Frequent monoallelic or skewed expression for developmental genes in CNS-derived cells and evidence for balancing selection.

Cellular mosaicism due to monoallelic autosomal expression (MAE), with cell selection during development, is becoming increasingly recognized as prevalent in mammals, leading to interest in understanding its extent and mechanism(s). We report here use of clonal cell lines derived from the CNS of adult female [Formula: see text] hybrid (C57BL/6 X JF1) mice to characterize MAE as neural stem cells (nscs) differentiate to astrocyte-like cells (asls). We found that different subsets of genes show MAE in the two populations of cells; in each case, there is strong enrichment for genes specific to the respective developmental state. Genes that exhibit MAE are 22% of nsc-specific genes and 26% of asl-specific genes. Moreover, the promoters of genes with MAE have reduced CpG dinucleotides but increased CpG differences between the two parental mouse strains. Extending the study of variability to wild populations of mice, we found evidence for balancing selection as a contributing force in evolution of those genes showing developmental specificity (i.e., expressed in either nsc or asl), not just for genes showing MAE. Furthermore, we found that genes showing skewed allelic expression (SKE) were similarly enriched among cell type-specific genes and also showed a heightened probability of balancing selection. Thus, developmental stage-specific genes and genes with MAE or SKE seem to make up overlapping classes subject to selection for increased diversity. The implications of these results for development and evolution are discussed in the context of a model with stochastic epigenetic modifications taking place only during a relatively brief developmental window.

Transcriptome-wide survey of mouse CNS-derived cells reveals monoallelic expression within novel gene families.

Monoallelic expression is an integral component of regulation of a number of essential genes and gene families. To probe for allele-specific expression in cells of CNS origin, we used next-generation sequencing (RNA-seq) to analyze four clonal neural stem cell (NSC) lines derived from Mus musculus C57BL/6 (B6)×Mus musculus molossinus (JF1) adult female mice. We established a JF1 cSNP library, then ascertained transcriptome-wide expression from B6 vs. JF1 alleles in the NSC lines. Validating the assay, we found that 262 of 268 X-linked genes evaluable in at least one cell line showed monoallelic expression (at least 85% expression of the predominant allele, p-value<0.05). For autosomal genes 170 of 7,198 genes (2.4% of the total) showed monoallelic expression in at least 2 evaluable cell lines. The group included eight known imprinted genes with the expected pattern of allele-specific expression. Among the other autosomal genes with monoallelic expression were five members of the glutathione transferase gene superfamily, which processes xenobiotic compounds as well as carcinogens and cancer therapeutic agents. Monoallelic expression within this superfamily thus may play a functional role in the response to diverse and potentially lethal exogenous factors, as is the case for the immunoglobulin and olfactory receptor superfamilies. Other genes and gene families showing monoallelic expression include the annexin gene family and the Thy1 gene, both linked to inflammation and cancer, as well as genes linked to alcohol dependence (Gabrg1) and epilepsy (Kcnma1). The annotated set of genes will provide a resource for investigation of mechanisms underlying certain cases of these and other major disorders.

Dual DNA methylation patterns in the CNS reveal developmentally poised chromatin and monoallelic expression of critical genes.

As a first step towards discovery of genes expressed from only one allele in the CNS, we used a tiling array assay for DNA sequences that are both methylated and unmethylated (the MAUD assay). We analyzed regulatory regions of the entire mouse brain transcriptome, and found that approximately 10% of the genes assayed showed dual DNA methylation patterns. They include a large subset of genes that display marks of both active and silent, i.e., poised, chromatin during development, consistent with a link between differential DNA methylation and lineage-specific differentiation within the CNS. Sixty-five of the MAUD hits and 57 other genes whose function is of relevance to CNS development and/or disorders were tested for allele-specific expression in F(1) hybrid clonal neural stem cell (NSC) lines. Eight MAUD hits and one additional gene showed such expression. They include Lgi1, which causes a subtype of inherited epilepsy that displays autosomal dominance with incomplete penetrance; Gfra2, a receptor for glial cell line-derived neurotrophic factor GDNF that has been linked to kindling epilepsy; Unc5a, a netrin-1 receptor important in neurodevelopment; and Cspg4, a membrane chondroitin sulfate proteoglycan associated with malignant melanoma and astrocytoma in human. Three of the genes, Camk2a, Kcnc4, and Unc5a, show preferential expression of the same allele in all clonal NSC lines tested. The other six genes show a stochastic pattern of monoallelic expression in some NSC lines and bi-allelic expression in others. These results support the estimate that 1-2% of genes expressed in the CNS may be subject to allelic exclusion, and demonstrate that the group includes genes implicated in major disorders of the CNS as well as neurodevelopment.

Monoallelic expression of multiple genes in the CNS.

The inheritance pattern of a number of major genetic disorders suggests the possible involvement of genes that are expressed from one allele and silent on the other, but such genes are difficult to detect. Since DNA methylation in regulatory regions is often a mark of gene silencing, we modified existing microarray-based assays to detect both methylated and unmethylated DNA sequences in the same sample, a variation we term the MAUD assay. We probed a 65 Mb region of mouse Chr 7 for gene-associated sequences that show two distinct DNA methylation patterns in the mouse CNS. Selected genes were then tested for allele-specific expression in clonal neural stem cell lines derived from reciprocal F(1) (C57BL/6xJF1) hybrid mice. In addition, using a separate approach, we directly analyzed allele-specific expression of a group of genes interspersed within clusters of OlfR genes, since the latter are subject to allelic exclusion. Altogether, of the 500 known genes in the chromosomal region surveyed, five show monoallelic expression, four identified by the MAUD assay (Agc1, p (pink-eyed dilution), P4ha3 and Thrsp), and one by its proximity to OlfR genes (Trim12). Thrsp (thyroid hormone responsive SPOT14 homolog) is expressed in hippocampus, but the human protein homolog, S14, has also been implicated in aggressive breast cancer. Monoallelic expression of the five genes is not coordinated at a chromosome-wide level, but rather regulated at individual loci. Taken together, our results suggest that at least 1% of previously untested genes are subject to allelic exclusion, and demonstrate a dual approach to expedite their identification.

Unique retrotransposon LINE-1 distribution at the Prader-Willi Angelman syndrome locus.

We analyzed the distribution of long interspersed nuclear elements (LINE)-1 (L1) along mouse autosomes at a 1-Mb scale, and found a unique combination of high density and strand asymmetry of L1 elements at the imprinted Prader-Willi syndrome/Angelman syndrome (PWS/AS) locus on mouse chromosome 7. This L1 signature overlaps the paternally expressed domain of the locus, excluding the maternally expressed Ube3a gene, and is conserved in rat and human. Unlike the PWS/AS locus, other instances of high L1 density and strand asymmetry in the mouse are not associated with imprinted regions and are not evolutionarily conserved in human. The evolutionary conservation of the L1 signature at the PWS/AS locus despite differences in composition of L1 elements between rodent and human, requires a mechanism for active perpetuation of L1 asymmetry during bursts of L1 activity, and indicates a possible functional role for L1 elements at this locus. Aside from the PWS/AS locus, rodents have a far greater correlation of L1 densities between DNA strands than do humans; we provide evidence that this difference in interstrand correlation between the two taxa is due largely to the difference in average age of the dominant L1 families.

Novel paternally expressed intergenic transcripts at the mouse Prader-Willi/Angelman Syndrome locus.

Gene expression profiling was performed on central nervous system (CNS) tissue from neonatal mice carrying the T9H translocation and maternal or paternal duplication of proximal Chromosomes 7 and 15. Our analysis revealed the presence of two novel paternally expressed intergenic transcripts at the Prader-Willi/Angelman Syndrome (PW/AS) locus. The transcripts were termed Pec2 and Pec3, for paternally expressed in the CNS. Imprinting of these transcripts was confirmed by sequencing of RT-PCR products in F(1) hybrids between Mus musculus musculus C57BL/6 and Mus musculus castaneus, following identification of single nucleotide polymorphisms between the two strains. Imprinting of Pec2 was also confirmed by Northern blot analysis. The two transcripts are separated by 0.5 Mb and are transcribed in the same orientation. They are located in a long interspersed transposable element (LINE)-rich region midway between the PW/AS imprinting center and the paternally expressed genes Ndn, Magel2, and Mkrn3, which are under imprinting center control. Our analysis also revealed imprinting of Magel2, Mkrn3, Ndn, Ube3a, and Usp29, as well as Pec2 and Pec3, in embryonic brain 15.5 dpc, and provided a survey of biallelically expressed genes on proximal Chrs 7 and 15 in embryonic and neonatal CNS.

Analysis of imprinting in mice with uniparental duplication of proximal chromosomes 7 and 15 by use of a custom oligonucleotide microarray.

We have developed an imprinting assay combining the use of mice carrying maternal or paternal duplication of chromosomal regions of interest with custom oligonucleotide microarrays. As a model system, we analyzed RNA from CNS tissue of neonatal mice carrying the reciprocal translocation T(7;15)9H and uniparental duplication of proximal Chr 7 and 15. The duplicated region includes the locus on proximal Chr 7 corresponding to the human Prader-Willi/Angelman Syndrome. The microarray contained 322 oligonucleotides, including probes to detect major genes involved in neural excitability and synaptic transmission, as well as known imprinted genes mapping to proximal Chr 7: Ndn, Snrpn, Mkrn3, Magel2, Peg3, and Ube3a. Imprinting of these genes in neonatal cortex and cerebellum was first confirmed by quantitative RT-PCR. Their inclusion on the microarray thus provided positive controls for evaluating the effect of background on the sensitivity of the assay, and for establishing the minimum level of expression required to detect imprinting. Our analysis extended previous work by revealing bi-allelic expression in CNS tissue of those queried genes mapping to proximal Chr 7 or 15, including the Gabrb3 gene, for which there have been conflicting reports. Microarray analysis also revealed no effect of the maternal or paternal disomy on expression levels of the unlinked genes detected, including those potentially implicated in the Prader-Willi or Angelman Syndrome. In addition, quantitative RT-PCR revealed a gene dosage effect in both cerebellum and cortex for all of the known imprinted genes assayed, except for Ube3a in cerebellum.

The Cdkn1a gene (p21Waf1/Cip1) is an inflammatory response gene in the mouse central nervous system.

We used high-density cDNA microarray analysis to examine changes in the gene expression profile of the hippocampus of C57BL/6 mice following intraperitoneal injection of lipopolysaccharide (LPS). Three hours after injection, the greatest increase in RNA expression was found for an expressed sequence tag subsequently identified as the Cdkn1a gene, coding for the cyclin-dependent kinase inhibitor p21(Waf1/Cip1). Northern blot hybridization confirmed the induction of Cdkn1a mRNA in the central nervous system (CNS), and also revealed similar increases in kidney, liver and heart. Induction of Cdkn1a expression was transient, reaching maximal levels in the CNS 3-6 h after LPS administration, and returning to untreated levels by 24 h. Combined use of laser capture microdissection and quantitative reverse transcription-polymerase chain reaction showed that there was a similar change in Cdkn1a expression for the pyramidal cell layer as for total hippocampus.