Copper depletion down-regulates expression of Alzheimer’s disease Amyloid- β precursor protein gene

(APP). in one Key targets for are that affect the regulation of the APP gene. Recent in vivo and in vitro studies have illustrated the importance of copper in Alzheimer’s disease neuropathogenesis and suggested a role for APP and amyloid- β in copper homeostasis. We hypothesised that metals and in particular copper might alter APP gene expression. To test the hypothesis, we utilised human fibroblasts over-expressing the Menkes protein (MNK), a major mammalian copper efflux protein. MNK deletion fibroblasts have high intracellular copper, while MNK over-expressing fibroblasts have severely depleted intracellular copper. We demonstrate that copper depletion significantly reduced APP protein levels and down-regulated APP gene expression. Furthermore, APP promoter deletion constructs identified the copper-regulatory region between –490 to +104 of the APP gene promoter in both basal MNK over-expressing cells and in copper-chelated MNK deletion cells. Overall these data support the hypothesis that copper can regulate APP expression and further support a role for APP to function in copper homeostasis. Copper-regulated APP expression may also provide a potential therapeutic target in Alzheimer’s disease.


INTRODUCTION
A number of neurodegenerative disorders including Alzheimer's disease (AD 1 ), Parkinson's disease, Amyotrophic lateral sclerosis and Prion disease have been closely linked to disturbances in copper homeostasis in the central nervous system and the brain (1,2). AD is the most common progressive neurodegenerative disorder in elderly people and is characterised by neuronal loss with the accumulation of senile plaques and neurofibrillary tangles. The major proteinaceous component of senile plaques is a 39-42 amino acid peptide, termed Amyloid-β peptide (Aβ) (3), that is proteolytically cleaved from the larger amyloid-β precursor protein (APP) (4).
An apparent relationship between copper homeostasis, APP and AD has been suggested (1). APP has a specific type II copper-binding site that binds copper with a Kd of 10 µM (5) and can modulate copper-induced toxicity and oxidative stress in primary mouse neuronal cultures (6). APP knockout mice have increased copper levels in the brain (7) while conversely, transgenic over-expression of APP and Aβ in mice correlates with reduced brain copper levels (8). In a cellular system overexpressing APP, increasing copper concentration can modulate APP processing, stimulating levels of cell-bound and secreted forms of APP with reduced production of Aβ (9). The importance of copper in AD pathology has also been demonstrated by the ability of Aβ to bind copper with a high affinity (10) promoting amyloid plaque aggregation and neurotoxicity (11). Conversely, treatment with a copper-zinc chelator can disaggregate Aβ both in vitro and in transgenic mouse models in vivo (12).
The APP gene is expressed in all major tissues but predominantly in the brain, where expression is primarily in neurons (13). Although there is both developmental and cell-type specific regulation of APP, expression in the adult is mostly ubiquitous (14). The regulation of the APP gene as a pathogenic factor for AD has received considerable attention. Down syndrome patients, who have an extra copy of the APP gene, invariably exhibit early onset AD-like pathology (15). Over-expression of APP in certain areas of the brain in AD patients also suggest that the regulation of APP might be an important factor in the neuropathology of AD (16). These observations illustrate the importance of elucidating mechanisms of APP gene regulation in the development of AD.
The human APP promoter closely resembles that of a typical housekeeping gene and contains the consensus sequences for the binding of several transcription factors (17,18). Nerve growth factor, interleukin-1, retinoic acid and various transcription factors are among several cellular mediators that cause an increase in APP mRNA levels in neuronal and non-neuronal cells (19). In contrast, thyroid hormone and interferon-gamma have been reported to down-regulate APP expression (20,21).
Recently, an iron responsive element located in the 5'-untranslated region has been implicated in the regulation of APP expression (22). We hypothesised that metals and in particular copper might alter APP gene expression.
To investigate the role of copper in APP gene regulation, we utilised a novel approach involving cultured human fibroblasts over-expressing the Menkes protein (MNK; encoded by ATP7A), a major mammalian copper translocating P-type ATPase involved in copper efflux (23)(24)(25). Cells lacking the MNK protein show high by guest on March 24, 2020 http://www.jbc.org/ Downloaded from intracellular copper levels due to the lack of active copper efflux, while cells transfected and over-expressing MNK have markedly reduced copper levels (25).
We report that depletion of intracellular copper results in significant reduction of APP gene expression. In addition, APP promoter analysis suggests putative metal regulatory elements may be involved in mediating the response to copper depletion to regulate APP gene expression. Overall, these data suggest that copper is a required co-factor for the basal regulation of the APP gene and supports the increasing evidence that APP is involved in copper homeostasis as a copper detoxification/efflux protein.

EXPERIMENTAL PROCEDURES
Cell lines -Human skin fibroblast cells, Me32a, were isolated from a classical Menkes disease patient (26). The patient carried a 4 bp deletion that resulted in a frame-shift mutation and premature stop codon of the MNK gene, whose expression was undetectable by western and northern analysis (25)(26)(27) (31). In addition, promoters were searched for ACE1/AMT1-like sequences (ALS), with no more than one base mismatch to the first two ACE1/AMT1 consensus residues (5' THNNGCTG 3'); and MRE-like sequences (MLS), with no more than one base mismatch from the last three MRE consensus residues (5' TGCRCNC 3').
Antibodies -Anti-MNK polyclonal antibodies raised to the MNK N-terminal region (24) were diluted 1:2500 for use in western blot analysis. Anti-APP (WO2) monoclonal antibodies raised to the Amyloid-β region (32) were diluted 1:1000 for use in immunofluorescence studies and 1:10000 for use in western blot analysis. Total cellular measurement of copper, zinc and iron -Cells were grown for 4 days in basal growth medium as described above. Cells were harvested as previously described (34) and copper, zinc and iron were analysed using a Perkin Elmer 5000 Atomic Absorption Spectrophotometer.
Indirect immunofluorescence -Cells were seeded in basal medium onto 13mm round cover-slips 48 hours prior to immunofluorescence. Cover-slips were washed in ice-cold phosphate-buffered saline (PBS) and fixed with 4% paraformaldehyde in PBS for 15 min, washed several times with PBS and permeabilised with 0.1% Triton X-100 in PBS for 5 min, washed several times in PBS then blocked overnight in PBS containing 1% bovine serum albumin (BSA). All antibody incubations were for 1 hour in 1% BSA/PBS. Monoclonal APP antiserum was visualised using Alexa-488-conjugated to goat anti-mouse IgG antibodies (Molecular Probes). Cover-slips were washed overnight in PBS then mounted using 100 mg/ml DABCO (Sigma) in 90% glycerol. Confocal microscopy was carried out using an Olympus BX60 microscope with a 60x PlanApo lens with a 1.40 N/A. This microscope was Total RNA extraction and Northern blot analysis -Total RNA was extracted from cell lines using the Total RNAeasy kit (QIAGEN). 5µg total RNA samples were denatured in formamide, formaldehyde, MOPS and ethidium bromide at 65°C for 15 min, cooled on ice, and electrophoresed on a 1.0% agarose-formaldehyde gel.
The gel was blotted on Hybond-N+ nitrocellulose filter and immobilised by 50 mM sodium hydroxide. Each filter was pre-hybridised in hybridisation buffer containing 50 X Denhardt's reagent, sheared Salmon sperm DNA, 0.5 M EDTA, 1 M Tris, 20% SDS and 1 M NaPO 4 for at least 2 hours. The filter was hybridised with a α-32 P-dATP labelled cDNA probe using the Prime-a-Gene labelling System (Promega).
After hybridisation, the filters were washed to a stringency of 0.01% SDS, 0.1 X SSC at 60°C. APP cDNA probe corresponded to 3kb fragment generated from human APP cDNA. Equal loading of samples was verified by re-hybridising the filter with the "housekeeping gene cDNA probe" rat glyceraldehyde phosphate dehydrogenase by guest on March 24, 2020 http://www.jbc.org/ Downloaded from (GAPDH). The filter were exposed for 2 hours to a PhosphorScreen (Molecular Dynamics) and scanned with a Typhoon 8600 PhosphorImager (Molecular Dynamics). Densitometer quantification was done with ImageQuant (Molecular Dynamics) software for both APP and GAPDH northern analysis.
Transient transfection and lysate preparation -Cells were seeded in 6-well plates 24 hours prior to transfection in basal medium. All cells were transfected with carbonate. The absorbance was measured on a Bio-Rad 2550 EIA plate reader at 420 nm.
Chloramphenicol acetyl transferase (CAT) promoter assay -The cell extracts from transfected cells were analysed for their CAT activity using a colorimetric ELISA assay according to manufacturer's instructions (Roche Molecular Biochemicals). Briefly, 50µg of protein (an amount within the linear range of the assay) was placed on anti-CAT coated microtiter plate modules and allowed to bind for 1 hour at 37°C. The plates were washed thoroughly after each step. Next, a digoxigenin-labelled anti-CAT antibody was added to the samples and incubated for 1 hour at 37°C. A subsequent antibody, anti-digoxigenin conjugated to peroxidase, was placed in the wells for 1 hour at 37°C. Finally, peroxidase substrate, ABTS (Roche Molecular Biochemicals), was added and the absorbance measured at 405 nM using an ELISA plate reader. CAT activity was determined from the ratio of pg CAT/milli unit β-Gal per µg protein.
Statistical analysis -One-way ANOVA of more than two means followed by Bonferroni's multiple comparison of mean's post test was performed for northern analysis, 64 Cu accumulation, cellular metal determination, and promoter assays using Graphpad Prism3 for Macintosh (GraphPad Software Inc.). Statistically significant was defined as P < 0.05.

RESULTS
(  (Fig. 1A) have dramatically altered copper levels, we investigated cells for intracellular levels of copper as well as zinc and iron.
Intracellular copper levels were measured using two independent approaches.
The first was to analyse cells for copper accumulation using the Cu radioisotope, 64 Cu, under steady state conditions. 64 Cu analysis showed MNK deletion and vector only control cells maintained copper levels significantly higher than normal human fibroblast cells with an average increase of 125% (Fig. 1B). MNK transfected cells maintained copper levels significantly lower than normal human fibroblast cells with an average decrease of 82% (Fig. 1B). MNK transfected cells also maintained copper levels significantly lower than MNK deletion cells with an average decrease of 92% (Fig. 1B).
The second method of analysis examined the total amount of copper present in cell pellets utilising atomic absorption spectroscopy. This analysis showed MNK deletion and vector only control cells contained significantly elevated copper levels compared to normal human fibroblast cells with an average increase of 65% (Fig.   1C). MNK transfected cells contained significantly reduced copper levels, with an average decrease of 79%, compared to normal human fibroblast cells (Fig. 1C). MNK transfected cells also contained significantly reduced copper levels compared to MNK deletion cells with an average decrease of 87% (Fig. 1C). There was no significant difference in the zinc and iron content of these fibroblast cell lines (Fig. 1D-E).
Overall, the cellular metal analysis demonstrated that by over-expression of MNK in a MNK deletion background, we could dramatically alter intracellular copper Immunofluorescence analysis showed MNK deletion, vector only control and normal human fibroblasts all had APP and/or Aβ levels similar to that observed in control human neurons ( Fig. 2A). Interestingly, APP and/or Aβ protein was not detectable via immunofluorescence in MNK transfected cells ( Fig. 2A).
APP protein expression in fibroblast lines was further examined using western blot analysis. APP protein isoforms were detected as bands between 90 and 110 kDa (Fig. 2B). MNK deletion, vector only control and normal human fibroblasts produced protein bands that were comparable to control human neurons (Fig. 2B). However, in MNK transfected cells, the protein bands detected were severely decreased in their intensity when compared to vector only control, normal human fibroblasts and control human neurons (Fig. 2B). Possible mechanisms for the reduced levels of APP protein in MNK transfected cell lines could involve effects at the transcriptional, translational or post-translational levels.
We performed northern analysis of APP mRNA levels in these fibroblast lines.
A single APP mRNA transcript band was observed in MNK deletion, vector only control and normal human fibroblast cell lines that corresponded to the transcript detected in control human neurons (Fig. 2C, upper panel). MNK transfected cells also had a single APP mRNA transcript. However, the observed band intensity was severely reduced compared to control lines and to the control housekeeping transcript by guest on March 24, 2020 http://www.jbc.org/ Downloaded from GAPDH (Fig. 2C, lower panel). Quantification of APP mRNA levels by densitometry showed no significant difference in APP expression between MNK deletion, vector only control, normal human fibroblast and human neuron cell lines (Fig. 2D). Notably, MNK transfected cells showed a significant decrease in APP expression with a down-regulation of approximately 83% compared to MNK deletion cells (Fig. 2D). These data provide strong support for the hypothesis that reduced copper levels result in down-regulation of APP gene expression in MNK transfected cultured human fibroblasts. (Table II suggested location).

The human APP and rhesus monkey APP gene promoters contain a number of
putative metal regulatory sequences -The human APP gene promoter (GenBank D87875) and rhesus monkey APP gene promoter sequences (GenBank AF067971) were searched for copper response elements (CuRE) found in yeast copper uptake genes FRE1, CTR1 and CTR3 promoters (28); ACE1/AMT1 metal response elements found in yeast copper detoxification genes CUP1, CRS5 and SOD1 promoters (29); and metal response elements (MRE) found in the mammalian copper detoxification gene metallothionein (MT ) promoter (30). In addition promoters were searched for ACE1/AMT1-like sequences and MRE-like sequences. Utilising this search criteria, a number of putative copper and metal response elements were identified in the human and rhesus monkey APP promoters ( Table 2). The contribution of these elements to a possible copper-responsive regulation of the APP gene has not yet been determined.
( Figure 3 suggested location). To confirm that the differential promoter response seen was due to low intracellular copper levels, we depleted copper levels in MNK deletion cells using the copper chelator, diamsar (38). MNK deletion cells were transiently transfected with promoter constructs and grown in basal media and media containing the copper chelator, diamsar for 48 hours. Analysis showed a significant difference between non-chelated and chelated cells with the -3416 to +104, -1131 to +104 and -490 to +104 APP promoter fusion constructs (Fig. 3D). Chelation with 5, 10 and 50µM diamsar was sufficient to down-regulate APP promoter activity by 53, 73 and 80% respectively (Fig. 3D).

Copper depletion decreases APP gene expression through the 5'-APP promoter
Overall analysis of APP promoter activity provided strong support for the hypothesis that reduced copper levels result in down-regulation of APP gene expression in cultured human fibroblasts. Furthermore, investigations utilising APP promoter deletion constructs narrowed the copper-responsive regulatory region to between -490 to +104 of the APP gene promoter in both basal MNK transfected cells and in copper-chelated MNK deletion cells (Fig. 3).
Since it has been demonstrated in various systems that genes involved in copper homeostasis can be regulated by cellular copper levels (28)(29)(30), as evidenced by the yeast copper uptake genes FRE1, CTR1 and CTR3 (28); yeast copper detoxification genes CUP1, CRS5 and SOD1 (29) and the mammalian copper detoxification gene MT (30), we hypothesised that copper may also play a role in the regulation of the APP gene. To investigate the role of copper in APP gene regulation, we utilised a system where intracellular copper levels could be genetically manipulated through altered expression of the MNK copper efflux protein (Table 1). Cells lacking the MNK protein show high intracellular copper levels due to reduced copper efflux (25). Here we report for the first time, evidence that decreased copper can downregulate expression of the human APP gene. This demonstrates a previously uncharacterised aspect of regulation of the human APP gene and further supports a role for the APP protein in copper homeostasis. This study also suggests that copperfactor(s) in the region -490 to +104 of the APP promoter are required for regulation of the APP gene.
The effect of decreased copper on APP gene regulation was observed in two independently derived MNK transfected fibroblast cell lines that constitutively express high levels of MNK. The use of two independent methods for determining the steady-state levels of copper in these fibroblast cell lines confirms that in MNK transfected cell lines the cellular copper levels are reduced by about 90% compared to the parental MNK deletion cell line (Fig. 1). These results are consistent with previous published copper data (25). Importantly, the levels of zinc, a known transcriptional co-factor, and iron, recently reported to be involved in regulating the 5' UTR of the APP gene (22), were unaffected (Fig. 1). The reduction of copper levels in MNK transfected cells correlates with a decrease in APP protein levels and on average an 83% down-regulation of APP gene expression detected in MNK transfected fibroblasts (Fig. 2). Elevated copper levels observed in MNK deletion and vector only control cells relative to normal human fibroblasts (Fig. 1), were not associated with altered APP expression (Fig. 2). These results suggest that copper-related down-regulation of APP gene expression is associated with the very low copper levels attained in MNK transfected fibroblast cell lines.
To evaluate the role of the APP gene promoter in the copper related downregulation of the APP gene we utilised APP promoter deletion constructs fused to a CAT reporter system. Analysis was performed under basal medium conditions, where it was predicted that MNK transfected cell lines would have decreased promoter activity compared to MNK deletion cells due to reduced copper levels (Fig.   1). MNK transfected cell lines demonstrated a significant decrease in basal APP promoter activity when compared to MNK deletion cells for all promoter deletion constructs analysed (Fig. 3B). MNK transfected cell lines are down-regulated by 75% on average when compared to promoter activity observed for MNK deletion and vector only control cell lines (Fig. 3B). This result is consistent with the approximate 83% down-regulation of APP mRNA expression observed from northern analysis.
Whilst copper supplementation of MNK transfected cells was shown to increase cellular copper levels (Table 3), APP promoter activity was not restored (Fig. 3C).
This may result from a lack of available copper in the pool responsible for activating APP promoter expression, and could be due to sequestration of copper via MNK in intracellular compartments or vesicles in cells over-expressing MNK (40,41).
Under copper-depleted conditions in MNK deletion cells, the APP promoter activity was significantly reduced by up to 80% using the Cu 2+ "cage" chelator diamsar (Fig. 3D). MNK deletion cells in the presence of 10µM and 50µM diamsar copper chelator showed a similar down-regulation of the APP promoter to that observed under basal medium conditions for MNK transfected cells (Compare Fig. 3B to Fig. 3D). This suggests that the putative regulatory element(s), responsible for the down-regulation of the APP gene observed in MNK transfected fibroblasts, are located within the region -490 to +104 of the APP promoter. It has recently been suggested that both APP and Aβ can function in copper efflux/detoxification (8,39). This was based on: i) the structural homology of the APP copper binding-domain to known copper chaperone and copper efflux protein copperbinding domains (39); ii) the extracellular localisation of the copper binding domain (4); iii) the findings that APP knockout mice show increases in brain copper levels (7), whilst transgenic overexpression of APP and Aβ results in reduced brain copper levels (8); iv) the observation that elevated copper levels results in a decrease in the amyloidogenic pathway and stimulation of the non-amyloidogenic pathway of APP cleavage, thus releasing the secreted APP ectodomain with concomitant efflux of copper (9).
The regulation of APP gene expression by copper described in the current studies also strongly supports a role for APP in copper efflux/detoxification. In conditions is similar to known transcriptional regulation mechanisms for copper detoxification/efflux genes in yeast and humans (29,30). Given the nonamyloidogenic processing of APP and in vivo reduction of Aβ under elevated copper conditions (9,42,43), it is possible that down-regulation of the APP gene under low copper conditions acts as a preventative measure to decrease the production of Aβ and effectively guard against amyloidogenic processing of APP. In summary, utilising a novel cellular copper efflux system, our results strongly suggest that at least in human fibroblasts copper is a co-factor in basal APP gene regulation. The data also further supports the role of APP to function in copper efflux/detoxification. The elucidation of the copper-regulation mechanisms of APP in human fibroblasts may provide new targets in developing therapeutic strategies in the treatment of AD that are designed to reduce the expression of the APP gene and the ensuing production of Aβ. These findings also suggest that current phase II clinical trials of copper-zinc chelators in the treatment of AD (12,44,45)