G‐quadruplex‐binding small molecules ameliorate C9orf72 FTD/ALS pathology in vitro and in vivo

Abstract Intronic GGGGCC repeat expansions in C9orf72 are the most common known cause of frontotemporal dementia (FTD) and amyotrophic lateral sclerosis (ALS), which are characterised by degeneration of cortical and motor neurons, respectively. Repeat expansions have been proposed to cause disease by both the repeat RNA forming foci that sequester RNA‐binding proteins and through toxic dipeptide repeat proteins generated by repeat‐associated non‐ATG translation. GGGGCC repeat RNA folds into a G‐quadruplex secondary structure, and we investigated whether targeting this structure is a potential therapeutic strategy. We performed a screen that identified three structurally related small molecules that specifically stabilise GGGGCC repeat G‐quadruplex RNA. We investigated their effect in C9orf72 patient iPSC‐derived motor and cortical neurons and show that they significantly reduce RNA foci burden and the levels of dipeptide repeat proteins. Furthermore, they also reduce dipeptide repeat proteins and improve survival in vivo, in GGGGCC repeat‐expressing Drosophila. Therefore, small molecules that target GGGGCC repeat G‐quadruplexes can ameliorate the two key pathologies associated with C9orf72 FTD/ALS. These data provide proof of principle that targeting GGGGCC repeat G‐quadruplexes has therapeutic potential.


Appendix supplementary methods Fluorescence anisotropy titration
Florescence anisotropy was performed for DB1246 and DB1247 but was not possible for DB1273 due to its weaker fluorescence. Desalted 5'-rGrGrGrC rCrGrGrGrG rCrCrGrGrG rGrCrCrGrG rGrG -3' (G-Q-RNA) (IDT, Coralville, IA, USA) was suspended in 50 mM KCl, 10 mM Tris-HCl (pH 7.4) in nuclease-free water to a final concentration of 1 mM. Samples were heated to 98 º C for 10 minutes, cooled for 45 minutes and stored at 4 ºC before use. Fluorescence polarization measurements were obtained on a Variant Cary Eclipse Fluorescence Spectrophotometer (Agilent Technologies, Santa Clara, CA) equipped with a polarization attachment. The excitation and emission polarizers were inserted between the light source and detector, respectively. Measurements were converted into anisotropy value by using Equations 1-2: = − * ℎ +2 * * ℎ (1) Where r is the anisotropy value; I vv is the fluorescence intensity of compound when excitation and emission polarizers are aligned in vertical position; I vh is the fluorescence intensity of compound when the excitation and emission polarizer are in vertical and horizontal position, respectively; I hv is the fluorescence intensity of compound when excitation and emission polarizer align in horizontal and vertical positions; I vv is the fluorescence intensity of compound when excitation and emission polarizer align in vertical position; GF stands for the grating factor of the polarizers. GF values were introduced to the equation as a means to correct the unequal sensitivity of the excitation and emission polarizers (Lea et al, 2011). To determine equilibrium constants the compounds were added into a quartz fluorescence cuvette at a fixed concentration of 50 nM and G-Q-RNA was titrated into the cuvette. The amount of G-Q-RNA was calculated before each titration experiment and the concentration of fluorophore was corrected after each addition for the slight dilution. Measurements were taken at the compound maximum emission wavelength. Both excitation and emission slits were at 20 nm. Each set of measurements were taken from a 50 s average reading measurement. The experimental buffer conditions were 50 mM KCl, 10 mM Tris-HCl pH 7.3 and 1 mM EDTA. The dissociation constant (K D ) of DB compounds to G-Qs was obtained by fitting the change in anisotropy value of the compound titration by using the following equation for a 1 to 1 binding reaction: (3) where r min is the anisotropy value of the unbound molecule; r max is the anisotropy value of fully bound molecule; L: total concentration of fluorophore; D: total concentration of receptor (G-Qs) (Roehrl et al, 2004) Southern blotting RNA free genomic DNA was extracted from cultured cells using the Qiagen Puregene kit following manufacturer's instructions. 5 μg of genomic DNA was digested using EcoRI and BamHI (NEB) and run on a 0.8% agarose gel at 60 V overnight. The gel was agitated for 30 min in 1L of denaturation buffer (National Diagnostics) followed by 1L of neutralisation buffer (National Diagnostics). DNA was transferred onto a positively charged nylon membrane (Roche Applied Science) via capillary action overnight using 20X SSC as transfer buffer. Following transfer, DNA was baked onto the nylon membrane at 80 ºC for 2 hours. Membrane was briefly wetted in water before being incubated in DIG Easy Hyb buffer (Roche) at 46 ºC for 3 hours. It was then transferred into fresh DIG Easy Hyb buffer with 100 μg/ml salmon sperm DNA (Thermo) and 1 kb DIG labelled probe (Fratta et al, 2014) and left overnight at 46 ºC. The membrane was washed for 15 minutes in 0.5X SSC with 0.1% SDS at a rising temperature from 46ºC to 65ºC, before being washed twice more at 65ºC. Following washes, the membrane was processed using the Roche DIG wash and block buffer set following manufacturer's instructions and using the AP-conjugated anti-digoxigenin (11093274910, Roche) at 1:10,000. The membrane was incubated in CDP-star ready to use (Roche) sealed in acetate and exposed to film (Roche Lumi-Film chemiluminescent detection film) for >1 hour before being developed using an X4 automatic processor (XOgraph).

Immunofluorescence staining of iPSC-neurons
iPSC-neurons were grown on laminin coated coverslips and were fixed in 4% paraformaldehyde in PBS for 15 minutes then washed three times in PBS supplemented with 0.3% Triton-X (PBST) and blocked with 5% BSA in PBST for 30 minutes. Cells were incubated overnight at 4 °C with primary antibody (ChAT 1:100 Millipore AB144P and TUJ1 1:1000 Biolegend 802001) in 5% BSA in PBST. Cells were washed in PBST three times then incubated with complementary Alexa Fluor secondary antibody (1:500) dilution for one hour at room temperature. Cells were washed once in PBST containing DAPI for ten minutes then twice more in PBST. Coverslips were mounted onto glass slides in Dako mounting medium. Images were acquired using an LSM700 confocal microscope (Zeiss) using a planapochromat 40x/1.4 NA oil immersion objective.

XTT assay in human fibroblasts
WI-38 non-transformed lung fibroblasts were seeded at a density of 6,000 cells/well in 96 well-plates in Eagle`s Minimum Essential Medium supplemented with 10% FBS, 1% nonessential amino acids and 1% L-glutamine and maintained at 37C with 5% CO 2 . 24 hrs later, fibroblasts were treated with 0.04, 0.1, 0.2, 0.4, 1, 2, 4, 10, 20 and 40 μM of DB1246, DB1247, DB1273 or cisplatin as a positive control with eight replicates. Cell viability was measured after 4 days using the XTT assay kit II (Roche) according to the manufacturer instructions.
Drosophila larvae quantitative RT-PCR Total RNA was extracted from 5 larvae for each replicate using Trizol (Invitrogen) according to the manufacturer's instructions and treated with DNAse I (Roche). cDNA was synthesized using 1 μg of total RNA for all samples, with a SuperScript III first strand cDNA synthesis kit (Invitrogen) and an equimolar mixture of oligo dT and random hexamer primers. Real-time qRT-PCR was carried out using Power SYBR Green Master Mix (Applied Biosystems) with previously described primers for the pUASTattB transgene (Tran et al, 2015) and normalised to α-tubulin. The amplified transcripts were quantified using the comparative Ct method and presented as normalized fold expression (ΔΔCt). Oligonucleotide sequences are provided in Table S2.

Drosophila confocal imaging
To investigate bioavailability of DB1273, daGAL4>36R first instar larvae were treated with vehicle, or 1 mM of DB1273, and dissected after 5-6 days at the third instar stage. 2-3 larvae were dissected per condition. Larval tissues were dissected in PBS and then fixed in 4 % PFA in PBS at room temperature for 20 mins. Larval tissues were washed in PBS and whole mounted onto glass slides in VECTASHIELD® mounting medium with DAPI (Vector Laboratories, H-1200). Z-stack images of larval tissues were acquired using an LSM710 confocal microscope (Zeiss) with a plan-apochromat 40x/1.4 NA oil immersion objective. Tissues were excited using the 488 nm laser excitation to detect DB1273. DB1273 fluorescence intensity was set to peak for individual larval tissues, using the same imaging settings for 0 mM and 1 mM treatments for each tissue.

Appendix supplementary figures
Appendix Fig. S1. Dose-response of small molecule binding to antisense GGCCCC repeats. A FRET assay was used to measure the difference in melting temperature (ΔT m ) of (G 2 C 4 ) 4 RNA or DNA G-quadruplexes in the presence of 0 -3 μM DB1246, DB1247 and DB1273, compared to vehicle (DMSO). Data is shown as mean ± s.d., n=3 independent experiments.
Appendix Fig. S2. Small molecules alone give no CD signal. (G 4 C 2 ) 4 RNA alone (red line) gives a characteristic CD spectrum, whereas DB1246, DB1247 and DB1273 alone (black lines) give no signal, confirming that the induced CD signal when RNA and small molecule are combined ( Figure 1D) is due to a specific small molecule-RNA interaction.
Appendix Fig. S3. (G 4 C 2 ) 4 RNA CD denaturation curves in the presence of each small molecule. CD spectra were measured during a thermal unfolding, with increasing temperature from 15 to 95 °C in steps of 10 °C. All three compounds increase the melting temperature (T m ) of the (G 4 C 2 ) 4 RNA. Values represent mean ± s.d. (n=3 independent experiments). T m was calculated using the van't Hoff equation.
Appendix Fig. S4. Fluorescence anisotropy titration of DB1246 and DB1247 to G 4 C 2 G-Q RNA. Change in anisotropy value r were plotted against the concentration of G 4 C 2 G-Q RNA. Compound concentrations were held near 50 nM in 10 mM Tris-HCl, 50 mM KCl in nucleasefree water (pH 7.4) with increasing concentration of RNA to obtain K D values for 410 nM ± 79 258 ± 60 nM for DB1246 and DB1247 respectively, using a 1 to 1 binding model as described in the methods section. The MSD immunoassay specifically detects (GP) 7 peptide but not other dipeptide repeat proteins (GR) 7 or (PR) 7 . (B) Representative standard curve of MSD response at different concentrations of (GP) 7 peptide calibrator, R 2 =0.99. Lower limit of detection is 0.034ng/ml. Poly(GP) is detected by MSD immunoassay in both C9orf72 (C9) patient iPSC-motor neurons (C) and iPSC-cortical neurons (D) and not in control (Ctrl) lines. Data in (C) are the average and SEM of 3 inductions per line (3 independent C9 lines and 2 Ctrl lines). Data in (D) is the average response of two technical replicates for one induction of the same lines as in (C).
Appendix Fig. S8. GGGGCC repeat G-quadruplex binding small molecules at 1 or 4 µM doses do not reduce poly(GP) in patient iPSC-motor neurons. Poly(GP) was detected by MSD immunoassay in C9orf72 patient iPSC-motor neurons treated for 4 days with 1 or 4 µM of DB1246, DB1247 and DB1273. Relative to those treated with DMSO as a control there is no reduction in poly(GP) in iPSC-motor neurons after treatment with the three compounds. Treatments on 3 independent C9orf72 iPSC lines with 1-6 differentiations per line. * p<0.05, ** p<0.01, one sample two-tailed t-test versus normalised control. p=0.0044 (DB1247 4 µM), p=0.0413 (DB1273 4 µM).