The Rosetta software suite for macromolecular modeling, docking, and design is widely used in pharmaceutical, industrial, academic, non-profit, and government laboratories. Despite its broad modeling capabilities, Rosetta remains consistently among leading software suites when compared to other methods created for highly specialized protein modeling and design tasks. Developed for over two decades by a global community of over 60 laboratories, Rosetta has undergone multiple refactorings, and now comprises over three million lines of code. Here we discuss methods developed in the last five years in Rosetta, involving the latest protocols for structure prediction; protein-protein and protein-small molecule docking; protein structure and interface design; loop modeling; the incorporation of various types of experimental data; modeling of peptides, antibodies and proteins in the immune system, nucleic acids, non-standard chemistries, carbohydrates, and membrane proteins. We briefly discuss improvements to the energy function, user interfaces, and usability of the software. Rosetta is available at www.rosettacommons.org.

Summary Naturally occurring, pharmacologically active peptides constrained with covalent crosslinks generally have shapes evolved to fit precisely into binding pockets on their targets. Such peptides can have excellent pharmaceutical properties, combining the stability and tissue penetration of small molecule drugs with the specificity of much larger protein therapeutics. The ability to design constrained peptides with precisely specified tertiary structures would enable the design of shape-complementary inhibitors of arbitrary targets. Here we describe the development of computational methods for de novo design of conformationally-restricted peptides, and the use of these methods to design 15–50 residue disulfide-crosslinked and heterochiral N-C backbone-cyclized peptides. These peptides are exceptionally stable to thermal and chemical denaturation, and twelve experimentally-determined X-ray and NMR structures are nearly identical to the computational models. The computational design methods and stable scaffolds presented here provide the basis for development of a new generation of peptide-based drugs.

Glutamate transporters regulate excitatory neurotransmission and prevent glutamate-mediated excitotoxicity in the CNS. To better study the cellular and temporal dynamics of the expression of these transporters, we generated bacterial artificial chromosome promoter Discosoma red [glutamate-aspartate transporter (GLAST)] and green fluorescent protein [glutamate transporter-1 (GLT-1)] reporter transgenic mice. Analysis of these mice revealed a differential activation of the transporter promoters not previously appreciated. GLT-1 promoter activity in the adult CNS is almost completely restricted to astrocytes, often and unexpectedly in a nonoverlapping pattern with GLAST. Spinal cord GLT-1 promoter reporter, protein density, and physiology were 10-fold lower than in brain, suggesting a possible mechanism for regional sensitivity seen in disease. The GLAST promoter is active in both radial glia and many astrocytes in the developing CNS but is downregulated in most astrocytes as the mice mature. In the adult CNS, the highest GLAST promoter activity was observed in radial glia, such as those located in the subgranular layer of the dentate gyrus. The continued expression of GLAST by these neural progenitors raises the possibility that GLAST may have an unanticipated role in regulating their behavior. In addition, GLAST promoter activation was observed in oligodendrocytes in white matter throughout many (e.g., spinal cord and corpus callosum), but not all (e.g., cerebellum), CNS fiber tracts. Overall, these studies of GLT-1 and GLAST promoter activity, protein expression, and glutamate uptake revealed a close correlation between transgenic reporter signals and uptake capacity, indicating that these mice provide the means to monitor the expression and regulation of glutamate transporters in situ.

SUMMARY The neuron-astrocyte synaptic complex is a fundamental operational unit of the nervous system. Astroglia play a central role in the regulation of synaptic glutamate, via neurotransmitter transport by GLT1/EAAT2. The astroglial mechanisms underlying this essential neuron-glial communication are not known. Here we show that presynaptic terminals are sufficient and necessary for GLT1/EAAT2 transcriptional activation and have identified the molecular pathway that regulates astroglial responses to presynaptic input. Presynaptic terminals regulate astroglial GLT1/EAAT2 via kappa B-motif binding phosphoprotein (KBBP), the mouse homologue of human heterogeneous nuclear ribonucleoprotein K (hnRNP K), which binds to an essential element of GLT1/EAAT2 promoter. This neuron-stimulated factor is required for GLT1/EATT2 transcriptional activation and is responsible for astroglial alterations in neural injury. Denervation of neuron-astrocyte signaling in vivo, by acute corticospinal tract transection, ricin-induced motor neuron death, or chronic neurodegeneration in amyotrophic lateral sclerosis (ALS) all result in reduced astroglial KBBP expression and transcriptional dysfunction of astroglial transporter expression. Our studies indicate that presynaptic elements dynamically coordinate normal astroglial function and also provide a fundamental signaling mechanism by which altered neuronal function and injury leads to dysregulated astroglia in CNS disease.

RNA molecules adopt three-dimensional structures that are critical to their function and of interest in drug discovery. Few RNA structures are known, however, and predicting them computationally has proven challenging. We introduce a machine learning approach that enables identification of accurate structural models without assumptions about their defining characteristics, despite being trained with only 18 known RNA structures. The resulting scoring function, the Atomic Rotationally Equivariant Scorer (ARES), substantially outperforms previous methods and consistently produces the best results in community-wide blind RNA structure prediction challenges. By learning effectively even from a small amount of data, our approach overcomes a major limitation of standard deep neural networks. Because it uses only atomic coordinates as inputs and incorporates no RNA-specific information, this approach is applicable to diverse problems in structural biology, chemistry, materials science, and beyond.