Imprinted expression of UBE3A in non-neuronal cells from a Prader-Willi syndrome patient with an atypical deletion

Abstract

Prader-Willi syndrome (PWS) and Angelman syndrome (AS) are two neurodevelopmental disorders most often caused by deletions of the same region of paternally inherited and maternally inherited human chromosome 15q, respectively. AS is a single gene disorder, caused by the loss of function of the ubiquitin ligase E3A (UBE3A) gene, while PWS is still considered a contiguous gene disorder. Rare individuals with PWS who carry atypical microdeletions on chromosome 15q have narrowed the critical region for this disorder to a 108 kb region that includes the SNORD116 snoRNA cluster and the Imprinted in Prader-Willi (IPW) non-coding RNA. Here we report the derivation of induced pluripotent stem cells (iPSCs) from a PWS patient with an atypical microdeletion that spans the PWS critical region. We show that these iPSCs express brain-specific portions of the transcripts driven by the PWS imprinting center, including the UBE3A antisense transcript (UBE3A-ATS). Furthermore, UBE3A expression is imprinted in most of these iPSCs. These data suggest that UBE3A imprinting in neurons only requires UBE3A-ATS expression, and no other neuron-specific factors. These data also suggest that a boundary element lying within the PWS critical region prevents UBE3A-ATS expression in non-neural tissues.

Figure 1.

Map of chromosome 15q11–q13 region. Map of chromosome 15q11–q13 between common break points 2 and 3 (BP2 and BP3). Blue dotted lines represent the regions deleted on the paternal allele and red dotted line represents the regions deleted on the maternal allele for the indicated cell lines. Blue and red boxes denote genes expressed exclusively from paternal and maternal alleles, respectively. Gray boxes denote genes expressed biallelically. Differential methylated regions (DMRs) are shown using circles where open and closed circles represent unmethylated and methylated alleles, respectively. Arrows indicate the direction of transcription. A solid blue line represents paternal SNURF/SNRPN transcripts expressed in most cell types, whereas a dashed blue line indicates neuron-specific transcripts, including UBE3A-ATS.

Figure 2.

Characterization of PWS SD iPSC lines. (A) Immunocytochemistry for the pluripotency markers NANOG, SSEA-4, TRA1-80 and TRA1-81. Cell nuclei were counterstained with DAPI. Scale bar = 100 μm. (B) Expression of genes within 15q11–q13 region iPSCs to confirm deletion in PWS SD iPSCs. GAPDH was used as en endogenous control, and data were normalized to NML 1-0 iPSCs. (C) Methylation analysis of PWS-IC within SNPRN, using a methylation-sensitive restriction endonuclease quantitative PCR assay, confirming the maintenance of genomic imprinting at the PWS-IC following reprogramming in PWS SD iPSCs. PWS del 1-7 and AS del 1-0 iPSCs were used as controls. Percent methylation was reported plus or minus the SD of three replicates.

Figure 3.

Imprinting of paternal UBE3A in non-neuronal cells. (A) Gene expression analysis of SNORD115 snoRNAs in iPSCs by qRT–PCR. GAPDH was used as an endogenous control, and data were normalized to NML 1-0 10 week neurons. (B) Analysis of UBE3A-ATS expression in PWS SD iPSCs. RT-19 primers were used to analyze the expression of the UBE3A-ATS (13), and GAPDH was used as a control. NML 1-0 10 week neurons were used as a positive control, since the UBE3A-ATS is exclusively expressed in neurons. (C) Expression analysis of UBE3A in iPSCs. GAPDH was used as an endogenous control, and data were normalized to NML 1-0 iPSCs. PWS del 1-7 and AS del 1-0 iPSCs were used as controls. (D) Allele-specific RT–PCR showed equal expression of UBE3A in NML 1-0 iPSCs. In PWS SD iPSCs, the paternal UBE3A is repressed while UBE3A-ATS is expressed (S = sense, ATS = antisense and GM = genomic DNA). (E) RNA FISH using a riboprobe that specifically detects UBE3A sense transcripts (green) and a BAC probe that detects SNORD115 transcripts (red). UBE3A is actively transcribed from both alleles, and SNORD115 transcripts are not detected in NML 1-0 iPSCs. Scale bars = 5 μm. In PWS SD iPSCs, 83% of SNORD115-positive nuclei showed monoallelic expression of UBE3A and 17% of that showed biallelic expression of UBE3A, as indicated by the asterisk.

Figure 4.

SNRPN upstream exon usage in PWS SD iPSCs neurons. (A) Map showing the organization of upstream SNURF-SNRPN exons (34). Upstream exons U1B, U1Bʹ, U1A and U2 are largely neuron specific, as indicated by the bracket. Exon U4 is expressed at low levels in a variety of tissues, including iPSCs. It is unclear whether U3 and U5 are brain specific. The location of the PWS-IC is indicated by a half-filled circle. (B) Immunocytochemistry for the neural marker MAP2 in PWS SD 2-8 10 week neurons. Cell nuclei were counterstained with DAPI. Scale bar = 100 μm. (C) Expression analysis of upstream SNRPN exons. GAPDH was used as an endogenous control, and data were normalized to NML 1-0 10 week neurons, since these exons are predominately expressed in neurons.

Figure 5.

Topotecan is less effective in repressing the UBE3A-ATS in PWS SD neurons. Relative expression of SNORD115 by qRT–PCR in 10 week neurons derived from PWS SD and AS iPSC lines treated with various concentrations of Topotecan. GAPDH was used as an endogenous control. **P-value ≤ 0.05.