Structure of RNA polymerase I transcribing ribosomal DNA genes.

RNA polymerase I (Pol I) is a highly processive enzyme that transcribes ribosomal DNA (rDNA) and regulates growth of eukaryotic cells. Crystal structures of free Pol I from the yeast Saccharomyces cerevisiae have revealed dimers of the enzyme stabilized by a 'connector' element and an expanded cleft containing the active centre in an inactive conformation. The central bridge helix was unfolded and a Pol-I-specific 'expander' element occupied the DNA-template-binding site. The structure of Pol I in its active transcribing conformation has yet to be determined, whereas structures of Pol II and Pol III have been solved with bound DNA template and RNA transcript. Here we report structures of active transcribing Pol I from yeast solved by two different cryo-electron microscopy approaches. A single-particle structure at 3.8 Å resolution reveals a contracted active centre cleft with bound DNA and RNA, and a narrowed pore beneath the active site that no longer holds the RNA-cleavage-stimulating domain of subunit A12.2. A structure at 29 Å resolution that was determined from cryo-electron tomograms of Pol I enzymes transcribing cellular rDNA confirms contraction of the cleft and reveals that incoming and exiting rDNA enclose an angle of around 150°. The structures suggest a model for the regulation of transcription elongation in which contracted and expanded polymerase conformations are associated with active and inactive states, respectively.

A framework for large-scale mapping of human settlement extent from Sentinel-2 images via fully convolutional neural networks.

Human settlement extent (HSE) information is a valuable indicator of world-wide urbanization as well as the resulting human pressure on the natural environment. Therefore, mapping HSE is critical for various environmental issues at local, regional, and even global scales. This paper presents a deep-learning-based framework to automatically map HSE from multi-spectral Sentinel-2 data using regionally available geo-products as training labels. A straightforward, simple, yet effective fully convolutional network-based architecture, Sen2HSE, is implemented as an example for semantic segmentation within the framework. The framework is validated against both manually labelled checking points distributed evenly over the test areas, and the OpenStreetMap building layer. The HSE mapping results were extensively compared to several baseline products in order to thoroughly evaluate the effectiveness of the proposed HSE mapping framework. The HSE mapping power is consistently demonstrated over 10 representative areas across the world. We also present one regional-scale and one country-wide HSE mapping example from our framework to show the potential for upscaling. The results of this study contribute to the generalization of the applicability of CNN-based approaches for large-scale urban mapping to cases where no up-to-date and accurate ground truth is available, as well as the subsequent monitor of global urbanization.

CENP-A arrays are more condensed than canonical arrays at low ionic strength.

The centromeric histone H3 variant centromeric protein A (CENP-A), whose sequence is the least conserved among all histone variants, is responsible for specifying the location of the centromere. Here, we present a comprehensive study of CENP-A nucleosome arrays by cryo-electron tomography. We see that CENP-A arrays have different biophysical properties than canonical ones under low ionic conditions, as they are more condensed with a 20% smaller average nearest-neighbor distance and a 30% higher nucleosome density. We find that CENP-A nucleosomes have a predominantly crossed DNA entry/exit site that is narrowed on average by 8°, and they have a propensity to stack face to face. We therefore propose that CENP-A induces geometric constraints at the nucleosome DNA entry/exit site to bring neighboring nucleosomes into close proximity. This specific property of CENP-A may be responsible for generating a fundamental process that contributes to increased chromatin fiber compaction that is propagated under physiological conditions to form centromeric chromatin.

Urban expansion simulation with an explainable ensemble deep learning framework.

Urban expansion simulation is of significant importance to land management and policymaking. Advances in deep learning facilitate capturing and anticipating urban land dynamics with state-of-the-art accuracy properties. In this context, a novel deep learning-based ensemble framework was proposed for urban expansion simulation at an intra-urban granular level. The ensemble framework comprises i) multiple deep learning models as encoders, using transformers for encoding multi-temporal spatial features and convolutional layers for processing single-temporal spatial features, ii) a tailored channel-wise attention module to address the challenge of limited interpretability in deep learning methods. The channel attention module enables the examination of the rationality of feature importance, thereby establishing confidence in the simulated results. The proposed method accurately anticipated urban expansion in Shenzhen, China, and it outperformed all the baseline methods in terms of both spatial accuracy and temporal consistency.