The G-quadruplex DNA stabilizing drug pyridostatin promotes DNA damage and downregulates transcription of Brca1 in neurons

Abstract

The G-quadruplex is a non-canonical DNA secondary structure formed by four DNA strands containing multiple runs of guanines. G-quadruplexes play important roles in DNA recombination, replication, telomere maintenance, and regulation of transcription. Small molecules that stabilize the G-quadruplexes alter gene expression in cancer cells. Here, we hypothesized that the G-quadruplexes regulate transcription in neurons. We discovered that pyridostatin, a small molecule that specifically stabilizes G-quadruplex DNA complexes, induced neurotoxicity and promoted the formation of DNA double-strand breaks (DSBs) in cultured neurons. We also found that pyridostatin downregulated transcription of the Brca1 gene, a gene that is critical for DSB repair. Importantly, in an in vitro gel shift assay, we discovered that an antibody specific to the G-quadruplex structure binds to a synthetic oligonucleotide, which corresponds to the first putative G-quadruplex in the Brca1 gene promoter. Our results suggest that the G-quadruplex complexes regulate transcription in neurons. Studying the G-quadruplexes could represent a new avenue for neurodegeneration and brain aging research.

Keywords: BRCA1; DNA damage; G-quadruplex; neurodegeneration; transcription.

Figure 1. Pyridostatin is neurotoxic for primary cortical neurons

(A) G-quadruplex is a non-canonical DNA secondary structure. Four guanine molecules (a single guanine is in red) can assemble into a square planar structure. The structure of a G-quadruplex is stabilized by hydrogen bonds between guanines and the interactions with a monovalent cation (Na⁺ or K⁺) resided in the central channel. (B) Repetitive guanine-rich DNA or RNA sequences have the potential to form G-quadruplex structures (C). (D) An example of survival analysis. Primary cortical neurons were transfected with mApple (a morphology and viability marker) and tracked with an automated microscope. Images collected every 24 h demonstrate the ability to return to the same field of neurons and to follow them over time. Each image is a montage of non-overlapping images captured in one well of a 24-well plate. Scale bar is 100 μm. A region from the original images at different time points is zoomed in to demonstrate longitudinal single-cell tracking (bottom panel). Red arrows depict two neurons that degenerate before 96 h after transfection. Note that neurites of the neuron 1 retract overtime. Scale bar is 20 μm. (E) Primary cortical neurons were transfected with mApple and treated with a vehicle or with different concentrations of pyridostatin (PDS; 0.01–5 μM). Transfected neurons were tracked with an automated microscope. Risk of death curves demonstrate that pyridostatin is neurotoxic. ***p<0.0001 (log-rank test). N.s., non-significant. Three hundred neurons were analyzed from three independent experiments. (F) Primary cortical neurons were treated with a vehicle (upper panel, control) or with 1 μM pyridostatin overnight (lower panel; PDS), fixed and stained with antibodies against MAP2c (red) and synapsin (green). Scale bar is 10 μm. (G) Quantification of the synapsin fluorescence intensity from (F). ***p<0.0001 (t-test). (H) Quantification of the neurite density from (F). MAP2c staining was used by the algorithm to identify and analyze neurites. ***p<0.001 (t-test). Three hundred neurons were analyzed from two independent experiments.

Figure 2. Pyridostatin promotes the formation of 53BP1-positive puncta in primary neurons

(A) Primary cortical neurons were transfected with GFP and mApple-53BP1trunc constructs, and then treated with a vehicle (left panel; control) or with 1 μM pyridostatin (right panel; PDS). Neurons were imaged with an automated microscope every 24 h for 3 days. Scale bar is 5 μm. (B) Quantification of the mApple-53BP1trunc puncta index from (A) at different times. The puncta index was estimated by measuring the standard deviation of the 53BP1 fluorescence intensity. Note that 53BP1 puncta index is higher in pyridostatin-treated neurons than control neurons. *p<0.01, **p<0.001, and ***p<0.0001 (t-test). A.u., arbitrary units. Two hundred neurons were analyzed from two independent experiments. (C) Primary cortical neurons were treated with a vehicle (upper panel; control) or with 1 μM pyridostatin (bottom panel; PDS) overnight, fixed, and stained for MAP2c (red), a marker of DNA damage 53BP1 (green), and with the nuclear Hoechst dye (blue). Scale bar is 10 μm. (D) Quantification of the 53BP1 puncta index from (C). Pyridostatin (PDS) increased the 53BP1 puncta index compared to control neurons (cont). ***p<0.0001 (t-test). A.u., arbitrary units. Three hundred neurons were analyzed from three independent experiments.

Figure 3. G-quadruplex-stabilizing drugs upregulate γH2A.X

(A) Primary cortical neurons were treated with a vehicle (control) or with 1 μM pyridostatin (PDS) or with 200 nM bortezomib (bortezomib) or with 10 μM cisplatin (cisplatin) overnight, fixed, and stained for MAP2c (red), a marker of DNA DSBs phosphorylated histone H2A variant X, γH2A.X (green), and with the nuclear Hoechst dye (blue). Scale bar is 10 μm. (B) The puncta index was estimated by measuring the standard deviation of the γH2A.X fluorescence intensity. Primary cortical neurons were treated with a vehicle (control, cont) or with 1 μM pyridostatin (PDS), or with 200 nM bortezomib (bort), or with 10 μM cisplatin (cisp) overnight, then fixed, and immunostained with antibodies against MAP2c and γH2A.X, and co-stained with the nuclear Hoechst dye (blue). ***p<0.0001 (t-test). N.s., non-significant (p_bort=0.1995; p_cis=0.8228). A.u., arbitrary units. Four hundred neurons were analyzed from three independent experiments. (C) Primary cortical neurons were treated with a vehicle (control, cont) or with a G-quadruplex stabilizing drug, TmPyP4 (1 μM, overnight), then fixed, immunostained against MAP2c and γH2A.X, and co-stained with the nuclear Hoechst dye (blue). ***p<0.0001 (t-test). A.u., arbitrary units. Four hundred neurons were analyzed from three independent experiments.

Figure 4. Pyridostatin downregulates BRCA1 levels in primary neurons

(A) A scheme of the Brca1 gene and its promoter showing putative G-quadruplex sequences. (B) Primary neurons were treated with a vehicle (control, -) or with different concentrations of pyridostatin (PDS; 1, 2 or 5 μM) overnight. Then neurons were collected, lysed, and samples were processed with western blotting. Note that pyridostatin reduced BRCA1 protein levels in a dose-dependent manner. Actin was used as a loading control. (C) Quantification of BRCA1 protein levels normalized to actin from (B). **p_{(cont-1 μM)}=0.005, **p_{(cont-2 μM)}=0.0015, **p_{(cont-5 μM)}=0.0043 (ANOVA). N.s., non-significant, p_{(1 μ M−2 μM)}=0.0547, p_{(1 μM−5 μM)}=0.0854. Results were pooled from five independent experiments. (D) Primary neurons were treated with a vehicle (control, -) or with a G-quadruplex stabilizing drug, TmPyP4 (1 μM), overnight. Neurons were lysed, and samples were processed by western blotting. TmPyP4 reduced BRCA1 protein levels. Actin was used as a loading control. (E) Quantification of BRCA1 protein normalized to actin from (D). *p<0.0276 (t-test). Results were pooled from four independent experiments. (F) Primary neurons were treated with a vehicle (control; -) or with 2 μM pyridostatin (PDS; +) overnight. The expression of Brca1 was determined by qRT-PCR relative to Actin. *p=0.0445 (t-test). Results were pooled from three independent experiments. (G) Primary neurons were treated with a vehicle (control; -) or with 2 μM pyridostatin (PDS; +) overnight. Relative expression of the Actin, Gapdh and Tbp genes was determined by qRT-PCR. *p=0.0112, **p=0.0039 (t-test). N.s., non-significant (p=0.7583). Results were pooled from three independent experiments. (H) Primary neurons were treated with a vehicle (control; -) or with 2 μM pyridostatin (PDS; +) overnight. The expression of Brca1 was determined by qRT-PCR normalized to Tbp. **p=0.0033 (t-test). Results were pooled from three independent reactions.

Figure 5. Ectopically expressed BRCA1 mitigates DNA damage associated with pyridostatin treatment

(A) Primary cortical neurons were transfected with GFP or with GFP-BRCA1 constructs, and then treated with a vehicle (control, GFP) or with 1 μM pyridostatin (GFP+PDS and GFP-BRCA1+PDS), overnight, fixed, and stained for a marker of DNA DSBs phosphorylated histone H2A variant X, γH2A.X (red). Scale bar is 10 μm. (B) The puncta index was estimated by measuring the standard deviation of the γH2A.X fluorescence intensity from (A). ***p<0.0001 (ANOVA). N.s., non-significant, p=0.243. A.u., arbitrary units. Two hundred neurons were analyzed from two independent experiments. (C) Primary cortical neurons were transfected with GFP or with GFP-BRCA1 constructs, and then treated with a vehicle (control, GFP) or with 1 μM pyridostatin (GFP+PDS, GFP-BRCA1+PDS), overnight, fixed, and stained for a marker of DNA DSBs 53BP1 (red). Scale bar is 10 μm. (D) The puncta index was estimated by measuring the standard deviation of the γH2A.X fluorescence intensity from (C). ***p<0.0001 (ANOVA). N.s., non-significant, p=0.441. A.u., arbitrary units. Two hundred neurons were analyzed from two independent experiments.

Figure 6. The G-quadruplex-specific antibody BG4 binds to a G-rich oligonucleotide derived from the promoter of Brca1

(A) The Cy5-conjugated BRCA1-G1 (G1), BRCA1-G2 (G2), and BRCA1-cont (cont) oligonucleotides (see Methods) were heat-denatured and then slow-cooled in the presence of K⁺ (KCl) to allow the formation of a secondary structure. 1.5 pmoles of each oligonucleotides (oligo) and 0 (a buffer alone), 30 or 50 ng of the BG4 antibody were incubated in a buffer, which contained 100 mM KCl. Note the bands that correspond to the free BRCA1-G1 (at the bottom) and the BRCA1-G1 bound to the BG4 antibody (at the gel top), both indicated with arrow heads. An asterisk marks the faint bands, which resulted from a loading dye. (B) The Cy5-conjugated BRCA1-G1 (G1), BRCA1-G2 (G2), and BRCA1-cont (cont) oligonucleotides were prepared in the presence of Na⁺ (NaCl). 1.5 pmoles of each oligonucleotides (oligo) and 0 (a buffer alone), 30 or 50 ng of the BG4 antibody were incubated in a buffer, which contained 100 mM NaCl. Note the bands that correspond to the free BRCA1-G1 (at the bottom) and the BRCA1-G1/BG4 complex (at the gel top), depictured with arrow heads. An asterisk marks the loading dye bands. (C) The Cy5-labeled BRCA1-G1 (G1), BRCA1-G2 (G2), and BRCA1-cont (cont) oligonucleotides were prepared in the presence of Li⁺ (LiCl). 1.5 pmoles of each oligonucleotides (oligo) and 0 (a buffer alone), 30 or 50 ng of the BG4 antibody were incubated in a buffer, which contained 100 mM LiCl. Note the bands that correspond to the free BRCA1-G1 (at the bottom) and a weak BRCA1-G1/BG4 complex (at the gel top), depictured with arrow heads. An asterisk marks the loading dye bands.

Figure 7. A model of the Brca1 gene downregulation by pyridostatin

(A) Transcription of the Brca1 gene in primary neurons under basal conditions. (B) Pyridostatin binds to and stabilizes the G-quadruplex structures present in the Brca1 gene promoter sequence. The pyridostatin/DNA complex stalls DNA polymerase during transcription and downregulates Brca1 gene expression. (1) Brca1 downregulation may then hinder DNA damage repair and DNA DSBs accumulate as a result. (2) In addition, DNA damage may occur due the action of endo-nucleases through a mechanism of transcription–coupled–repair poisoning [25].