Hydration-mediated G-protein-coupled receptor activation.

The Rhodopsin family of G-protein­coupled receptors (GPCRs) comprises the targets of nearly a third of all pharmaceuticals. Despite structural water present in GPCR X-ray structures, the physiological relevance of these solvent molecules to rhodopsin signaling remains unknown. Here, we show experimental results consistent with the idea that rhodopsin activation in lipid membranes is coupled to bulk water movements into the protein. To quantify hydration changes, we measured reversible shifting of the metarhodopsin equilibrium due to osmotic stress using an extensive series of polyethylene glycol (PEG) osmolytes. We discovered clear evidence that light activation entails a large influx of bulk water (∼80­100 molecules) into the protein, giving insight into GPCR activation mechanisms. Various size polymer osmolytes directly control rhodopsin activation, in which large solutes are excluded from rhodopsin and dehydrate the protein, favoring the inactive state. In contrast, small osmolytes initially forward shift the activation equilibrium until a quantifiable saturation point is reached, similar to gain-of-function protein mutations. For the limit of increasing osmolyte size, a universal response of rhodopsin to osmotic stress is observed, suggesting it adopts a dynamic, hydrated sponge-like state upon photoactivation. Our results demand a rethinking of the role of water dynamics in modulating various intermediates in the GPCR energy landscape. We propose that besides bound water, an influx of bulk water plays a necessary role in establishing the active GPCR conformation that mediates signaling.

Activation of the G-Protein-Coupled Receptor Rhodopsin by Water.

Visual rhodopsin is an important archetype for G-protein-coupled receptors, which are membrane proteins implicated in cellular signal transduction. Herein, we show experimentally that approximately 80 water molecules flood rhodopsin upon light absorption to form a solvent-swollen active state. An influx of mobile water is necessary for activating the photoreceptor, and this finding is supported by molecular dynamics (MD) simulations. Combined force-based measurements involving osmotic and hydrostatic pressure indicate the expansion occurs by changes in cavity volumes, together with greater hydration in the active metarhodopsin-II state. Moreover, we discovered that binding and release of the C-terminal helix of transducin is coupled to hydration changes as may occur in visual signal amplification. Hydration-dehydration explains signaling by a dynamic allosteric mechanism, in which the soft membrane matter (lipids and water) has a pivotal role in the catalytic G-protein cycle.

Native Mass Spectrometry Reveals the Simultaneous Binding of Lipids and Zinc to Rhodopsin.

Rhodopsin, a prototypical G-protein-coupled receptor, is responsible for scoptic vision at low-light levels. Although rhodopsin's photoactivation cascade is well understood, it remains unclear how lipid and zinc binding to the receptor are coupled. Using native mass spectrometry, we developed a novel data analysis strategy to deconvolve zinc and lipid bound to the proteoforms of rhodopsin and investigated the allosteric interaction between lipids and zinc binding. We discovered that phosphatidylcholine bound to rhodopsin with a greater affinity than phosphatidylserine or phosphatidylethanolamine, and that binding of all lipids was influenced by zinc but with different effects. In contrast, zinc binding was relatively unperturbed by lipids. Overall, these data reveal that lipid binding can be strongly and differentially influenced by metal ions.

Multi-Parametric Fusion of 3D Power Doppler Ultrasound for Fetal Kidney Segmentation Using Fully Convolutional Neural Networks.

Kidney development is key to the long-term health of the fetus. Renal volume and vascularity assessed by 3D ultrasound (3D-US) are known markers of wellbeing, however, a lack of real-time image segmentation solutions preclude these measures being used in a busy clinical environment. In this work, we aimed to automate kidney segmentation using fully convolutional neural networks (fCNNs). We used multi-parametric input fusion incorporating 3D B-Mode and power Doppler (PD) volumes, aiming to improve segmentation accuracy. Three different fusion strategies and their performance were assessed versus a single input (B-Mode) network. Early input-level fusion provided the best segmentation accuracy with an average Dice similarity coefficient (DSC) of 0.81 and Hausdorff distance (HD) of 8.96 mm, an improvement of 0.06 DSC and reduction of 1.43 mm HD compared to our baseline network. Compared to manual segmentation for all models, repeatability was assessed by intra-class correlation coefficients (ICC) indicating good to excellent reproducibility (ICC 0.93). The framework was extended to support multiple graphics processing units (GPUs) to better handle volumetric data, dense fCNN models, batch normalization and complex fusion networks. This work and available source code provides a framework to increase the parameter space of encoder-decoder style fCNNs across multiple GPUs and shows that application of multi-parametric 3D-US in fCNN training improves segmentation accuracy.