Human fibroblasts (control: GM07492; HD: GM04281, 17/68 CAG in HTT gene, HD2: GM09197, 21/151 CAG in HTT gene) were obtained from Coriell Repository (Camden, NJ, USA). Fibroblasts were cultured in Minimal Essential Medium (Sigma–Aldrich, St. Louis, MO, USA) with 10% Fetal Bovine Serum (Sigma–Aldrich) supplemented with GlutaMAX (Life Technologies, Carlsbad, CA, USA), non-essential amino acids (Sigma–Aldrich), and Antibiotic Antimycotic Solution (Sigma–Aldrich) at 37°C. All cell cultures were checked for mycoplasma contamination.

Oligonucleotide were synthesized by Future Synthesis (Poznan, Poland) or IDT (Coralville, IA, USA), and duplexes were annealed according to the manufacturer’s instructions. The LNA oligomer was synthetized by Exiqon (Vedbæk, Denmark). The sequences of the synthetic ONs and oligomer used in this study are presented in Table 1.

Cell transfections were performed using Lipofectamine 2000 (Life Technologies) according to the manufacturer’s instructions. The reagents were transfected at 50 nM, unless stated otherwise in figure legends. Oligonucleotide treatment lasted for 4 h; after that time medium was changed. Material for subsequent analyses was collected after 48 h. For fixation, cells were seeded directly on cover slips prior to transfection. The transfection efficiency was monitored using BlockIT fluorescent siRNA (Life Technologies).

Fluorescence in situ hybridization and IF procedures were performed as described previously (Urbanek and Krzyzosiak, 2016). Briefly, cells were fixed in 4% paraformaldehyde at 4°C for 30 min. Permeabilization was performed with 2% acetone for 5 min, and overnight incubation in 70% EtOH. Prior to hybridization, cells were prehybridized in 30% formamide, 2x SSC buffer for 30 min at RT. Hybridization was performed in hybridization buffer (30% formamide, 2x SSC, 200 ng/ml ssDNA, 0.02% BSA, 10% dextran sulfate, 2 mM vanadyl-ribonucleoside, 2 nM probe) at 37°C overnight. Washing was performed with 30% formamide, 2x SSC and 1x SSC at RT. For biotin-labeled probes washings were performed only in 1x SSC and subsequently samples were blocked with 1% BSA buffer for 1 h at RT and incubated with IgG Fraction Monoclonal Mouse Anti-Biotin antibody (1:200, 200-542-211, Jackson ImmunoResearch; West Grove, PA, USA) overnight at 4°C. For signal enhancement samples were additionally incubated for 1 h with AffiniPure F(ab′)₂ Fragment Donkey Anti-Mouse IgG (H+L) antibody labeled with Alexa488 (1:500, 715-546-150, Jackson ImmunoResearch; West Grove, PA, USA). For RNase treatment, cells were subjected to 0.01 μg/ml RNase A for 1.5 h after the permeabilization step. For IF, samples were blocked with 1% BSA buffer for 1 h at RT, incubated with primary anti-huntingtin antibody (1:200, MAB2166, Millipore) at 4°C overnight and with secondary antibody AffiniPure F(ab′)₂ Fragment Donkey Anti-Mouse IgG (H+L) antibody labeled with Alexa488 (1:500) for 1 h. For combined RNA FISH and IF, first, FISH was performed with only 1x SSC washings. Next, samples were blocked with 1% BSA for 30 min and incubated with anti-huntingtin antibody for 3 h and secondary antibody for 45 min. For endosomes visualization anti-Rab5 (1:200, 12666T, Cell Signaling Technology; Danvers, MA, USA) and anti-EEA1 (1:100, 12666T, Cell Signaling Technology) antibodies and secondary antibody AffiniPure F(ab′)₂ Fragment Donkey Anti-Rabbit IgG (H+L) antibody labeled with Alexa488 (1:500, 711-546-152, Jackson ImmunoResearch) were used. Nuclei were imaged with SlowFade Gold with DAPI (Life Technologies). Images were captured with a Leica SP5 confocal microscope.

For CAG-specific ONs detection in cells we used DNA probe Cy3-(CAG)₆CA (Metabion, Germany), for foci imaging we used either LNA-modified DNA probe Cy3-(CTG)₆CT (Exiqon, Denmark) or three HTT-specific 5′-Biotin-labeled DNA probes: GGTAAAAGCAGAACCTGAGCGGCCGTCCATCTTGGACCCGT, TCGAAGGCCTTCATCAGCTTTTCCAGGGTCGCCAT, TCCATAGCGATGCCCAGAAGTTTCTGAAATTCTGGAG (Metabion).

The western blot analysis for the HTT protein (17/68Q tract) was performed as previously described (Fiszer et al., 2011). Briefly, 30 μg of total protein was run on a Tris-acetate sodium dodecyl sulfate (SDS)-polyacrylamide gel (1.5 cm, 4% stacking gel/4.5 cm, 5% resolving gel, acrylamide:bis-acrylamide ratio of 49:1) in XT Tricine buffer (Bio-Rad, Hercules, CA, USA) at 130 V in an ice-water bath. Subsequently, the proteins were wet-transferred to a nitrocellulose membrane (Sigma–Aldrich). All of the immunodetection steps were performed using the SNAPid system (Millipore). The primary antibodies anti-huntingtin (1:1000, MAB2166, Millipore) and anti-plectin (1:1000, ab83497, Abcam, Cambridge, UK) and secondary antibodies anti-mouse HRP-conjugate (1:2000, A9917, Sigma–Aldrich) and anti-rabbit HRP-conjugate (1:2000, 711-035-152, Jackson ImmunoResearch) were used in a PBS/0.1% Tween-20 buffer containing 0.25% non-fat milk. The immunoreaction was detected using WesternBright Quantum HRP Substrate (Advansta, Menlo Park, CA, USA). The protein bands were scanned directly from the membrane using a camera and were quantified using Gel-Pro Analyzer.

Total RNA was isolated from fibroblast cells using TRIzol reagent (Sigma–Aldrich) and a Direct-zol Kit (Zymo Research, Irvine, CA, USA) according to the manufacturer’s instructions. For isolation of RNA fractions Cytoplasmic and Nuclear RNA Purification kit (Norgen Biotek, Corp., Thorold, ON, Canada) was used according to the manufacturer’s instructions. The RNA concentration was measured using a DeNovix spectrophotometer (Wilmington, DE, USA). An amount of 500 ng of total RNA or 200 ng of fractionated RNA was reverse transcribed at 55°C using Superscript III (Life Technologies) and random hexamer primers (Promega, Madison, WI, USA). cDNA was used for qPCR using LightCycler 480 SYBR Green I Master (Roche, Basel, Switzerland) with denaturation at 95°C for 10 min, followed by 45 cycles of denaturation at 95°C for 10 s, annealing at 60°C for 15 s, and elongation at 72°C for 20 s, with HTT, GAPDH, or U6-specific primers (sequences are listed in Supplementary Table S2) on the Light Cycler 480 II (Roche). Data pre-processing and normalization were performed using LightCycler 480 SW 1.5.1 software.

ImageJ software (Fiji distribution) was used for image analysis. Images were adjusted with respect to brightness, contrast, and smooth effects. For the statistical analysis of images, a Python script for ImageJ was prepared.

All experiments were repeated at least three times. The statistical significance of changes in gene expression levels (real-time PCR, western blotting, and IF level) was assessed using a one-sample t-test, with an arbitrary value of 1 assigned to cells treated with control siRNA (siLUC). Selected data were compared using an unpaired t-test with Welch’s correction to assess the allele-selectivity of silencing (normal vs. mutant allele silencing). The statistical significance of the number of foci, nuclear RNA signal and percent of aggregate-positive cells was assessed using a one-way ANOVA and multiple comparisons testing with post hoc Dunnett’s tests. For FISH and IF analyses, at least 30 cells were analyzed for each experimental condition. p-values of <0.05 were considered significant.

We used a set of chemically synthesized ONs that differed with regard to their (I) targeted sequence (specific HTT sequence or CAG repeats), (II) anticipated mechanism of activity (RNAi-based, RNase H-inducing, and a “blocker”), and (III) pattern of chemical modification (pure RNA ONs and RNAs containing 2′-fluoro (2′F), 2′-O-methylo (2′OMe), and phosphorothioate (PTO) modifications, DNA gapmers with 2′OMe and PTO, as well as LNA oligomer). All ON sequences are presented in Table 1 with references if they were used in previous studies. Specifically, we used siRNA and ASO targeting HTT-specific sequences (siHTT and ASO HTT, respectively) and a set of CAG repeat-targeting ONs: unmodified RNA duplex (CAG/CUG), miRNA-like siRNAs with single mismatches (A2 and G2), chemically modified siRNA (A2F and A2M), antisense oligonucleotide (ASO CTG) and a short LNA blocker (LNA CTG) (Figure 2A).

First, we aimed to determine the cellular distribution of ONs targeting CAG repeat sequences. Selected siRNAs (CAG/CUG, A2, G2, A2F, and A2M), ASO CTG, and LNA CTG were delivered by lipid-based transfection. After 48 h, cells were fixed and small-RNA FISH was performed using probes that interact with the ONs in a 1:1 stoichiometry. However, we were not able to visualize LNA reagents owing to their short length. First, we obtained images of reagents in control fibroblasts (Figure 1A). We observed predominantly cytoplasmic localization of RNA interference reagents, both unmodified and chemically modified. Owing to the differences in binding strengths of the probes to various ONs, we did not quantify differences between reagents. The ONs mainly localized to cytoplasmic vesicles (Figure 1B), similar to the localization observed for control fluorescent siRNA, BlockIT (data not shown). Observed ONs did not localize within endosomes marked with EEA1 and Rab5 proteins showing that these are probably vesicles formed by the lipofection (Supplementary Figures S1A,B). It is worth noting that using this experimental approach, we could observe only cellular spots that contained multiple ON molecules, but not the localization of single molecules that could represent sites of interactions with target sequences. We also observed the passenger strand of CAG/CUG, which showed similar localization to that of the guide strand (Figure 1C). ASO localized within both the cytoplasm and nucleoplasm; however, a strong bias toward the nucleus was observed. Non-transfected cells and cells transfected with control siRNA siLUC were used as negative controls. Next, we examined whether the CAG repeat-expanded tract present in HD cells affects the localization of reagents targeting the CAG sequence. We did not observe significant differences in localization between HD and control fibroblasts, indicating that the presence of the mutation in cells does not alter reagent localization (Figure 1D).

In our earlier studies we defined the effectiveness of ON treatment at the RNA level using RT-PCR (separate analyses for normal and mutant alleles) and at the protein level using western blot (Fiszer et al., 2011, 2013, 2016). To monitor treatment outcomes more broadly, we added in the current study a microscopic analyses of CAG RNA foci and huntingtin levels and localization. We used a CTG probe that was demonstrated in previous studies (Urbanek and Krzyzosiak, 2016) to successfully visualize RNA foci in cellular models of polyQ diseases (Figure 2B). Both analyzed HD lines demonstrated nuclear RNA foci, which were observed in about 50% cells. RNA foci were not detected after RNase treatment (Supplementary Figure S2). As shown for LNA reagents, for which their binding to repeat sequences may block subsequent RT-PCR (Rué et al., 2016), we demonstrated that the reagents did not interfere with the RNA FISH procedure by visualizing HTT mRNA with HTT sequence-specific probes. Probes targeting the first exon of HTT mRNA were used to successfully visualize RNA foci in HD fibroblasts. In control fibroblasts, the signal was rather uniform within the nucleoplasm and cytoplasm, with increased signals strength observed in nucleoli (Figure 2C). An additional feature of HD that may serve as an indicator of an effective therapeutic approach is the presence of HTT protein aggregates. We observed huntingtin aggregates in the cytoplasm and nucleus in both HD fibroblast cell lines, and no such aggregates were visible in the control cell line (Figure 2D). Thus in addition to western blotting, IF may be used to observe protein level changes resulting from ON treatment.

We investigated the silencing of HTT expression at the mRNA and protein levels using selected ONs under the same experimental conditions as used for microscopic analyses, i.e., 48 h after the transfection of HD fibroblasts with 50 nM ONs. We assessed total HTT mRNA levels using qRT-PCR and primers located downstream the CAG repeat tract (Figures 2A, 3A). The most significant decreases in HTT mRNA were observed for siHTT (to ∼40% of the control level) and for both ASOs (to ∼65% of the control level). CAG/CUG reagent decreased level of HTT mRNA to ∼85%. A separate analysis of HTT alleles by semi-quantitative RT-PCR revealed a similar general trend of HTT silencing by these ONs and there was no indication of allele-selective inhibition by ASO reagents (data not shown). At the selected time point, we did not observe decrease in HTT mRNA level by A2, G2, A2F, and A2M ONs. Additionally, for selected set of ONs we performed RNA fractionation to analyze changes in nuclear and cytoplasmic transcript level (Supplementary Figures S3A,B). We tested siHTT, A2, ASO HTT, and ASO CTG in this assay. The results were generally consistent with qRT-PCR for total transcript level described above. siHTT caused lowering of HTT mRNA already in the nucleus (to ∼70%) but more prominent effects were observed in the cytoplasm what in total resulted in HTT mRNA decrease to ∼25%. For A2 we observed slight increase of HTT transcript in the nucleus and slight decrease in the cytoplasm. As expected, the effects of ASOs activity were mainly nuclear as these ONs decreased HTT mRNA level to ∼45% already in the nucleus. For all nine ONs we performed western blotting to analyze normal and mutant huntingtin levels separately (Figure 3B). For ONs that lowered HTT mRNA (siHTT and ASOs), significant decreases in protein levels were also observed, to ∼25%, ∼30%, and ∼60% of the control level for siHTT, ASO HTT and ASO CTG, respectively. Neither siHTT, nor any of the used ASOs acted in an allele-selective way, decreasing both alleles with similar strength. Consistent with previous reports, selected CAG repeat-targeting reagents caused the allele-selective lowering of huntingtin, with a high preference for the mutant protein observed for A2, G2, and A2F. Significant allele-selectivity was also observed for A2M ON, but not for CAG/CUG siRNA or LNA CTG.

Next, we analyzed RNA foci after ONs treatment to investigate whether mRNAs within the nucleus are accessible to ON reagents, whether these ONs exert their activity within the nucleus, and whether they can decrease RNA foci. Our results showed that not all tested ONs are able to significantly decrease the number of RNA foci, despite altering the RNA or protein levels. The results differed for probes used; however, a similar tendency to reduce number of RNA foci was observed (Figures 4A,B). Untreated HD fibroblasts had a mean of 12.4 foci per foci-positive cell, in agreement with our previous reports. Cells treated with siLUC and BlockIT exhibited similar means, with 11.2 and 14.3 foci per cell, respectively. Most significantly RNA foci-disrupting ONs were A2, with a mean of 6.6 foci per cell, ASO CTG, with a mean of 6.1, and LNA CTG, with a mean of 6.8 foci per cell (Figure 4C). Most of the tested ONs significantly decreased the number of foci per cell, but with different efficiencies. siHTT, CAG/CUG, G2, A2F, and A2M reduced foci number to a mean of 7.9, 7.4, 7.3, 7.5, and 8.1 foci per cell, respectively. We did not observe significant foci alterations using ASO HTT. Next, we analyzed nuclear signals from the CTG probe (Figure 4D). All HTT-specific and CAG-specific ONs, except for LNA CTG, decreased level of detected RNA in the nucleus. siHTT and CAG/CUG reagents also triggered visible reductions in overall signal intensity from the cell. Differences between results obtained for measurement of nuclear level of transcript with RNA FISH and qRT-PCR after fractionation (Supplementary Figure S3A), are probably related to different specificity of the methods. Next, we compared results from foci analysis with total RNA level results obtained with qRT-PCR (Figure 3A). There was no clear correlation between the reduction in RNA levels and the RNA foci phenotype (Spearman correlation, p = 0.8603).

To further analyze the therapeutic potential of selected ONs, we performed IF experiments to detect huntingtin protein (both normal and mutant) in cells. The results of the microscopic analysis of HTT protein levels (Figures 5A,B and Supplementary Figure S5) were in agreement with results obtained by western blotting (Figure 3B). However, we could additionally observe differences in protein localization after ON treatment and the presence and number of mutant protein aggregates. Protein aggregates were observed in cells with or without RNA foci showing that there is no connection between formation of RNA foci and protein aggregates (Supplementary Figure S4). We observed that after treatment with ASO CTG, the protein tended to localize around the cytoplasmic vesicles containing ONs (Figure 5C). Moreover, in a substantial portion of cells, the protein localized around the nucleus after ON treatment (Figure 5D). Next, we measured the percentage of aggregate-positive cells (Figure 5E). With the decrease in mutant protein levels observed by western blotting, we also observed decreases in the number of protein aggregates. In untreated and siLUC-treated fibroblasts, we observed 1–3 aggregates in about 60% of cells. Protein aggregates localized mainly in the cytoplasm. Rarely, cells with a high number of cytoplasmic protein aggregates were observed and these cells also had nuclear aggregates (Figure 2C). We observed significant decrease in aggregate-positive cells up to 35–38% after treatment with A2M, G2, and A2F (Figure 5E).