Bidirectional Transcription at the PPP2R2B Gene Locus in Spinocerebellar Ataxia Type 12

Abstract

Background: Spinocerebellar ataxia type 12 (SCA12) is a neurodegenerative disease caused by expansion of a CAG repeat in the PPP2R2B gene.

Objective: In this study, we tested the hypothesis that the PPP2R2B antisense (PPP2R2B-AS1) transcript containing a CUG repeat is expressed and contributes to SCA12 pathogenesis.

Methods: Expression of PPP2R2B-AS1 transcript was detected in SCA12 human induced pluripotent stem cells (iPSCs), iPSC-derived NGN2 neurons, and SCA12 knock-in mouse brains using strand-specific reverse transcription polymerase chain reaction. The tendency of expanded PPP2R2B-AS1 (expPPP2R2B-AS1) RNA to form foci, a marker of toxic processes involving mutant RNAs, was examined in SCA12 cell models by fluorescence in situ hybridization. The apoptotic effect of expPPP2R2B-AS1 transcripts on SK-N-MC neuroblastoma cells was evaluated by caspase 3/7 activity. Western blot was used to examine the expression of repeat associated non-ATG-initiated translation of expPPP2R2B-AS1 transcript in SK-N-MC cells.

Results: The repeat region in the PPP2R2B gene locus is bidirectionally transcribed in SCA12 iPSCs, iPSC-derived NGN2 neurons, and SCA12 mouse brains. Transfected expPPP2R2B-AS1 transcripts induce apoptosis in SK-N-MC cells, and the apoptotic effect may be mediated, at least in part, by the RNA secondary structure. The expPPP2R2B-AS1 transcripts form CUG RNA foci in SK-N-MC cells. expPPP2R2B-AS1 transcript is translated in the alanine open reading frame (ORF) via repeat-associated non-ATG translation, which is diminished by single-nucleotide interruptions within the CUG repeat and MBNL1 overexpression.

Conclusions: These findings suggest that PPP2R2B-AS1 contributes to SCA12 pathogenesis and may therefore provide a novel therapeutic target for the disease. © 2023 International Parkinson and Movement Disorder Society.

Keywords: SCA12; antisense.

Fig. 1.

Bidirectional transcription at PPP2R2B gene locus. (A) PPP2R2B-AS1 map. Primer locations for SS-RT-PCR (Linkered F1 and R1) in B-D are shown. On the sense strand, 7B7D and 7C7D are the two splice variants of PPP2R2B exon 7 that contains the CAG repeat. On the antisense strand, location of SP1 binding sites were predicted using PROMO. 3’ end of PPP2R2B-AS1 was identified by 3’ RACE using gene-specific primer R2 and Generacer 3’ primer. The 5’ end of PPP2R2B-AS1 exon 1 is not characterized. (B-D) Detection of normal and expanded PPP2R2B and PPP2R2B-AS1 by SS-RT-PCR in human iPSCs (B), iPSC derived cortical excitatory neurons differentiated by NGN2 overexpression (C), and cortices (CX) and cerebella (CB) of a humanized SCA12 knock-in mouse model with either 10 or 80 repeats (KI-10 or KI-80). PCR bands amplified from two normal alleles in human control iPSCs or iPSC derived neurons could not be separated on an agarose gel due to size similarity. Only homozygous (homo) mice were included. N=3 independent experiments; representative gel images are shown. PCRs using water (no template) as negative controls. +RT indicates RT-PCR products, while no reverse transcription (−RT) control is a mock reverse transcription containing all the RT-PCR reagents, except the reverse transcriptase.

Fig. 2.

Overexpression of an expanded PPP2R2B-AS1 transcript is toxic to SK-N-MC cells. (A) Caspase 3/7 assay indicates toxicity of an expanded PPP2R2B-AS1-(CUG)73 transcript compared to a normal PPP2R2B-AS1-(CUG)10 transcript. ATXN2-AS constructs are included as positive controls. N=5 biological replicates. The Caspase 3/7 activity in ATXN2-AS-(CTG)22 or PPP2R2B-AS1-(CTG)10 transfected SK-N-MC cells was normalized to 100. (B) Single nucleotide interruptions in the CUG repeat decrease the toxicity of expanded PPP2R2B-AS1 transcript. N=4 biological replicates. The Caspase 3/7 activity in PPP2R2B-AS1-(CTG)10 transfected SK-N-MC cells was normalized to 100. (C) Relative expression level of exogenous PPP2R2B-AS1 transcripts in SK-N-MC cells normalized by GADPH level. qPCR primers used to measure exogenous PPP2R2B-AS1 are located on the vector backbone. N=3 biological replicates. The PPP2R2B-AS1/GAPDH in PPP2R2B-AS1-(CTG)10 transfected SK-N-MC cells was normalized to 1. Mean ± SEM are shown. One-way ANOVA and Tukey posthoc test, *p<0.05, **p<0.01, ***p<0.001.

Fig. 3.

Expanded PPP2R2B-AS1 transcripts form CUG RNA foci. (A-C) In situ hybridization using a CAG DNA/LNA probe indicates PPP2R2B-AS1-(CUG)73 and –int(CUG)73 transcripts form nuclear CUG RNA foci when transfected into SK-N-MC neuroblastoma cells. % of cells with foci and number of foci per cell were quantified (M). Mean ± SEM are shown. N=3 biological replicates, One-way ANOVA, *p<0.05, **p<0.01, compared with PPP2R2B-AS1-(CTG)10, ^^p<0.01, compared with PPP2R2B-AS1-(CTG)73. (D-L) GFP-MBNL1 shows nuclear diffused localization in cells that overexpress PPP2R2B-AS1-(CUG)10 (D-F), but becomes sequestered to CUG RNA foci formed by PPP2R2B-AS1-(CUG)73 (G-I) or -Int(CUG)73 (J-L) transcripts. Scale bar = 10um.

Fig. 4.

The expanded CUG repeat in PPP2R2B-AS1 transcript may be translated into polyalanine (polyAla) containing proteins via RAN translation. (A) Schematic presentation of the tagged PPP2RAB-AS-(CTG)n-3T construct to determine the presence of RAN product. (B-I) Anti-HA staining revealed polyAla RAN protein in PPP2R2B-AS1-(CTG)73–3T and PPP2R2B-AS1-int(CTG)73–3T overexpressing SK-N-MC cells (C and D), compared with the PPP2R2B-AS1-(CTG)10–3T control cells (B), while anti-Flag or anti-Myc staining showed negative results (F-4 and J-4). SK-N-MC cells transfected with RVG-Lamp2b-HA (E), PPP2R2B-Flag (I), and 7B7D-polySer-Myc (M), were used as positive controls for immunostaining using anti-HA, Flag, and Myc tag antibodies, respectively. (N-O) RAN-polyAla-HA protein expression is decreased in PPP2R2B-AS1-int(CTG)73–3T overexpressing cells, compared with PPP2R2B-AS1-(CTG)73–3T. Asterisks mark nonspecific bands. N=4 biological replicates. The RAN-polyAla-HA/β-actin ratio in PPP2R2B-AS1-(CTG)73–3T transfected SK-N-MC cells was normalized to 1. Mean ± SEM are shown. (P-Q) GFP-MBNL1 overexpression decreased RAN-polyAla-HA expression in PPP2R2B-AS1-(CTG)73 expressing SK-N-MC cells. N=3 biological replicates. The RAN-polyAla-HA/β-actin ratio in SK-N-MC cells co-transfected with PPP2R2B-AS1-(CTG)73–3T and GFP was normalized to 1. Mean ± SEM are shown. One-way ANOVA and Tukey posthoc test, *p<0.05, **p<0.01, ***p<0.001.

Fig. 5.

Effect of MBNL1 ZNFs on the production of RAN polyAla proteins from the expPPP2R2B-AS1 transcript. (A). Schematic presentation of the GFP-MBNL1, GFP-MBNL1-Δ251–388, and GFP-MBNL1-Δ1–240 constructs. (B-M). In situ hybridization using a CAG DNA/LNA probe indicates exogenous GFP-MBNL1 and GFP-MBNL1-Δ251–388, but not GFP-MBNL1-Δ1–240, colocalizes with nuclear CUG RNA foci formed by the PPP2R2R-AS1-(CUG)73 transcripts in SK-N-MC neuroblastoma cells. N=3 biological replicates. Representative images are shown. Scale bar = 10um. Arrows point to CUG RNA foci. (N-O) Overexpression of GFP-MBNL1 or GFP-MBNL1-Δ251–388, but not that of GFP-MBNL1-Δ1–240, decreased RAN-polyAla-HA protein expression in PPP2R2B-AS1-(CTG)73–3T expressing SK-N-MC cells. N=4 biological replicates. The RAN-polyAla-HA/β-actin ratio in SK-N-MC cells co-transfected with PPP2R2B-AS1-(CTG)73–3T and GFP was normalized to 1. Mean ± SEM are shown. One-way ANOVA and Tukey posthoc test, *p<0.05, **p<0.01, ***p<0.001. (P-Q) Overexpression of GFP-MBNL1, or GFP-MBNL1-Δ251–388, but not that of GFP-MBNL1-Δ1–240, decreased polyAla-Myc expression in JPH3-Ex2A-(CTG)55-Myc expressing SK-N-MC cells. N=4 biological replicates, Mean ± SEM are shown. The JPH3-Ex2A-polyAla-Myc/β-actin ratio in SK-N-MC cells co-transfected with JPH3-Ex2A-(CTG)55-Myc and GFP was normalized to 1. One-way ANOVA and Tukey posthoc test, *p<0.05, **p<0.01, ***p<0.001.