Microarray analysis of gene/transcript expression in Angelman syndrome: deletion versus UPD

Abstract

Angelman syndrome (AS) is a neurodevelopmental disorder due to a functional deficit, usually a deletion, of the UBE3A gene located in the 15q11-q13 chromosome region. We report the first microarray analysis of gene expression in AS using a custom cDNA microarray to compare expression patterns from lymphoblastoid cell lines from control males and AS subjects with a 15q deletion or uniparental paternal disomy 15. Expression patterns of genes known to be biallelically expressed or paternally or maternally expressed were consistent with expectations. We detected paternal or maternal allelic bias in the expression of several genes and transcripts (e.g., GABRA5, GABRB3, WI-14946). Additionally, mechanisms controlling paternal allele expression appear to be faithfully replicated in each paternal chromosome in individuals with paternal disomy. Our results indicate that interconnected mechanisms can produce subtle and unexpected changes in gene expression that may help explain the phenotypic differences observed among the genetic subtypes of AS.

Fig. 1.

Schematic diagram of human chromosomal region 15q11–q13. White boxes represent genes expressed from the paternal chromosome only, black boxes represent genes expressed from the maternal chromosome only, and boxes with diagonal lines represent genes expressed from both chromosomes. The imprinting centers (IC) for AS (black) and PWS (white) are shown. BP1, BP2, and BP3 represent the two proximal breakpoints and one distal breakpoint observed in both AS and PWS associated with the typical 15q11–q13 deletions. Nucleotide numbers represent the proximal and distal nucleotide base pairs for GCP5 and HERC2, respectively, relative to the chromosome 15p terminus and taken from the UCSC Genome Browser (http://genome.ucsc.edu).

Fig. 2.

Replicate hybridization spots from the custom cDNA array. Target sequences from the AS subjects resulting from 15q11–q13 deletions (Del) or uniparental paternal disomy (UPD) were labeled with Alexa 647 (red) and target sequences from controls were labeled with Alexa 555 (green).

Fig. 3.

Comparative expression levels of genes/transcripts. Data represent the normalized means ± the standard error. Each probe was replicated 10 times on each of four arrays, two labeled with Alexa 555 and two labeled with Alexa 647. (A) Control and biallelically expressed genes/transcripts. (B) Paternally expressed genes/transcripts from 15q11–q13. (C) Genes/transcripts with variant expression levels [panel 1, PWS UPD < PWS deletion and control; panel 2, paternal bias; panel 3, PWS deletion > control & PWS UPD; panel 4, maternal bias (panel 1–panel 4 adapted from Bittel et al. [5]; used by permission of the publisher)]. Red lettering indicates sequences that map outside the 15q11–q13 region.

Fig. 4.

Representative quantitative RT-PCR using primers specific for GABRB3. Total RNA was extracted from actively growing lymphoblast cultures and equal quantities were used in a QuantiTect all-in-one RT-PCR using Sybr green fluorescence to quantify the amplicon. The C_T was set at the narrowest portion of the logarithmic phase of amplification (C, control; U, AS UPD; D, AS deletion). The cycle difference between the control (27.7 cycles) and AS UPD (26.9 cycles) was 0.8 cycle, which corresponds to a fold change in expression of 1.7 indicating a higher expression in AS UPD. For the deletion group, the cycle was 28.2, which represents a difference of 0.5 cycle, which corresponds to a −1.4-fold change in expression compared with the control, indicating a lower expression in AS deletion.