Collaborative self-incompatibility

Most flowering plants contain male and female gametophytes but do not self-fertilize and instead exclusively mate with other individuals. One adaptation that prevents self-fertilization is self-incompatibility, the genetic recognition and rejection of self-pollen. In many plants, self-incompatibility is controlled by male- and female-specificity determinants encoded by multiallelic S genes. When the S haplotypes of the male and female gametophytes match, S-RNases (female-specific determinants) have a cytotoxic effect inside the self-pollen tube. The S-locus F-box (SLF) protein has been identified as a male-specific determinant and has been hypothesized to recognize non-self S-RNAses and mediate their degradation. Seiji Takayama and colleagues now report (Science 330, 796–799, 2010) that there are at least three types of SLF proteins in Petunia. The authors first cloned nine additional SLF alleles from pollen complementary DNA (cDNA), constructed transgenes and tested the effect of each of these transgenes on self-incompatibility. These in vivo pollination experiments identified three classes of SLF proteins, each affecting different subsets of non-self S haplotypes. Co-immunoprecipitation experiments showed that each type of SLF protein interacts with a specific repertoire of S-RNAses. The authors propose a 'collaborative non-self recognition' mechanism for self-incompatibility in Petunia, in which each SLF type recognizes a subset of non-self S-RNAses, and all SLF types are needed to recognize the entire repertoire of S-RNAses. PC

Uveal melanoma metastasis suppressor

Uveal melanoma is a cancer of the eye that frequently progresses to fatal metastasis. Anne Bowcock and colleagues now report (Science published online, doi:10.1126/science.1194472, 4 November 2010) that class 2 uveal melanomas, which are at high risk for metastasis, frequently harbor somatic mutations in the gene encoding the BRCA1-associated protein BAP1. The authors performed exome sequencing on two class 2 uveal melanomas and discovered that both tumors harbored somatic, inactivating mutations in BAP1. They then sequenced BAP1 in 29 additional class 2 uveal melanomas and found that BAP1 was mutated in the majority of these tumors, with concomitant loss of the wildtype BAP1 allele. Conversely, they identified only a single BAP1 mutation among 26 class 1 uveal melanomas, indicating that BAP1 mutations are largely specific for the more aggressive class 2 subtype. The authors also knocked down BAP1 expression in a uveal melanoma cell line using RNA interference (RNAi) and found that the treated cells developed morphological and gene expression characteristics typical of class 2 tumors. These findings strongly suggest that inactivation of BAP1 is a key step in the progression of uveal melanomas toward metastasis. KV

ATRX and tandem repeats

Mutation of ATRX causes mental retardation and α-thalassaemia with down regulation of α-globin expression (ATR-X syndrome). ATRX binds chromatin at telomeric and heterochromatic repeats, but its role in regulating gene expression is not well understood. Now, Richard Gibbons and colleagues report genome-wide maps of ATRX binding sites (Cell 143, 367–378, 2010). The authors used ChIP-seq in primary human erythroid cells and mouse embryonic stem cells and showed that CpG islands and G-rich tandem repeats are the major euchromatic targets of ATRX. ATR-X syndrome is characterized by variability in α-globin expression between individuals with the same ATRX mutation, which manifests as varying proportions of red blood cells with hemoglobin-H inclusions and variable degrees of thalassaemia. Gibbons and colleagues identified a strong correlation between the length of a tandem repeat at the α-globin locus and the proportion of red blood cells with hemoglobin-H inclusions in 43 individuals with ATR-X syndrome. The authors also showed that the length of a tandem repeat had a cis effect on expression of another ATRX target, NME4. These findings suggest that the phenotypic variability of ATR-X syndrome may be influenced by differences in tandem repeat lengths at ATRX target genes. EN

Apical junction regulator

A defining characteristic of most epithelia is apical-basal cell polarity that partitions the 'top' and 'bottom' of cell layers. Cell-cell junctions located on the apical side of cells are important structures in the establishment of apical-basal polarity. However, the transcriptional networks that regulate the proteins that form cell-cell junctions are not well understood. Now, Kai M. Schmidt-Ott and colleagues report that the transcription factor Grainyhead-like 2 (Grhl2) is required for the expression of apical junctional components in a few types of epithelial tissues in mouse (Development 137, 3835–3845, 2010). Comparison of gene expression profiles in epithelial and non-epithelial cells identified five transcription factors, including Grhl2, whose expression was perfectly correlated with E-cadherin (an apical junction component) expression. The authors generated Grhl2 knockout mice and found that expression of E-cadherin and Claudin 4 (Cldn4), another apical junctional component, was severely reduced in Grhl2 mutants. Subsequent chromatin immunoprecipitation experiments showed that Grhl2 binds to cis-regulatory regions that drive Cdh1 (the E-cadherin gene) and Cldn4 expression. Reporter constructs with the Grhl2 binding sites deleted and mutated showed that Grhl2 is required to activate the Cldn4 promoter. The study suggests that Grhl2 directly regulates two key components of the apical junction complex. PC

Host control of HIV-1

Paul de Bakker and colleagues report a genome-wide association study for host control of HIV-1 (Science published online, doi:10.1126/science.1195271, 4 November 2010). They use a unique multiethnic cohort including 974 HIV-1 controllers, individuals who show a low level of plasma virus load maintained over at least one year in the absence of treatment, including individuals of European, African-American and Hispanic ancestry. They identify over 300 SNPs meeting genome-wide significance levels, all of which were located within the major histocompatibility complex (MHC). Previous studies have implicated variants in the HLA region in HIV-1 disease progression. Here, the authors used several approaches to fine map the HLA association to HIV-1 control. Using stepwise logistic regression, they identified four independent SNPs, which together account for 19% of the variance in the European sample. They used a new method for imputation of classical HLA alleles and amino acids and tested these for association. They identified three amino acids in HLA-B as being the most significantly associated. These amino acids are found in the peptide binding groove, suggesting HLA-mediated peptide presentation of viral peptides is involved in host control. These associations to SNPs, classical HLA alleles and amino acids were also replicated in the Swiss HIV Cohort. OB

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