Regulation of alpha-synuclein expression by poly (ADP ribose) polymerase-1 (PARP-1) binding to the NACP-Rep1 polymorphic site upstream of the SNCA gene

Abstract

Alleles at NACP-Rep1, the polymorphic microsatellite repeat located approximately 10 kb upstream of the alpha -synuclein gene (SNCA), are associated, in some reports, with differing risks of sporadic Parkinson disease (PD). We showed previously that NACP-Rep1 acts as a negative modulator of SNCA transcription, with an effect that varied threefold among different NACP-Rep1 alleles. Given that duplications and triplications of SNCA have been implicated in familial Parkinson disease (PD), even a 1.5-2-fold increase in alpha -synuclein expression may, over many decades, contribute to PD. Thus, the association of different NACP-Rep1 alleles with PD may be a consequence of polymorphic differences in transcriptional regulation of SNCA. Here we aimed to identify the factor(s) that bind to NACP-Rep1 and potentially contribute to SNCA transcriptional modulation, by pulling down proteins that bind to NACP-Rep1 and identifying them by mass spectrometry. One of these proteins was poly-(ADP-ribose) transferase/polymerase-1 (PARP-1), a DNA-binding protein and transcriptional regulator. Electrophoresis mobility shift and chromatin immunoprecipitation assays showed specific binding of PARP-1 to NACP-Rep1. Inhibition of PARP-1's catalytic domain increased the endogenous SNCA mRNA levels in cultured SH-SY5Y cells. Furthermore, PARP-1 binding to NACP-Rep1 specifically reduced the transcriptional activity of the SNCA promoter/enhancer in luciferase reporter assays. This down-regulation effect of PARP-1 depended on NACP-Rep1 being present in the construct and was abrogated by inhibiting PARP-1's catalytic activity with 3-aminobenzamide. The association of different NACP-Rep1 alleles with PD may be mediated, in part, by the effect of PARP-1, as well as other factors, on SNCA expression.

Figure 1

NACP-Rep1 binds protein/s from nuclear SH-SY5Y extracts. EMSAs were performed with DIG-labeled double-stranded 80 bp oligo including the entire NACP-Rep1 sequence. Lane 1 shows the DIG-NACP-Rep1 only. SH-SY5Y nuclear extract was used in the binding assay (lanes 2 and 3). Unlabeled ds-NACP-Rep1 oligo was used as the specific competitor (lane 3). Unlabeled poly[dCI] and poly[d(AT)] were used in all lanes as a nonspecific competitor to avoid nonspecific DNA-binding. The plus (+) sign indicates what is included in each reaction while a minus sign (−) indicates what is excluded from the reaction. The number in the upper panel indicates the lane number. A bracket indicates the proteins-NACP-Rep1 complex. The lower smear, designated by an asterisk (*), indicates the DIG-NACP-Rep1.

Figure 2

Identification of PARP-1 as NACP-Rep1 binding protein. A, Commassie stained gel of the NACP-Rep1 binding proteins that were purified from SH-SY5Y nuclear extract. The arrow indicates the 120-kDa protein’s band that was used for mass spectrometry identification. B, Mascot results of the mass spectrometry identification of the 120-kDa band. PARP-1 was identified by three peptides with the highest score value (178).

Figure 3

PARP-1 binds to the NACP-Rep1 sequence. A, EMSAs were performed with DIG-labeled double-stranded 80-bp oligo including the entire NACP-Rep1 sequence. Lane 1 shows the DIG-NACP-Rep1 only. Purified human PARP-1 was used in the binding assay. Anti-PARP-1 antibody was used for the super shift reaction. Unlabeled ds-NACP-Rep1 oligo was used as the specific competitor, and unlabeled ds-SNCA-exon 4 oligo was used as the unspecific competitor. The plus sign (+) indicates what is included in each reaction, whereas a minus sign (−) indicates what is excluded from the reaction. The number in the upper panel is the lane number. Close arrow indicates the PARP-1-NACP-Rep1 complex (shifted band). Open arrow indicates the PARP-1-antibody-NACP-Rep1 complex (super-shift band). B, ChIPs were carried out using anti-PARP-1 antibody and IgG antibody. The input control represents the whole DNA of SH-SY5Y cells. Following the ChIP reactions, a PCR analysis was performed using three different sets of primers. Amplification around the NACP-Rep1 region, control amplifications of the SNCA 3′ UTR and the Inpp5b gene. The yeast DNA and no DNA control reactions for background failed to amplify. The arrow indicates the PCR product of the NACP-Rep1 region (270 bp). The unblackened arrowhead indicates the PCR product of the SNCA 3′ UTR (215 bp), and the blackened arrowhead indicates the PCR product of Inpp5b (284 bp).

Figure 4

Real-time analysis of SNCA mRNA levels in SH-SY5Y cell after 3-AB treatment. A, Validation curve of the ΔΔ real-time assay for relative quantization of SNCA mRNA relatively to GAPDH mRNA. A relative efficiency plot of SNCA and GAPDH was formed by plotting the log input amount (ng of total RNA) versus the ΔCt [Ct(SNCA)−Ct(GAPDH)]. The slop is 0.0638<0.1, which indicates the validation of the ΔΔCt calculation in the range of 1–100 ng RNA. B, Cells were treated with 3AB in different concentrations (0, 3, 10, 15, 30, and 60 mM), and the relative levels of SNCA mRNA to GAPDH mRNA were assessed by real-time PCR and were analyzed by the ΔΔ method. The 3-AB concentration (mM) is indicated on the X-axis, and the fold increase of SNCA mRNA is indicated on the Y-axis. Each data point represents the average of three repeats.

Figure 5

Fold expression of luciferase activity derived by full-length and NACP-Rep1–deleted pASP constructs in SH-SY5Y cells before and after 3-AB treatment. A, Cells were transfected with each construct or pGL 3-Basic. The relative activity for each pASP or pGL 3-Basic plasmid was calculated by dividing the luminescence intensity of the firefly luciferase by that of the cotransfected Renilla luciferase in each independent aliquot of cells and then averaging the three relative luciferase activities seen. The fold expression for pASP and pASP-ΔNACP was then determined by dividing the average relative activity of each construct to that of the average obtained with pGL 3-Basic. B, The ratio of fold expression of luciferase activity derived by the full-length pASP and pASP-ΔNACP in SH-SY5Y cells upon 3-AB treatment. Cells were cotransfected with each of the two constructs (full-length and NACP-Rep1 deletion) or pGL 3-Basic and pRL. In each experiment for each construct, three well were untreated and three were treated with 30 mM 3-AB for 48 h. For each construct, two to three independent experiments were performed. The relative activity for each pASP, with and without 3-AB treatment or pGL 3-Basic plasmid, was calculated by dividing the luminescence intensity of the firefly luciferase by that of the cotransfected Renilla luciferase in each independent aliquot of cells and then averaging the three relative luciferase activities seen. The fold expression for each pASP with and without 3-AB treatment was then determined by dividing the average relative activity of each construct to that of the average obtained with pGL 3-Basic. Then, the ratio of the fold expression for each of the 3-AB treated cells transfected with each construct (pASP and pASP-ΔNACP) relative to the untreated cells transfected with the same construct was determined. The average of the ratios of two to three independent experiments performed on separate days was calculated. A bar in the positive scale of the Y-axis indicates an increase in luciferase activity upon 3-AB treatment, and a bar in the negative scale of the Y-axis indicates a decrease in luciferase activity upon 3-AB treatment.