|
|
Widespread Monoallelic Expression of Human Autosomal Genes« Back to Volume 24, Issue 1, May 2008 - Table of Contents With certain exceptions, it is generally assumed that maternally and paternally-derived copies (alleles) of each gene are expressed at comparable levels in humans. The first exception is inactivation of most of the genes residing on the X-chromosome in females—so called X-inactivation. Half of the cells in an embryo on average randomly inactivate the paternal X chromosome and half inactivate the maternal X chromosome around the time of implantation. The second exception involves imprinting of autosomal genes, such as IGF-2, on a parent-of-origin basis. A third exception is a small group of autosomal genes that are subject to random monoallelic expression; these include genes encoding odorant receptors, T cell receptors, interleukins, and natural killer cell receptors. There is new evidence that monoallelic expression of autosomal genes may be much more extensive than previously believed. Gimelbrant et al exploited the growing number of single nucleotide polymorphisms (SNPs) and advances in gene chip (array) technology to survey allele-specific transcription of about 4,000 genes in lymphoblastoid cell lines from 3 individuals. They took advantage of the observation that once a cell decides to express one of 2 alleles, the clonal descendents of this cell continue to express the selected allele. Since lymphoblastoid cells are polyclonal, they were able to derive clonal B cell lines using single-cell cloning. To perform the genome-wide screen for monoallelic transcription, the investigators developed protocols to distinguish polymorphic allele expression based on detection of SNPs in nuclear RNA, which is enriched in intronic RNA, where most SNPs associated with genes reside. Conversion of this RNA to double-stranded cDNA and analysis on a SNP array generated “transcriptosome-derived genotypes” that allowed monoallelic expression to be identified. Filters were used to minimize cDNA genotyping artifacts. About 10% of SNPs were reliably called from this analysis, which was expected since most of the other SNPs are likely present in regions of the genome that are not expressed by B cell lines that were studied. As proof-of-concept, the investigators first showed that random inactivation of X-chromosome genes could be detected in the clonal cell lines and then demonstrated as an example of their approach that monoallelic expression of the amyloid precursor protein gene could be detected. They next turned to genome-wide screening. On the array used for analysis, there were SNPs present for ~11,000 genes. They were able to detect allele-specific transcription for ~4,000 genes in 2 or more cell clones. Of the ~4,000 genes examined, 2.2% were detected as monoallelically expressed with multiple informative SNPs per gene per clone. An additional 7.3% of assessed genes were identified as monoallelically expressed based on a single informative SNP per gene per clone. The genes included both B cell-specific genes and ubiquitously expressed genes. The investigators suggested a conservative estimate that over 1,000 genes are subject to random monoallelic expression in humans. Several interesting observations were made. For example, the choice of expressed allele was made independently for each gene within a given clonal cell line. This is in contrast to the chromosomal-wide coordination characteristic of X-inactivation. Another finding was that a disproportionately large fraction of genes coding for cell surface proteins—transmembrane receptors and surface proteins was detected. The authors concluded by suggesting that at least 1,000 human genes display random monoallelic transcription that could contribute to genetic diversity within tissues of an individual as well as between individuals. A commentary by Ohlsson1 notes that although monoallelic expression has been known in humans, this study by Gimelbrant expands the concept further especially by documenting it in a much larger number of genes than previously appreciated. He briefly discusses possible mechanisms that could account for the phenomenon as well as its potential role in modulating disease. Editor’s CommentThis is one of several publications in recent years that challenges what we were taught about mendelian genetics. Of note, several genes relevant to human growth disorders were identified as displaying monoallelic expression including the growth hormone receptor gene (GHR) and genes that harbor mutations responsible for Ellis van Creveld syndrome (EVC) and the trichorhinophalengeal syndrome 1 (TRPS1). It seems quite plausible that monoallelic expression of these genes could contribute to the clinical variability of these conditions. Lymphoblastoid cells have very different functions compared to chondrocytes, osteoblasts and other cells that contribute to skeletal growth; and their patterns of gene expression may differ dramatically. Screening the latter cells for monoallelic transcription would be technically much more difficult than for lymphoblastoid cells, but it would likely reveal monoallelic expression of additional growth related genes. William A. Horton, MD References - (linked to
|
Volume 24 Number 1
May 2008
www.GGHjournal.com
ISSN 1932-9032
)