Transcriptional Regulation of the Beta-Synuclein 5′-Promoter Metal Response Element by Metal Transcription Factor-1

The progression of many human neurodegenerative disorders is associated with an accumulation of alpha-synuclein. Alpha-synuclein belongs to the homologous synuclein family, which includes beta-synuclein. It has been proposed that beta-synuclein may be a natural regulator of alpha-synuclein. Therefore controlling beta-synuclein expression may control the accumulation of alpha-synuclein and ultimately prevent disease progression. The regulation of synucleins is poorly understood. We investigated the transcriptional regulation of beta-synuclein, with the aim of identifying molecules that differentially control beta-synuclein expression levels. To investigate transcriptional regulation of beta-synuclein, we used reporter gene assays and bioinformatics. We identified a region −1.1/−0.6 kb upstream of the beta-synuclein translational start site to be a key regulatory region of beta-synuclein 5′-promoter activity in human dopaminergic cells (SH-SY5Y). Within this key promoter region we identified a metal response element pertaining to a putative Metal Transcription Factor-1 (MTF-1) binding site. We demonstrated that MTF-1 binds to this 5′-promoter region using EMSA analysis. Moreover, we showed that MTF-1 differentially regulates beta-synuclein promoter binding site, as well as beta-synuclein mRNA and protein expression. This effect of MTF-1 on expression was found to be specific to beta-synuclein when compared to alpha-synuclein. Understanding the regulation of synucleins and how they interact may point to molecular targets that could be manipulated for therapeutic benefit. In this study we showed that MTF-1 differentially controls the expression of beta-synuclein when compared to its homolog alpha-synuclein. This could potentially provide a novel targets or pathways for therapeutic intervention and/or treatment of synucleinopathies.


Introduction
Neurodegenerative disorders constitute a disease category that the World Health Organization calculates will become the world's second leading cause of death by the year 2040, overtaking cancer.The progression of many human neurodegenerative disorders is associated with an accumulation of alpha-synuclein (a-Syn), including Alzheimer's disease (AD) and Parkinson's diseases (PD), dementia associated with Lewy body disease (DLB), diffuse Lewy body disease (DLBD) and multiple system atrophy (MSA).This group of diseases are now known as synucleinopathies and are characterised by the presence of Lewy bodies (LB), the intracytoplasmic inclusions/aggregates found in dopaminergic neurons containing alpha-synuclein (a-Syn) [1].Abnormal protein-protein interactions may allow the precipitation of a-Syn, which facilitates the formation of these extracellular and intracellular aggregates.The formation of these deposits can be induced by a number of substances, including metal ions [2].Indeed, high levels of metals have been identified in the substantia nigra of PD brains [3].Aggregated a-Syn potentially inhibits proteasomal activity [4], which may be a cause of the observed reduction in proteasome activity in the substantia nigra of PD patients [5].Studies in families affected by dementia, PD and DLBD have genomic amplifications encompassing the a-Syn gene, leading to a proportional increase in a-Syn mRNA and protein in brain tissue [6][7][8].These cases support the hypothesis that increased expression of a-Syn causes disease.
a-Syn belongs to the homologous synuclein family, which includes beta-synuclein (b-Syn).Expression patterns and levels of a-Syn and b-Syn most closely overlap with highest levels throughout the brain [9].Although b-Syn is not detected in LB or form fibrils like a-Syn [10], b-Syn mutants have been identified in DLB patients [11] and b-syn protein is found to be abundant in neurofibrillary lesions of patients with AD [12].Moreover, in both the mouse brain and the human substantia nigra, a-Syn mRNA decreases and b-Syn mRNA increases with age [13].In contrast to control patients, there is a dramatic increase in a-Syn and decrease in b-Syn mRNA levels in the substantia nigra of PD, DLBD and a LB variant of AD patients [14].These changes were specific for the substantia nigra, the dopaminergic neuron-containing region of the brain most severely affected in synucleinopathies.This concentration reversal of synuclein transcript levels with disease suggests the balance of a-Syn and b-Syn expression may be important, which is supported by several studies [15,16].This suggests that b-Syn may be a natural negative regulator of a-Syn.
Transcriptional Regulation of the synucleins has been reported.Polymorphisms of the dinucleotide repeat complex NACP-Rep1 (10.7 kb upstream of the translational start site are associated with AD and PD [17,18].Different NACP-REP1 alleles have varying repressive effects on a-Syn promoter driven reporter activity in SH-SY5Y cells [19].Assessment of regions of the a-Syn promoter suggest the presence of activator sites between 21.5/21.9kb and repressor sites between 26.2/29.8kb upstream of the translational start site [19].Regulators of b-Syn expression have not been reported.Since b-Syn may be a natural negative regulator of a-Syn, identifying transcriptional regulators of b-Syn is essential.In particular, b-Syn is a good target for disease therapies as there are no apparent abnormalities when it is overexpressed in transgenic mice [16].Manipulation of transcriptional regulators of b-Syn expression would increase b-Syn levels, inhibit aggregation of a-Syn and potentially prevent disease progression.This study aimed to investigate the regulation of the b-Syn promoter.

PCR and Cloning
Several b-Syn promoter constructs of varying length and the open reading frames of a-syn, b-Syn, c-Syn and MTF-1 were cloned.Briefly, PCR products were generated using primers into which restriction endonuclease sites were engineered (Table 1).For the promoter constructs, PCRs were performed using a genomic clone that contained the beta-synuclein gene (ImaGenes, RZPDB737).For mammalian protein expression analysis, PCRs were performed using cDNA generated from SH-SY5Y cells, these methods have been previously described [20].
PCRs were carried out in a total volume of 50 ml, containing 200 mmol/L dNTPs, 0.4 mM of each primer, 1.5 mmol/L MgCl 2 , 1U of Phire Taq DNA Polymerase (Finnzymes, Labtech International LTD, UK), 1 ml reaction buffer, and 20 to 100 ng of DNA.Thermal cycling conditions were as follows: 2 min at 94uC (initial denaturation); 35 cycles of 30 s at 94uC, 30 s at 55-65uC, 40 s at 72uC; and a final extension of 2 min at 72uC.PCR products were cloned into either the luciferase promoter reporter vector, pGL3-basic (Promega Corporation, UK) or the mammalian expression vector, pcDNA3.1+(Invitrogen, UK).Constructs were confirmed by restriction endonuclease digestions and DNA sequencing and large scale DNA preparations were made with GeneJET TM Pasmid Miniprep Kit (Fermentas, UK).

Promoter Assays
Transfections into SH-SY5Y and U-87 MG cells were performed in 24-well plates seeded at 5610 4 cells/well 24 h prior to transfection.Transfections were performed using FuGENE6H transfection reagent (Roche, Mannheim, Germany) as per manufacturer's instructions.To control for variation in transfection efficiency among replicates, promoter constructs were co-transfected with the Renilla luciferase vector, pRL-TK, (Promega).At 48 h post-transfection, SH-SY5Y and U-87 MG cells were harvested and firefly and Renilla luciferase chemiluminescence were measured using the Dual-LuciferaseH Reporter Assay System (Promega) in a BMG FLUOstar Omega platereader (BMG Labtech GmbH, Offenburg, Germany).Luciferase activity was calculated as the ratio of firefly to Renilla luciferase activity.Each construct was tested in triplicate with at least three independent transfection experiments.
For experiments investigating copper, SH-SY5Y cells were plated 24 hr prior to transfection in the presence of serum-free media.To reduce copper levels we replaced fetal calf serum with B27 supplement (Invitrogen).The concentration of copper in defined medium is 0.1 mM [21].Cells were then supplemented with no copper, 10 mM or 50 mM at time of transfection.

Site-Directed Mutagenesis
Site-directed mutagenesis was used to introduce a point mutation at 2774 bp of the putative b-Syn promoter, specifically within a metal response element (MRE).PWO polymerase (Roche) was used with the following oligonucleotides in PCR reactions as per manufacturers instructions: Forward, 59-CC-GGGCACTGCTGCGTTCCTGCTCAG-39; Reverse, 59-GTC-CTGAGCAGGAACGCAGCAGTGC-39.MRE is shown in bold with mutant base underlined.Thermal cycling conditions were as a pEGFP-PrP (prion protein) was a gift from Dr Catia Sorgato and was subcloned into pcDNA3.1+using HindIII and EcoR1 restriction sites and is not listed.pcDNA3.1+a-Syn,pcDNA 3.1+NRF2, pcDNA3.1+MTF-1have been described previously [50,51]

Electrophoretic Mobility Shift Assays (EMSAs)
MTF-1 transcription factor binding to the MRE was determined by using EMSA.SH-SY5Y cells were stably transfected with empty expression vector (pcDNA3+) or pcDNA3.1+MTF-1.Cells were transfected as described above, except after 24 h transfection, cells containing vectors were selected for using media containing 400 mg/ml geneticin (Sigma).After several rounds of selection and maintenance, cells were washed in ice cold PBS and the protein was extracted using the NE-PER Nuclear and Cytoplasmic Extraction kit (Thermo Scientific, UK).Protein was quantified using the Bradford assay (Bio-Rad Laboratories Ltd, Hercules, CA).Synthetic biotin and non-biotin labelled probes (MWG, Germany) for the MRE (shown in bold) for normal MRE (Forward, 59-CCCGGGCACTGCTGCGCTCCTGCTCAGGACC-39; Reverse, GGTCCTGAGCAGGAGCGCAGCAGTGCCCGGG-39) and mutant MRE (Forward, 59-CCCGGGCACTGCTGCG-TTCCTGCTCAGGACC-39; Reverse, 59-GGTCCTGAGCAG-GAACGCAGCAGTGCCCGGG-39) were obtained for EMSA analysis.Binding reactions containing equal amounts of protein (15 mg) and probes were performed using the LightShiftH Chemiluminescent EMSA kit (Thermo Scientific, UK) as permanufacturer's instructions.Protein-DNA complexes were electrophoresed on 6% non-denaturing polyacrylamide gels, and transferred to Hybond-N+ (GE Healthcare, UK); signals were detected using LightShiftH Chemiluminescent EMSA kit (Thermo Scientific).All EMSA assays were performed for three sample replicates in each group.

Quantitative PCR
Cells stably transfected with pcDNA3.1+or pcDNA3.1+MTF-1were washed in ice cold PBS and RNA was obtained using TRI reagent (Sigma) as per manufacturer's instructions.Five micrograms of RNA was used from each sample for the synthesis of cDNA in a 20 ml volume with oligo(dT) 20 (Invitrogen) and SuperScript III reverse transcriptase (Invitrogen).The reactions were incubated at 50uC for 1 hr and then inactivated by heating at 70uC for 15 min.RNaseH (Invitrogen) was added, 1 ml to each preparation, and incubated for 20 min at 37uC.Six microlitres of diluted cDNA was used as template in quantitative PCR (QPCR) reactions in a 15 ml volume with 7.5 ml iTaq SYBR Green TM Supermix with ROX (Bio-Rad), 0.3 ml 100 nM of each primer (Table 2).Primers were designed, where possible, to span exons and to generate a fragment of between 80 and 130 bp.The QPCR assays were performed on a the Step One Plus TM Real-Time PCR System (Applied Biosystems, Foster City, CA) using the following PCR conditions: 2 min at 95uC followed by 40 cycles of 15 s denaturing at 95uC and 1 min annealing at 60uC and 15 s acquiring the fluorescence data (temperature dependent on the gene being assayed).A RT-negative control was used to assess whether any residual genomic DNA remained and a notemplate control was included to examine primer-dimer formation.QPCR experiments were excluded for analysis if the reaction efficiency was outside the range of 0.9 -1.1 or if the r 2 value of the linear regression was less than 0.95.Furthermore, the inter-experiment variability between replicates was assessed by measuring sample coefficient of variation and excluding samples if this was over 30%.
For the QPCR analysis, the relative quantification method was used.Our approach incorporated a standard curve to allow determination of the relative number of cDNA molecules of a given gene by using serial dilutions of an arbitrary cDNA sample.Expression of the gene of interest was normalised by the geometric mean of expression from three reference genes (GAPDH, ACTB and B2M).QPCR was performed with cDNA samples in triplicate from five independent replicates of cDNA from SH-SY5Y cells over-expressing various proteins.

Western Immunoblotting
Cells stably transfected with pcDNA3.1+or pcDNA3.1+MTF-1were washed in ice cold PBS and the protein was obtained as described above.Thirty micrograms of cytosolic protein was used for identifying a-Syn (Roche) and b-Syn (Millipore, UK) protein levels.All experiments were performed in triplicate.Protein band intensities were measured using Quantity One densitometry software (Bio-Rad).Further details of Western immunoblot analysis have been described previously [22].

Statistical Analysis
We evaluated statistical significance for the promoter assays, QPCR and Western immunoblot analysis with paired t-tests on GraphPad Prism software (GraphPad, San Diego, CA).

Basal promoter activity
The putative b-syn promoter region was assessed using a Dual-LuciferaseH reporter assay system.The genomic organisation of b-Syn has previously been described [23].A region spanning ,10.8 kb 5-prime of the translational start site, pertaining to nine promoter fragments, were analysed (Figure 1A).Parallel reporter gene assays of the nine constructs demonstrated a significant difference in basal transcription rate, in SH-SY5Y cells, between pGL3b 21.1/20.6 kb and the other b-syn promoter constructs, which had minimal or negligible activity above background (Figure 1B).Background is measured by empty pGL3basic vector.All the b-Syn promoter constructs had minimal or negligible activity in the U-87 MG astrocytoma cells (Figure 1C).

Transcription Factor regulation of the b-Syn promoter
To investigate the regulation of b-Syn 21.1/20.6kb, we performed a transcriptional binding site analysis using the Transcription Element Search System (TESS) [24] and identified a putative metal transcription factor-1 site pertaining to a metal response element (MRE-TGCGCTC).To explore this further we looked at the effect of cells over-expressing MTF-1 on the b-Syn 21.1/20.6 kb promoter fragment.Figure 2A shows that in cells over-expressing pcDNA3.1+MTF-1there was a statistically significant increase in b-Syn 21.1/20.6 kb promoter stably-transfected with the empty expression vector, pcDNA3.1+.We also tested the affect of MTF-1 on 21.1/20.6 kb in U-87 MG cells (Figure 2B) and although the basal activity of 21.1/20.6 kb was only slightly above empty vector (pGL3basic), there was a similar trend for that observed in the SH-SY5Y cells.We also investigated whether MTF-1 affected two larger promoter constructs containing the MRE (23.7kb/ATG & -0.9 kb/ATG) and a larger fragment without the MRE (26.0/20.9kb) (Figure 2C).All three constructs were not significantly affected by MTF-1 overexpression, however, the two MRE-containing promoter fragments did trend to the effect observed for the b-Syn 21.1/20.6 kb construct.This less dramatic effect on the two larger MREcontaining promoter fragments could be due to the presence of repressor elements counteracting the effect of MTF-1 on the MRE, which may also explain the initial lower basal activity observed for these two promoter constructs (Figure 1B).We investigated whether the addition of copper from a low basal level affected the activity of the 21.1/20.6 kb promoter construct and found that in serum-free media, where copper levels are 5.2 nM [25], the activity of the 21.1/20.6 kb fragment was abolished, even in the presence of over-expressed MTF-1 (Figure 2E).When we added copper at 10 mM and 50 mM the basal activity was restored, as well as MTF-1 induction of the promoter, although the overall activity was not as high as in normal conditions (Figure 2A).In normal serum media the concentration of copper is ,2 mM, suggesting that other factors in normal serum, not present in the B27 supplement, maybe important for normal promoter activity.These results suggest that copper is important for proper b-syn promoter function, as well as MTF-1 regulation of the promoter.

MTF-1 binding to the b-Syn promoter MRE
To assess whether this increase in promoter activity was specific to the MTF-1 binding site or an indirect effect of MTF-1 overexpression, we abolished the MRE binding site in b-Syn 21.1/ 20.6 kb by site-directed mutagenesis.Figure 2D shows that the inserted mutation at 2774 bp of the b-syn promoter, corresponding to the putative MTF-1 binding site, abolishes promoter activity to near basal levels, regardless of MTF-1 over-expression, as can be seen when compared to pGL3basic.Therefore, it appears that this binding site is critical for b-Syn promoter activity.
To determine whether the observed MTF-1 regulation of the b-Syn promoter was a consequence of MTF-1 binding to the b-syn or an indirect effect of MTF-1 on the b-syn promoter activity, we performed EMSAs.EMSA analysis showed that in cells over-expressing the transcription factor MTF-1 there was an increase in binding for the normal MRE-containing probe compared to the mutant probe as indicated by the band intensity shift when compared to unlabelled probe (Figure 3A & B).In comparison to pcDNA3.1+over-expressing cells, this effect was less intense (Figure 3C), although more intense than the mutant probe in cells over-expressing MTF-1, suggesting that endogenous levels of MTF-1 are present.These results show that MTF-1 binds to MRE located in the 21.1/20.6 kb region of the b-Syn 59promoter.

MTF-1 regulates b-Synuclein expression
We looked to see whether the observed interaction of MTF-1 with the b-Syn promoter, as well as the increase in promoter activity due to over-expression of MTF-1, affected mRNA and protein levels.QPCR analysis was performed to see whether overexpression of MTF-1 affected b-Syn transcription levels.Figure 4A shows that over-expression of MTF-1 has a statistically significant effect on b-Syn gene expression when compared to control cells stably transfected with the empty vector, pcDNA3.1+.This effect of MTF-1 on gene expression is specific to b-Syn as shown when compared to the effect on the other synucleins, a-Syn and c-Syn (Figure 4A).
Western immunoblot analysis shows that in cells overexpression of MTF-1 there was a significant effect on b-Syn protein expression, with a 30-fold increase in protein abundance, when compared to control cells (Figure 4B).For comparison, we also looked at the effect of MTF-1 over-expression on a-Syn protein levels and found no significant change in expression.

Discussion
We analysed the beta-Synuclein (b-Syn) putative 59-promoter and identified a region 21.1/20.6 kb upstream of the translational start site to be important in the activity of a luciferase reporter construct in human dopaminergic SH-SY5Y cells.Moreover, we identified a metal response element (MRE) in this promoter region and showed that it was regulated by metal transcription factor-1 (MTF-1) binding.We also demonstrated that MTF-1 regulates b-Syn expression at both the gene and protein levels.
Human dopaminergic SH-SY5Y cells are commonly used as a model system in the study of Parkinson's disease (PD).SH-SY5Y cells can either be undifferentiated or differentiated and it has been suggested that there are potential using the former as a model system in for these studies [26].However, we have previously shown that undifferentiated SH-SY5Y cells express key dopaminergic markers [27], including the dopamine transporter, and were therefore confident in using these cells for the current study.
MTF-1 is a zinc-finger protein transcription factor that induces expression of metallothioneins and other genes involved in metal homeostasis in response to heavy metals such as cadmium, zinc, copper, and silver [28][29][30][31].The protein is a nucleocytoplasmic shuttling protein that accumulates in the nucleus upon heavy metal exposure and binds to promoters containing a metal-responsive element (MRE; core consensus sequence TGCRCNC) [32].MTF-1 is also believed to be activated under other stressful conditions, other than heavy metal load, including oxidative stress and hypoxia [33].
Emerging evidence now suggests that metal homeostasis is strongly linked to the oxidative stress response in many cells types [34].PD is a progressive neurodegenerative disorder in which oxidative stress and metal toxicity is implicated as a major causative factor [35,36].Parkin, a ubiquitin ligase, has been linked to familial form of PD.A Drosophila disease model lacking Parkin, given zinc-supplemented food showed improved condition, including extended lifespan and improved motoric abilities [37].As a zinc-sensing protein, MTF-1 could conceivably be a way through which zinc-supplementation manifests its positive effects.Moreover, overexpression of MTF-1 was found to improve the life-span of Drosophila exposed to oxidative stress and metal toxicity suggesting a common mechanism for protection against both oxidative stress and metal toxicity [38].Therefore, MTF-1 could potentially have beneficial effect in synucleinopathies.
There is currently no literature suggesting a role of b-Syn in relation to metal homeostasis.However, b-Syn has been suggested to bind copper [39,40].A recent study has shown that it has for copper as a-Syn and that the co-ordination of copper binding is similar [40].a-Syn not only binds copper but copper binding increases its aggregation [41][42][43] and results in the formation of toxic oligomeric species that kill cells both exogenously and endogenously [22,27].As b-Syn has been shown to decrease the aggregation of a-Syn [44], it is conceivable that this process might involve prevention of copper binding to a-Syn by either binding copper itself or a physical interaction with a-Syn that inhibits copper binding.Alternatively, b-Syn might outcompete a-Syn for copper due to increased expression.In this study we showed that copper appears necessary for b-syn promoter function, as well MTF-1 induction of the b-Syn promoter.Indeed, MTF-1 has been shown to act as a coppersensing transcriptional activator to regulate human prion gene expression [45], a protein also relevant to neurodegeneration.Increased expression/activity of MTF-1 in response to altered copper levels in the cell could conceivably result in increased levels of b-Syn, protecting the cell from adverse copper binding by a-Syn that could lead to the formation of toxic copper binding oligomers.
MTF-1 altered b-Syn gene regulation and mRNA and protein expression levels.As can be observed in Figure 4 there was a marked increase in protein expression (,30-fold) when compared to the increase in mRNA expression (,2-fold).Simply measuring an mRNA expression change does not necessary reflect a change in protein expression [46].There is support in the literature for discrepancies between mRNA and corresponding protein levels; Pradet-Balade et al. (2001) suggested that the abundance of a protein at a given mRNA expression level may vary 30-fold [47].Furthermore, in Figure 4A the control level (no MTF-1) is already relatively high but this doesn't lead to protein expression that can be detected in the western blot (Figure 4B).This is likely because the turnover of b-Syn mRNA is quite rapid and the protein synthesis machinery is unable to compete, but in the presence of MTF-1 the rate of mRNA synthesis is greater which leads to significant amounts of mRNA being utilised for protein expression.Thus the apparent doubling of mRNA is not informative as to the expected level of protein expression..As previously highlighted, b-Syn may be a natural negative regulator of the disease-associated a-Syn.Furthermore, transgenic mice over-expressing b-Syn, showed a marked reduction in a-Syn aggregation and protein expression [44].The importance of the balance between levels of the synucleins is highlighted by the observation that b-Syn inhibits a-Syn aggregation in vitro and in vivo [2,48,49].It's conceivable that due to cellular stresses, such as heavy metal toxicity or oxidative stress, MTF-1 expression is induced, which consequently alters the regulation of several genes, including b-Syn.b-Syn, in turn, could potentially prevent a-Syn aggregation and/or toxic species.This process may be disrupted in patients with synucleinopathies, as in contrast to control patients, there is a dramatic increase in a-Syn and decrease in b-Syn mRNA levels in the substantia nigra of PD, DLBD and a LB variant of AD patients [14].Therefore MTF-1 could provide a potential target for disease therapy and this needs to be explored further.

Figure 1 .
Figure 1.b-Synuclein 59-genomic structure and basal promoter activity.(A) Depicts the 59-genomic structure of human b-Synuclein gene.Exons are depicted as closed boxes, and 59-region and introns are black lines.Black box indicates the translational start site in exon 3. Lower panel, luciferase constructs ranging across the 59-region.* Indicates metal response element (MRE-TGCGCTC).(B & C) Reporter gene assays using Dual-Luciferase TM shows the basal activity of b-Syn promoter fragments and empty pGL3basic vector in SH-SY5Y and U87 MG cells respectively.RLU = relative luciferase units.doi:10.1371/journal.pone.0017354.g001

Figure 3 .Figure 4 .
Figure 3. EMSA analysis.Nuclear extracts from SH-SY5Y cells were stably transfected with (A & C) pcDNA3.1+MTF-1or (B) pcDNA3.1+.A & B are probed with normal probe and C with mutant probe.Lane 1 is the migration of free-probe in the absence of nuclear extract and therefore no shift observed.Lane 2 is either the MRE-containing double-stranded DNA probe (A & C) and shows a signal shift due to transcription factor binding or the mutant MRE-containing probe (B) and shows a diminished signal shift.Lane 3 shows that the signal shift can be inhibited from excess non-labelled probe.doi:10.1371/journal.pone.0017354.g003

Table 1 .
Oligonucleotides used for cloning.