Ancient polymerase

The archaeal RNA polymerase shares structural similarities with eukaryotic RNA polymerase II, but differs in that the archaeal polymerase requires only two accessory factors whereas the eukaryotic version requires several. The overall structure of the archaeal RNA polymerase was recently described, but some of the subunits were missing. Now, Abrescia and colleagues complete the picture with the crystal structure of the 13-subunit RNA polymerase from Sulfolobus shibatae at 3.35 Å. The complete structure reveals two new subunits: Rpo8, which is equivalent to the eukaryotic Rbp8, and Rpo13, which has no ortholog in the eukaryotic complex. Both of these subunits were absent from the earlier archaeal polymerase structures. Although the function of Rpo13 is still unknown, its location suggests that it has a role in initiation and elongation. The authors indicate that Rpo13 could assist formation of the transcription bubble once the pre-initiation complex is formed. The structure of the RNA polymerase allied with modeling of the atomic details of Rpo13 suggest that its C terminus is about 7 Å from the phosphate group of the nucleotide at a position 8 residues downstream from the start site of the modeled nontemplate DNA strand. The authors suggest that Rpo13 acts as a 'lock point' against which the main body of the polymerase cleft can push or twist DNA. This could explain how Rpo13 performs some of the roles attributed to eukaryote-specific general transcription factors. (PLoS Biol. 7, e1000102, 2009) MH

Oct4 and counting

Dosage compensation in mice and humans involves the inactivation of one X chromosome in females during early differentiation to achieve similar X-associated gene expression in both sexes. Recent data have suggested a link between pluripotency factors and X chromosome inactivation, which would explain how the latter process is coupled to differentiation. Lee and colleagues now find that the mouse X-inactivation center (XIC), a region key to X inactivation and X chromosome counting, contains sites that bind Sox2 and the pluripotency factor Oct4. The predicted Oct4 and Sox2 sites are associated with regulation of Tsix, a noncoding transcript that is the antisense repressor of the Xist noncoding RNA. Oct4 and Sox2 binding sites were confirmed by EMSA and quantitative ChIP in female mouse embryonic stem (ES) cells. As the sites lie close to the Ctcf and Yy1 binding elements, both previously implicated in dosage compensation, the authors tested and confirmed interactions between Oct4-Ctcf and Sox2-Yy1. Similar to the Ctcf knockdown phenotype, Oct4 depletion resulted in defects in pairing of the two X chromosomes, as well as decreased Tsix expression and a corresponding increase in Xist levels in differentiating ES cells. In addition, both the Oct4 and Sox2 binding sites are needed for upregulation of a Tsix-reporter fusion. As expected, these defects upon Oct4 knockdown are coupled to biallelic expression of Xist, indicating that counting of X chromosomes does not occur correctly. Oct4 expression decreases during differentiation, consistent with the idea that X inactivation would be regulated by this factor. It is proposed that Oct4 ultimately helps to determine X chromosome choice and counting when Oct4 levels decrease to the point where it can bind to only one of the two X chromosomes. How Oct4 ends up localized to one X chromosome remains an interesting question for future study. (Nature, advance online publication, doi:10.1038/nature08098, 17 June 2009) SL

Insulating Pol II

Insulators are chromatin boundary elements that help to regulate gene activity by ensuring appropriate interactions between cis-regulatory elements. A number of well-defined insulators present in the Drosophila melanogaster Hox clusters have previously been defined as key to the ordered expression of these genes during embryogenesis and can engage with regulated promoters through long-range interactions. The Hox clusters also contain several genes, lab, Antp, Ubx and abd-B, with stalled RNA polymerase II (Pol II) at their promoters. Chopra, Levine and colleagues now bring these phenomena together by arguing that promoters that carry a stalled polymerase can also behave as insulators. The authors placed promoter regions upstream of a LacZ reporter and between a white reporter and strong enhancer element (IAB-5), finding that those derived from Hox genes with a stalled or paused polymerase could block enhancer activity, whereas lack of a stalled polymerase correlated with no enhancer-blocking activity. The authors show that the promoters per se do not compete for the IAB-5 enhancer, arguing against a model in which one promoter is simply competing more strongly for interaction with the enhancer. To test whether the stalled Pol II might be involved in enhancer-blocking activity, the authors decreased the dose of factors previously implicated in Pol II pausing, DRB sensitivity–inducing factor (DSIF) and negative elongation factor (NELF). In heterozygous mutants for these factors, enhancer-blocking activity is decreased. Further work will reveal whether known insulator proteins are involved in these activities and how promoters carrying stalled polymerases can act as insulators. (Genes Dev. published online, doi:10.1101/gad.1807309, 10 June 2009) SL

Breaking the blood-brain barrier

Our central nervous system (CNS) is guarded by several layers of protection: the skull and vertebrae form a rigid shield and a special system of membranes (meninges) envelops the CNS, which is bathed in cerebrospinal fluid. Finally, many substances or particles in circulation cannot reach the CNS, owing to the so-called blood-brain barrier (BBB). This restriction applies to antibodies, hormones, most microorganisms and many drugs; in fact, this poses a challenge for the delivery of therapeutic drugs to the CNS. The BBB is conferred by tight and adherens junctions that mediate interactions between the endothelial cells forming the CNS capillaries; astrocyte projections also seem to contribute to the barrier. However, the BBB is not impenetrable: the bacterium Neisseria meningitidis can get into the bloodstream, break the BBB and invade the meninges, causing high-mortality-rate meningitis. N. meningitidis uses appendages called type 4 pili to adhere to endothelial cells and form microcolonies. Coureuil and colleagues now investigate how this results in crossing of the BBB. Using a human brain endothelial cell line that forms a monolayer with tight junctions in culture, the authors observed that N. meningitidis recruited proteins that determine cell polarity to the site of bacteria-endothelial cell interaction. This in turn led to the redistribution of components of the tight and adherens junctions to these same sites. The resulting depletion of junction proteins from the cell-cell interface increased the permeability and caused formation of visible gaps between cells, allowing bacteria to cross the monolayer. Only bacteria expressing type 4 pili can do this, and it will be fascinating to learn about the nature of the signal triggered by these appendages that causes such a dramatic effect on endothelial cells. (Science express, published online 11 June 2009, doi 10.1126/science.1173196) IC

Written by Inês Chen, Maria Hodges & Sabbi Lall