Analyzing the prenylome

Protein prenylation refers to the attachment of farnesyl or geranylgeranyl groups to target proteins via prenyltransferases. To explore the total mammalian prenylome, Nguyen et al. developed a biotin-labeled geranylpyrophosphate substrate, as well as a series of prenyltransferases capable of conjugating the substrate to target proteins. This approach allowed all prenylated proteins to be detected by western blotting and mass spectrometry.

Nguyen, U.T.T. et al. Nat. Chem. Biol. advance online publication 15 February 2009.


Identifying lincRNAs

Mammalian genomes encode a number of large intervening noncoding RNAs (lincRNAs), but their functional importance is not well-understood. Guttman et al. used chromatin immunoprecipitation followed by massively parallel sequencing to identify chromatin domains indicative of active transcriptional units residing outside of protein-coding gene loci. They found that a majority of the lincRNAs were highly evolutionarily conserved, indicating that they are likely functional. They also describe an approach for predicting the function of these lincRNAs.

Guttman, M. et al. Nature 458, 223–227 (2009).


Whole cell electron microscopy

de Jonge et al. describe a method for imaging whole cells in their native liquid state, using scanning transmission electron microscopy. Using gold nanoparticles as tags, they imaged single epidermal growth factor molecules bound to their receptors with a spatial resolution of 4 nanometers, higher than current super-resolution imaging methods. The rapid imaging speed may allow snapshots of dynamic information to be captured.

de Jonge, N. et al. Proc. Natl. Acad. Sci. USA 106, 2159–2164 (2009).


A gene expression atlas of the mouse brain

Ng et al. present the Anatomic Gene Expression Atlas (AGEA) of the adult C57Bl/6J mouse brain. This publicly available resource was created from more than 4,000 gene expression profiles from in situ hybridization data from the Allen Brain Atlas. The AGEA allows users to explore three-dimensional gene expression–based correlation and clustering maps of the brain, and a search tool allows users to retrieve lists of enriched genes at points of interest.

Ng, L. et al. Nat. Neurosci. 12, 356–362 (2009).

Protein biochemistry

Predicting membrane protein structures

Membrane protein structures are notoriously difficult to solve experimentally, but they also pose a challenge for in silico structural prediction. Barth et al. describe a computational method for determining the structures of large membrane proteins by enforcing constraints on helix-helix packing interactions as predicted from the protein sequence or from experiments. They obtained near-native models for 9 of 12 membrane proteins.

Barth, P. et al. Proc. Natl. Acad. Sci. USA 106, 1409–1414 (2009).