodies and subjected to indirect immunofluorescence microscopy. The yellow regions in the Merge panels (panels 5, 6, 11 and 12) represent the co-localized material.

Figure 3
Stabilized Cln6 mutants localize to the ER

U373Cln6wt (panels 1–6) and U373Cln6G123D (panels 7–12) cells treated with (+) or without (−) the proteasome inhibitor ZL3VS were probed with anti-HA (panels 1, 2, 7 and 8) and anti-PDI (panels 3, 4, 9 and 10) antibodies and subjected to indirect immunofluorescence microscopy. The yellow regions in the Merge panels (panels 5, 6, 11 and 12) represent the co-localized material.

Cln6G123D and Cln6M241T polypeptides interact with the dislocation complex

The dislocation of ER proteasomal substrates is mediated by the interaction with both ER and cytosolic proteins. Derlin-1, an ER membrane protein essential for the dislocation of certain ER degradation substrates, together with the cytosolic complex of AAA (ATPase associated with various cellular activities) p97–Ufd1–Npl4, facilitates the extraction of ER proteins out of the ER [26]. To further define the processing of the Cln6G123D and Cln6M241T mutant proteins, we examined whether Cln6G123D and Cln6M241T form a complex with the specific components of the dislocation apparatus, Derlin-1 and p97. Derlin-1 was recovered from U373Cln6wt, U373Cln6G123D and U373Cln6M241T cells treated with or without ZL3VS (10 μM) for 5 h using an anti-Derlin-1 antibody. The immunoprecipitates and total cell lysates were subjected to immunoblotting using anti-HA (Figure 4A, lanes 1–14) and anti-Derlin-1 (Figure 4A, lanes 15–28) antibodies. As expected, Derlin-1 polypeptides were recovered from all samples (Figure 4A, lanes 15–28). Consistent with previous results (Figure 2), Cln6G123D and Cln6M241T were stabilized in the presence of a proteasome inhibitor (Figure 4A, lanes 11 and 13). Both Cln6G123D and Cln6M241T polypeptides were co-precipitated with Derlin-1 molecules from proteasome inhibitor-treated U373Cln6G123D and U373Cln6M241T cells (Figure 4A, lanes 4 and 6). The lack of a polypeptide signal from U373 and U373Cln6wt cells (Figure 4A, lanes 1, 2 and 7) confirmed the specificity of the anti-Derlin-1 immunoprecipitation. These results suggest that Derlin-1 is probably involved in the dislocation of the Cln6G123D and Cln6M241T mutants.

Cln6 mutants complex with the dislocation apparatus

Figure 4
Cln6 mutants complex with the dislocation apparatus

(A) Derlin-1 was recovered from U373, U373Cln6wt, U373Cln6G123D and U373Cln6M241T cells treated with (+) or without (−) the proteasome inhibitor ZL3VS using an anti-Derlin-1 antibody. The immunoprecipitates (lanes 1–7 and 15–21) and total cell lysates (lanes 8–14 and 22–28) were subjected to immunoblot analysis using anti-HA (lanes 1–14) and anti-Derlin-1 (lanes 15–28) antibodies. Cln6 proteins were recovered from U373Cln6wt (B and C), U373Cln6G123D (B) and U373Cln6M241T (C) cells treated with (+) or without (−) the proteasome inhibitor ZL3VS. The immunoprecipitates and U373Cln6wt total cell lysates (B and C) were subjected anti-p97 (B and C, lanes 1–5) and anti-HA (B and C, lanes 6–10) immunoblot analysis. Derlin-1, p97 and Cln6 proteins are indicated. Molecular masses are shown at the left-hand side.

Figure 4
Cln6 mutants complex with the dislocation apparatus

(A) Derlin-1 was recovered from U373, U373Cln6wt, U373Cln6G123D and U373Cln6M241T cells treated with (+) or without (−) the proteasome inhibitor ZL3VS using an anti-Derlin-1 antibody. The immunoprecipitates (lanes 1–7 and 15–21) and total cell lysates (lanes 8–14 and 22–28) were subjected to immunoblot analysis using anti-HA (lanes 1–14) and anti-Derlin-1 (lanes 15–28) antibodies. Cln6 proteins were recovered from U373Cln6wt (B and C), U373Cln6G123D (B) and U373Cln6M241T (C) cells treated with (+) or without (−) the proteasome inhibitor ZL3VS. The immunoprecipitates and U373Cln6wt total cell lysates (B and C) were subjected anti-p97 (B and C, lanes 1–5) and anti-HA (B and C, lanes 6–10) immunoblot analysis. Derlin-1, p97 and Cln6 proteins are indicated. Molecular masses are shown at the left-hand side.

We next examined a possible interaction between Cln6G123D and Cln6M241T and p97 (Figures 4B and 4C). Cln6 polypeptides recovered from U373Cln6wt, U373Cln6G123D and U373Cln6M241T cells treated with or without ZL3VS (5 μM) for 12 h were subjected to immunoblot analysis. As expected, significant amounts of Cln6G123D and Cln6M241T proteins were recovered from the proteasome inhibitor-treated cells (Figures 4B and 4C, lane 9). p97 exclusively co-precipitated with Cln6G123D and Cln6M241T from ZL3VS-treated cells (Figures 4B and 4C, lane 4), but not with Cln6wt (Figures 4A and 4B, lanes 1 and 2). Collectively, these results suggest that Derlin-1 and p97 probably mediate the extraction of Cln6G123D and Cln6M241T polypeptides from the ER into the cytosol.

Cln6G123D and Cln6M241T are dislocated in a SEL1L-dependent manner

Ubiquitination of ER degradation substrates through a specific E3 ligase is a pre-requisite for the dislocation of most degradation substrates [27]. In fact, ER localized membrane proteins that complex with E3 ligases play an important role in the dislocation process. SEL1L, an ER membrane protein that complexes with the E3 ligase Hrd1 [HMG-CoA (3-hydroxy-3-methylglutaryl-CoA) reductase degradation protein 1], Derlin-1, Derlin-2 and p97 [28,29] has been implicated in the dislocation of both membrane and soluble degradation substrates [17,30]. Is SEL1L involved in the degradation of Cln6G123D and Cln6M241T? To address this question, we examined the steady state levels of Cln6 polypeptides in cells that stably express shRNA (small-hairpin RNA) against SEL1L (shRNA–SEL1L) or GFP (green fluorescent protein) (shRNA–GFP) [17]. The shRNA–GFP cells were utilized as a negative control. U373Cln6wt, U373Cln6G123D and U373Cln6M241T cells expressing shRNA–SEL1L or shRNA–GFP were treated with the proteasome inhibitor ZL3VS (10 μM) for up to 5 h. The total cell lysates were subjected to immunoblot analysis to determine the levels of Cln6 polypeptides (Figure 5). Cln6wt levels were not altered upon the inclusion of the proteasome inhibitor or expression of shRNAs (Figure 5, lanes 1–3 and 10–12). As expected, in shRNA–GFP-expressing cells, the levels of Cln6G123D and Cln6M241T increased upon treatment with the proteasome inhibitor (Figure 5, lanes 4–6 and 7–9 respectively). Strikingly, significant levels of both Cln6G123D and Cln6M241T polypeptides were observed in SEL1L-knockdown cells (Figure 5, lanes 13–18). Since SEL1L was knocked down by only ∼50% (see Supplementary Figure S3 at http://www.bioscirep.org/bsr/029/bsr0290173add.htm), a greater attenuation of Cln6G123D and Cln6M241T degradation would probably be observed upon complete knockdown of SEL1L. The GAPDH immunoblot confirmed equal loading of the cell lysates (Figure 5, lanes 20–38). These results strongly support the hypothesis that Cln6G123D and Cln6M241T degradation is dependent on SEL1L and is regulated by ER quality control.

Knockout of SEL1L in cells attenuates Cln6G123D and Cln6M241T degradation

Figure 5
Knockout of SEL1L in cells attenuates Cln6G123D and Cln6M241T degradation

Total cell lysates from U373Cln6wt, U373Cln6G123D and U373Cln6M241T cells that express shRNA against either GFP (lanes 1–9 and 20–28) or SEL1L (lanes 10–19 and 29–38) were treated with the proteasome inhibitor ZL3VS for various incubation times (h). The samples were subjected to immunoblot analysis using anti-HA (lanes 1–19) and anti-GAPDH (lanes 20–38) antibodies. The Cln6 proteins and GAPDH are indicated. Molecular masses are shown at the left-hand side.

Figure 5
Knockout of SEL1L in cells attenuates Cln6G123D and Cln6M241T degradation

Total cell lysates from U373Cln6wt, U373Cln6G123D and U373Cln6M241T cells that express shRNA against either GFP (lanes 1–9 and 20–28) or SEL1L (lanes 10–19 and 29–38) were treated with the proteasome inhibitor ZL3VS for various incubation times (h). The samples were subjected to immunoblot analysis using anti-HA (lanes 1–19) and anti-GAPDH (lanes 20–38) antibodies. The Cln6 proteins and GAPDH are indicated. Molecular masses are shown at the left-hand side.

DISCUSSION

NCLs are neurodegenerative childhood diseases caused by mutations within nine different cln genes [31]. Cln mutants are either truncated versions of the protein or contain point mutations which yield enzymatically inactive polypeptides or molecules that never reach their final destination. The early processing events within the ER can dictate the fate of these mutant proteins. In fact, specific genetic diseases, including cystic fibrosis, emphysema, Fabri disease, Gaucher disease, GM1 gangliosidosis and Tay–Sachs disease are caused by mutations within proteins that are subjected to ER quality control [32,33]. The present study shows for the first time that the Cln6 mutants Cln6G123D and Cln6M241T implicated in the onset of NCL type 6 are unstable and degraded in a proteasome-dependent manner. Hence the development of NCL may be attributed to the instability of the mutant polypeptide, implying that ER quality control contributes to the onset of disease.

ER quality control has evolved to prevent the egress of defective membrane or secretory polypeptides from the ER. In most cases, the misfolded proteins are recognized by ER chaperones such as BiP, calnexin, and PDI [34,35] and extracted through the ER membrane via a proteinaceous channel to be degraded by the proteasome. The recruitment of the degradation substrate to the dislocon occurs through an interaction with membrane-bound Ubcs (ubiquitin-conjugating enzymes) (i.e. Ubc7, Ubc1 and Ubc6), E3 ubiquitin ligases [i.e. gp78, Hrd1/Der3, Doa10, HsHrd1 (human Hrd1), trc8 and TEB4] [27], and factors, such as the Derlin family of proteins and SEL1L. Once at the site of dislocation, p97 complexed with Npl4 and Ufd1 extracts the ubiquitin-conjugated dislocated substrate through the membrane [29,36]. The N-linked glycan attached to the polypeptide would then be removed efficiently, followed by deubiquitination and degradation by the proteasome. The mutant Cln6 proteins are degraded as classical misfolded proteins, as observed by their association with the ER chaperones calnexin and BiP (K. Oresic and D. Tortorella, unpublished work), as well as with Derlin-1 and p97 (Figure 4). Since the knockdown of SEL1L caused an increase in the levels of the Cln6 mutants, it is possible that SEL1L contributes to the instability of Cln6G123D and Cln6M241T proteins. Collectively, these results are consistent with the theory that Cln6G123D and Cln6M241T mutant proteins are recognized and disposed of by the ER dislocation/degradation apparatus.

The Cln6G123D and Cln6M241T mutants are degraded with fast kinetics (Figure 1). Given the fact that the Cln6 protein spans the membrane seven times, we would expect the degradation kinetics to be similar to the degradation kinetics of CFTR (cystic fibrosis transmembrane conductance regulator) mutants [37]. Why are these aberrant proteins degraded in such a fast manner? Since both Cln6 mutations lie within the transmembrane regions, the amino acid changes may disrupt the stability of the transmembrane domains and induce a drastically misfolded protein that is quickly discarded for degradation by the proteasome. The severity of the misfolded state of the protein can dictate directly the half-life of the polypeptide. Therefore Cln6G123D would be in a more native state than Cln6M241T. Nevertheless, the dramatic instability of the Cln6 mutants could prove to be a useful model system to study how multi-membrane spanning proteins are dislocated across the ER membrane.

The degradation of aberrant Cln6 polypeptides implies that those proteins are thwarted in their cellular function, which eventually leads to the accumulation of insoluble proteolipid material in the lysosomes. It remains to be clarified whether increasing the stability of these mutants would allow the Cln mutant to regain at least partial wt activity. Therefore proteasome inhibitors may be considered as potential therapeutics for NCL type 6 if the stabilized Cln6 mutant would retain at least partial activity. In fact, understanding the process of ER quality control in the development of specific genetic diseases has already yielded promising results in the treatment of Fabri disease caused by mutations in α-galactosidase polypeptides. Namely, 1-deoxygalactonojirimycin, a competitive inhibitor of α-galactosidase, could stabilize an α-galactosidase mutant protein and correct its trafficking defect [38]. In addition, chemical chaperones could rescue some lysosomal storage diseases caused by inherited mutations, such those mutations which cause Gaucher disease, GM1 gangliosidosis and Tay–Sachs disease [3941]. The present study serves as a preliminary guide for identifying potential new treatments against NCL that could be examined in the animal models available. A number of spontaneous animal forms of NCL have been described in sheep, dogs and mice [4244]. The mouse models of NCLs, Ppt1−/− [45], Ppt1Δex4 knockout [46], Tpp1 neoins R446H knockout [47], Cln5−/− knockout [48], Cln6nclf spontaneous mutant [49], Cln8mnd spontaneous mutant [50], Cln3−/− knockout [51], Cln3Δex7/8 knock-in [52] and Cln3 knockout [53] demonstrate phenotypes that closely resemble the major features of human NCLs. They all show clinically relevant behavioural and pathological changes that indicate a progressive neurodegenerative disorder. The value of animal models for neurodegenerative disorders lies largely in the utility of these models in evaluating potential therapeutic interventions in the future [54]

Abbreviations

     
  • BiP

    immunoglobulin heavy-chain-binding protein

  •  
  • ER

    endoplasmic reticulum

  •  
  • GAPDH

    glyceraldehyde-3-phosphate dehydrogenase

  •  
  • GFP

    green fluorescent protein

  •  
  • HA

    haemagglutinin

  •  
  • Hrd1

    HMG-CoA (3-hydroxy-3-methylglutaryl-CoA) reductase degradation protein 1

  •  
  • NCL

    neuronal ceroid lipofuscinoses

  •  
  • NP-40

    Nonidet P40

  •  
  • PDI

    protein disulfide-isomerase

  •  
  • SEL1L

    sel-1 suppressor of lin-12-like (Caenorhabditis elegans)

  •  
  • shRNA

    small-hairpin RNA

  •  
  • Ubc

    ubiquitin-conjugating enzyme

  •  
  • wt

    wild-type

  •  
  • ZL3VS

    carboxybenzyl-leucyl-leucyl-leucyl vinyl sulfone.

We thank Caroline Ng for critical analysis of the manuscript.

FUNDING

This work was supported by the National Institutes of Health [grant number AI060905, U19 AI062623]. D.T. is partially supported by the Irma T. Hirschl Trust.

References

References
Rider
 
J. A.
Rider
 
D. L.
 
Batten disease, a twenty-eight-year struggle: past, present and future
Neuropediatrics
1997
, vol. 
28
 (pg. 
4
-
5
)
Goebel
 
H. H.
Wisniewski
 
K. E.
 
Current state of clinical and morphological features in human NCL
Brain Pathol
2004
, vol. 
14
 (pg. 
61
-
69
)
Kohlschutter
 
A.
Gardiner
 
R. M.
Goebel
 
H. H.
 
Human forms of neuronal ceroid-lipofuscinosis (Batten disease): consensus on diagnostic criteria, Hamburg 1992
J. Inherit. Metab. Dis.
1993
, vol. 
16
 (pg. 
241
-
244
)
Wisniewski
 
K. E.
Rapin
 
I.
Heaney-Kieras
 
J.
 
Clinico-pathological variability in the childhood neuronal ceroid-lipofuscinoses and new observations on glycoprotein abnormalities
Am. J. Med. Genet. Suppl.
1988
, vol. 
5
 (pg. 
27
-
46
)
Prasad
 
V. V.
Pullarkat
 
R. K.
 
Brain lysosomal hydrolases in neuronal ceroid-lipofuscinoses
Mol. Chem. Neuropathol.
1996
, vol. 
29
 (pg. 
169
-
179
)
Palmer
 
D. N.
Tyynela
 
J.
van Mil
 
H. C.
Westlake
 
V. J.
Jolly
 
R. D.
 
Accumulation of sphingolipid activator proteins (SAPs) A and D in granular osmiophilic deposits in miniature Schnauzer dogs with ceroid-lipofuscinosis
J. Inherit. Metab. Dis.
1997
, vol. 
20
 (pg. 
74
-
84
)
Kyttala
 
A.
Lahtinen
 
U.
Braulke
 
T.
Hofmann
 
S. L.
 
Functional biology of the neuronal ceroid lipofuscinoses (NCL) proteins
Biochim. Biophys. Acta
2006
, vol. 
1762
 (pg. 
920
-
933
)
Qu
 
D.
Teckman
 
J. H.
Omura
 
S.
Perlmutter
 
D. H.
 
Degradation of a mutant secretory protein, α1-antitrypsin Z, in the endoplasmic reticulum requires proteasome activity
J. Biol. Chem.
1996
, vol. 
271
 (pg. 
22791
-
22795
)
Hirano
 
K.
Zuber
 
C.
Roth
 
J.
Ziak
 
M.
 
The proteasome is involved in the degradation of different aquaporin-2 mutants causing nephrogenic diabetes insipidus
Am. J. Pathol.
2003
, vol. 
163
 (pg. 
111
-
120
)
Halaban
 
R.
Svedine
 
S.
Cheng
 
E.
Smicun
 
Y.
Aron
 
R.
Hebert
 
D. N.
 
Endoplasmic reticulum retention is a common defect associated with tyrosinase-negative albinism
Proc. Natl. Acad. Sci. U.S.A.
2000
, vol. 
97
 (pg. 
5889
-
5894
)
Kida
 
E.
Golabek
 
A. A.
Wisniewski
 
K. E.
 
Cellular pathology and pathogenic aspects of neuronal ceroid lipofuscinoses
Adv. Genet.
2001
, vol. 
45
 (pg. 
35
-
68
)
Wheeler
 
R. B.
Sharp
 
J. D.
Schultz
 
R. A.
Joslin
 
J. M.
Williams
 
R. E.
Mole
 
S. E.
 
The gene mutated in variant late-infantile neuronal ceroid lipofuscinosis (CLN6) and in nclf mutant mice encodes a novel predicted transmembrane protein
Am. J. Hum. Genet.
2002
, vol. 
70
 (pg. 
537
-
542
)
Gao
 
H.
Boustany
 
R. M.
Espinola
 
J. A.
Cotman
 
S. L.
Srinidhi
 
L.
Antonellis
 
K. A.
Gillis
 
T.
Qin
 
X.
Liu
 
S.
Donahue
 
L. R.
, et al 
Mutations in a novel CLN6-encoded transmembrane protein cause variant neuronal ceroid lipofuscinosis in man and mouse
Am. J. Hum. Genet.
2002
, vol. 
70
 (pg. 
324
-
335
)
Heine
 
C.
Quitsch
 
A.
Storch
 
S.
Martin
 
Y.
Lonka
 
L.
Lehesjoki
 
A. E.
Mole
 
S. E.
Braulke
 
T.
 
Topology and endoplasmic reticulum retention signals of the lysosomal storage disease-related membrane protein CLN6
Mol. Membr. Biol.
2007
, vol. 
24
 (pg. 
74
-
87
)
Oresic
 
K.
Noriega
 
V.
Andrews
 
L.
Tortorella
 
D.
 
A structural determinant of human cytomegalovirus US2 dictates the down-regulation of class I major histocompatibility molecules
J. Biol. Chem.
2006
, vol. 
281
 (pg. 
19395
-
19406
)
Tortorella
 
D.
Story
 
C. M.
Huppa
 
J. B.
Wiertz
 
E. J.
Jones
 
T. R.
Bacik
 
I.
Bennink
 
J. R.
Yewdell
 
J. W.
Ploegh
 
H. L.
 
Dislocation of type I membrane proteins from the ER to the cytosol is sensitive to changes in redox potential
J. Cell Biol.
1998
, vol. 
142
 (pg. 
365
-
376
)
Mueller
 
B.
Lilley
 
B. N.
Ploegh
 
H. L.
 
SEL1L, the homologue of yeast Hrd3p, is involved in protein dislocation from the mammalian ER
J. Cell Biol.
2006
, vol. 
175
 (pg. 
261
-
270
)
Hochstenbach
 
F.
David
 
V.
Watkins
 
S.
Brenner
 
M. B.
 
Endoplasmic reticulum resident protein of 90 kilodaltons associates with the T- and B-cell antigen receptors and major histocompatibility complex antigens during their assembly
Proc. Natl. Acad. Sci. U.S.A.
1992
, vol. 
89
 (pg. 
4734
-
4738
)
Teixeira
 
C. A.
Espinola
 
J.
Huo
 
L.
Kohlschutter
 
J.
Persaud Sawin
 
D. A.
Minassian
 
B.
Bessa
 
C. J.
Guimaraes
 
A.
Stephan
 
D. A.
Sa Miranda
 
M. C.
, et al 
Novel mutations in the CLN6 gene causing a variant late infantile neuronal ceroid lipofuscinosis
Hum. Mutat.
2003
, vol. 
21
 (pg. 
502
-
508
)
Scott
 
D. C.
Schekman
 
R.
 
Role of Sec61p in the ER-associated degradation of short-lived transmembrane proteins
J. Cell Biol.
2008
, vol. 
181
 (pg. 
1095
-
1105
)
Wang
 
Q.
Li
 
L.
Ye
 
Y.
 
Regulation of retrotranslocation by p97-associated deubiquitinating enzyme ataxin-3
J. Cell Biol.
2006
, vol. 
174
 (pg. 
963
-
971
)
Wiertz
 
E. J.
Tortorella
 
D.
Bogyo
 
M.
Yu
 
J.
Mothes
 
W.
Jones
 
T. R.
Rapoport
 
T. A.
Ploegh
 
H. L.
 
Sec61-mediated transfer of a membrane protein from the endoplasmic reticulum to the proteasome for destruction
Nature
1996
, vol. 
384
 (pg. 
432
-
438
)
Fiebiger
 
E.
Hirsch
 
C.
Vyas
 
J. M.
Gordon
 
E.
Ploegh
 
H. L.
Tortorella
 
D.
 
Dissection of the dislocation pathway for type I membrane proteins with a new small molecule inhibitor, eeyarestatin
Mol. Biol. Cell
2004
, vol. 
15
 (pg. 
1635
-
1646
)
Omura
 
S.
Fujimoto
 
T.
Otoguro
 
K.
Matsuzaki
 
K.
Moriguchi
 
R.
Tanaka
 
H.
Sasaki
 
Y.
 
Lactacystin, a novel microbial metabolite, induces neuritogenesis of neuroblastoma cells
J. Antibiot. (Tokyo)
1991
, vol. 
44
 (pg. 
113
-
116
)
Baker
 
B. M.
Tortorella
 
D.
 
Dislocation of an endoplasmic reticulum membrane glycoprotein involves the formation of partially dislocated ubiquitinated polypeptides
J. Biol. Chem.
2007
, vol. 
282
 (pg. 
26845
-
26856
)
Lilley
 
B. N.
Ploegh
 
H. L.
 
A membrane protein required for dislocation of misfolded proteins from the ER
Nature
2004
, vol. 
429
 (pg. 
834
-
840
)
Kostova
 
Z.
Tsai
 
Y. C.
Weissman
 
A. M.
 
Ubiquitin ligases, critical mediators of endoplasmic reticulum-associated degradation
Semin. Cell Dev. Biol.
2007
, vol. 
18
 (pg. 
770
-
779
)
Lilley
 
B. N.
Ploegh
 
H. L.
 
Multiprotein complexes that link dislocation, ubiquitination, and extraction of misfolded proteins from the endoplasmic reticulum membrane
Proc. Natl. Acad. Sci. U.S.A.
2005
, vol. 
102
 (pg. 
14296
-
14301
)
Ye
 
Y.
Shibata
 
Y.
Kikkert
 
M.
van Voorden
 
S.
Wiertz
 
E.
Rapoport
 
T. A.
 
Inaugural article: recruitment of the p97 ATPase and ubiquitin ligases to the site of retrotranslocation at the endoplasmic reticulum membrane
Proc. Natl. Acad. Sci. U.S.A.
2005
, vol. 
102
 (pg. 
14132
-
14138
)
Kopito
 
R. R.
 
Aggresomes, inclusion bodies and protein aggregation
Trends Cell Biol.
2000
, vol. 
10
 (pg. 
524
-
530
)
Haltia
 
M.
 
The neuronal ceroid-lipofuscinoses: from past to present
Biochim. Biophys. Acta
2006
, vol. 
1762
 (pg. 
850
-
856
)
Hebert
 
D. N.
Molinari
 
M.
 
In and out of the ER: protein folding, quality control, degradation, and related human diseases
Physiol. Rev.
2007
, vol. 
87
 (pg. 
1377
-
1408
)
Aridor
 
M.
 
Visiting the ER: the endoplasmic reticulum as a target for therapeutics in traffic related diseases
Adv. Drug Deliv. Rev.
2007
, vol. 
59
 (pg. 
759
-
781
)
Brodsky
 
J. L.
 
The protective and destructive roles played by molecular chaperones during ERAD (endoplasmic-reticulum-associated degradation)
Biochem. J.
2007
, vol. 
404
 (pg. 
353
-
363
)
Nakatsukasa
 
K.
Brodsky
 
J. L.
 
The recognition and retrotranslocation of misfolded proteins from the endoplasmic reticulum
Traffic
2008
, vol. 
9
 (pg. 
861
-
870
)
Ye
 
Y.
Meyer
 
H. H.
Rapoport
 
T. A.
 
The AAA ATPase Cdc48/p97 and its partners transport proteins from the ER into the cytosol
Nature
2001
, vol. 
414
 (pg. 
652
-
656
)
Cheng
 
S. H.
Gregory
 
R. J.
Marshall
 
J.
Paul
 
S.
Souza
 
D. W.
White
 
G. A.
O'Riordan
 
C. R.
Smith
 
A. E.
 
Defective intracellular transport and processing of CFTR is the molecular basis of most cystic fibrosis
Cell
1990
, vol. 
63
 (pg. 
827
-
834
)
Yam
 
G. H.
Zuber
 
C.
Roth
 
J.
 
A synthetic chaperone corrects the trafficking defect and disease phenotype in a protein misfolding disorder
FASEB J.
2005
, vol. 
19
 (pg. 
12
-
18
)
Beck
 
M.
 
New therapeutic options for lysosomal storage disorders: enzyme replacement, small molecules and gene therapy
Hum. Genet.
2007
, vol. 
121
 (pg. 
1
-
22
)
Brady
 
R. O.
 
Emerging strategies for the treatment of hereditary metabolic storage disorders
Rejuvenation Res.
2006
, vol. 
9
 (pg. 
237
-
244
)
Sakuraba
 
H.
Sawada
 
M.
Matsuzawa
 
F.
Aikawa
 
S.
Chiba
 
Y.
Jigami
 
Y.
Itoh
 
K.
 
Molecular pathologies of and enzyme replacement therapies for lysosomal diseases
CNS Neurol. Disord. Drug Targets
2006
, vol. 
5
 (pg. 
401
-
413
)
Ahmad-Annuar
 
A.
Tabrizi
 
S. J.
Fisher
 
E. M.
 
Mouse models as a tool for understanding neurodegenerative diseases
Curr. Opin. Neurol.
2003
, vol. 
16
 (pg. 
451
-
458
)
Cooper
 
J. D.
Russell
 
C.
Mitchison
 
H. M.
 
Progress towards understanding disease mechanisms in small vertebrate models of neuronal ceroid lipofuscinosis
Biochim. Biophys. Acta
2006
, vol. 
1762
 (pg. 
873
-
889
)
Hafezparast
 
M.
Ahmad-Annuar
 
A.
Wood
 
N. W.
Tabrizi
 
S. J.
Fisher
 
E. M.
 
Mouse models for neurological disease
Lancet Neurol.
2002
, vol. 
1
 (pg. 
215
-
224
)
Gupta
 
P.
Soyombo
 
A. A.
Atashband
 
A.
Wisniewski
 
K. E.
Shelton
 
J. M.
Richardson
 
J. A.
Hammer
 
R. E.
Hofmann
 
S. L.
 
Disruption of PPT1 or PPT2 causes neuronal ceroid lipofuscinosis in knockout mice
Proc. Natl. Acad. Sci. U.S.A.
2001
, vol. 
98
 (pg. 
13566
-
13571
)
Jalanko
 
A.
Vesa
 
J.
Manninen
 
T.
von Schantz
 
C.
Minye
 
H.
Fabritius
 
A. L.
Salonen
 
T.
Rapola
 
J.
Gentile
 
M.
Kopra
 
O.
Peltonen
 
L.
 
Mice with Ppt1Deltaex4 mutation replicate the INCL phenotype and show an inflammation-associated loss of interneurons
Neurobiol. Dis.
2005
, vol. 
18
 (pg. 
226
-
241
)
Sleat
 
D. E.
Wiseman
 
J. A.
El-Banna
 
M.
Kim
 
K. H.
Mao
 
Q.
Price
 
S.
Macauley
 
S. L.
Sidman
 
R. L.
Shen
 
M. M.
Zhao
 
Q.
, et al 
A mouse model of classical late-infantile neuronal ceroid lipofuscinosis based on targeted disruption of the CLN2 gene results in a loss of tripeptidyl-peptidase I activity and progressive neurodegeneration
J. Neurosci.
2004
, vol. 
24
 (pg. 
9117
-
9126
)
Kopra
 
O.
Vesa
 
J.
von Schantz
 
C.
Manninen
 
T.
Minye
 
H.
Fabritius
 
A. L.
Rapola
 
J.
van Diggelen
 
O. P.
Saarela
 
J.
Jalanko
 
A.
Peltonen
 
L.
 
A mouse model for Finnish variant late infantile neuronal ceroid lipofuscinosis, CLN5, reveals neuropathology associated with early aging
Hum. Mol. Genet.
2004
, vol. 
13
 (pg. 
2893
-
2906
)
Bronson
 
R. T.
Donahue
 
L. R.
Johnson
 
K. R.
Tanner
 
A.
Lane
 
P. W.
Faust
 
J. R.
 
Neuronal ceroid lipofuscinosis (nclf), a new disorder of the mouse linked to chromosome 9
Am. J. Med. Genet.
1998
, vol. 
77
 (pg. 
289
-
297
)
Cooper
 
J. D.
Messer
 
A.
Feng
 
A. K.
Chua-Couzens
 
J.
Mobley
 
W. C.
 
Apparent loss and hypertrophy of interneurons in a mouse model of neuronal ceroid lipofuscinosis: evidence for partial response to insulin-like growth factor-1 treatment
J. Neurosci.
1999
, vol. 
19
 (pg. 
2556
-
2567
)
Mitchison
 
H. M.
Bernard
 
D. J.
Greene
 
N. D.
Cooper
 
J. D.
Junaid
 
M. A.
Pullarkat
 
R. K.
de Vos
 
N.
Breuning
 
M. H.
Owens
 
J. W.
Mobley
 
W. C.
, et al 
Targeted disruption of the Cln3 gene provides a mouse model for Batten disease. The Batten Mouse Model Consortium
Neurobiol Dis.
1999
, vol. 
6
 (pg. 
321
-
334
)
Cotman
 
S. L.
Vrbanac
 
V.
Lebel
 
L. A.
Lee
 
R. L.
Johnson
 
K. A.
Donahue
 
L. R.
Teed
 
A. M.
Antonellis
 
K.
Bronson
 
R. T.
Lerner
 
T. J.
MacDonald
 
M. E.
 
Cln3(Deltaex7/8) knock-in mice with the common JNCL mutation exhibit progressive neurologic disease that begins before birth
Hum. Mol. Genet.
2002
, vol. 
11
 (pg. 
2709
-
2721
)
Katz
 
M. L.
Shibuya
 
H.
Liu
 
P. C.
Kaur
 
S.
Gao
 
C. L.
Johnson
 
G. S.
 
A mouse gene knockout model for juvenile ceroid-lipofuscinosis (Batten disease)
J. Neurosci. Res.
1999
, vol. 
57
 (pg. 
551
-
556
)
Wendt
 
K. D.
Lei
 
B.
Schachtman
 
T. R.
Tullis
 
G. E.
Ibe
 
M. E.
Katz
 
M. L.
 
Behavioral assessment in mouse models of neuronal ceroid lipofuscinosis using a light-cued T-maze
Behav. Brain Res.
2005
, vol. 
161
 (pg. 
175
-
182
)

Supplementary data