Stress-induced epigenetic changes

Evidence suggests that environmental stresses can lead to epigenetic modifications of the genome that can be inherited, although the molecular mechanisms of this process are not known. Now, Shunsuke Ishii and colleagues report that stress-induced phosphorylation of the transcription factor dATF-2 in Drosophila can disrupt heterochromatin, which can be transmitted to the next generation (Cell 145, 1049–1061, 2011). The authors analyzed flies carrying the marker gene white close to centromeric heterochromatin. Loss of Atf-2 led to derepression of white, suggesting Atf-2 is required for the generation and maintenance of heterochromatin. Immunostaining experiments showed that dATF-2 localizes to heterochromatin and that heat shock in early fly embryos disrupts this localization. The authors then mated adult females (that were exposed during embryogenesis) to unstressed males. The authors observed derepression of the white gene in progeny, indicating that the defective heterochromatin was transmitted. The inheritance of the heat-shock effect was tested over multiple generations, and, although continued heat-shock treatment sustained white derepression over multiple generations, it eventually returned to the normal state. The authors suggest that stress-induced epigenetic changes are mediated by regulation of chromatin structure. PC

Self-recognition in social organisms

The capacity to discriminate between self and nonself is a critical trait for most organisms. The social amoebae Dictyostelium discoideum can aggregate and form multicellular fruiting bodies in which 20–30% of cells form a stalk and die, whereas 70–80% of cells survive as spores. The altruistic behavior of stalk cells is not well understood, although D. discoideum shows kin discrimination as a defense against cheater cells that sporulate but do not contribute to the stalk. Gad Shauldsky and colleagues now report that allelic pairs of tgrB1 and tgrC1 are necessary and sufficient for self-recognition in D. discoideum (Science published online, doi:10.1126/science.1203903, 23 June 2011). Cells deficient in tgrB1 and tgrC1 arrest in the aggregate stage and show compromised spore production, suggesting that these genes are required for self-recognition. The authors cloned these genes from four other wild strains, recombined them into the tgrB1−tgrC1− cells and mixed them with the parental strain. Cells carrying the wild replacement genes autosegregated from the parental strain, and each population formed separate fruiting bodies and spores appropriately. These results suggest that a matching pair of alleles at these genes is sufficient to confer self-recognition. The authors propose that TgrB1 and TgrC1 are cell-surface proteins that mediate self-recognition by direct binding. PC

STAT1 and fungal infections

Jos van der Meer and colleagues (N. Engl. J. Med. 365, 54–61, 2011) report that mutations in STAT1 underlie an autosomal dominant form of chronic mucocutaneous candidiasis marked by susceptibility to fungal infections of the skin, nails and mucous membranes. The authors performed immunological studies on peripheral blood mononuclear cells from members of a Dutch family with this disorder and determined that the interleukin-12 and interleukin-23 signaling pathways were impaired in the affected individuals. On the basis of these findings, the authors selected 100 candidate genes for analysis and identified a missense variant in STAT1 (p.Arg274Trp) that segregated with disease in the three affected family members. Further analyses of four additional British and Dutch families identified the same p.Arg274Trp variant in one family and a second missense variant (p.Ala267Val) in the three other families. For each variant, haplotype analyses were consistent with a single mutational origin. Notably, some individuals carrying the p.Arg274Trp variant also developed autoimmune phenotypes, supporting a key role for STAT1 both in the response to fungal infections and in the regulation of immune homeostasis. KV

Open chromatin and hematologic traits

Formaldehyde-assisted isolation of regulatory elements (FAIRE) identifies regions of open chromatin that indicate sites of regulatory activity. Now Dirk Paul and colleagues report FAIRE-generated maps of open chromatin in megakaryocyte and erythroblastoid cell lines at 62 loci associated with hematologic quantitative traits, coronary artery disease and myocardial infarction (PLoS Genet. 7, e1002139, 2011). The authors identified 254 and 251 sites of open chromatin in megakaryocyte and erythroblastoid cells, respectively, with 147 sites in common. They found that the majority of open chromatin sites occur at noncoding regions, frequently near transcription start sites. The authors focused on an open chromatin site at 7q22.3 that coincides with an allele associated with increased mean platelet volume (MPV) and decreased platelet count. They show that there is a megakaryocyte-specific nucleosome-depleted region at this locus, and the MPV-associated allele has lower binding affinity for the EVI1 transcription factor and associates with higher levels of expression of the nearby gene PIK3CG in platelets and macrophages. This work leads to hypotheses that EVI1 is a negative regulator of PIK3CG expression and that increased PIK3CG expression leads to increased MPV, and it suggests a possible functional mechanism for this MPV-associated locus. EN

Rare variants across populations

Carlos Bustamante and colleagues estimate the allele frequency spectrum of genetic variants across continental human populations on the basis of an analysis of 1000 Genomes Project data sets (Proc. Natl. Acad. Sci. USA published online, doi:10.1073/pnas.1019276108, July 5, 2011). They directly compared low- and high-coverage sequencing data sets from the 1000 Genomes pilot project and developed an error model used to estimate and correct for biases in using only the low-coverage data set. Then they modeled the joint allele site frequency spectrum between pairs of populations in the low-coverage data set and applied this to estimate demographic parameters for an out-of-Africa demographic model. In turn, they used this inferred demographic model to predict the number of variants that can be discovered by increasing the sample sizes of these data sets. For comparison, they also used a jack-knife approach to estimate the number of segregating sites in a sample. They found that most genetic variation is rare, most of the rare variants show limited sharing across continental populations and rare variants show less sharing than common variants between closely related populations. Such limited sharing of rare variation between continental populations represents additional challenges for replication of rare variant association studies. OB

Written by Orli Bahcall, Pamela Colosimo, Emily Niemitz & Kyle Vogan