Acidic microenvironment is commonly observed in tumour tissues, including glioblastoma (GBM), the most aggressive and lethal brain tumour in adults. Acid sensing ion channels (ASICs) are neuronal voltage-insensitive sodium channels, which are sensors of extracellular protons. Here we studied and functionally characterized ASICs in two primary glioblastoma stem cell lines as cell culture models. We detected transcripts of the ACCN2 and ACCN3 genes, coding for ASIC1 and ASIC3, respectively, but not transcripts of ACCN1 (coding for ASIC2). Available microarray data confirmed that ACCN1 is downregulated in glioma. Western blotting confirmed expression of ASIC1 and ASIC3, the most proton-sensitive ASICs, in both GBM cell lines. We characterized ASICs functionally using whole-cell patch clamp and detected different types of acid-sensitive currents. Some of these currents had kinetics typical for ASICs and were sensitive to specific toxin inhibitors of ASIC1a or ASIC3, demonstrating that the GBM cell lines express functional ASIC1a and ASIC3 that may enable GBM cells to sensitively detect extracellular pH in a tumour tissue. Microarray data revealed that expression of ACCN2 and ACCN3 is associated with improved survival of patients suffering from gliomas, suggesting that preserved susceptibility to extracellular pH may impair tumour growth.

Activation of acid-sensing ion channels (ASICs) contributes to neuronal death during stroke, to axonal degeneration during neuroinflammation, and to pain during inflammation. Although understanding ASIC gating may help to modulate ASIC activity during these pathologic situations, at present it is poorly understood. The ligand, H ؉ , probably binds to several sites, among them amino acids within the large extracellular domain. The extracellular domain is linked to the two transmembrane domains by the wrist region that is connected to two anti-parallel ␤-strands, ␤1 and ␤12. Thus, the wrist region together with those ␤-strands may have a crucial role in transmitting ligand binding to pore opening and closing. Here we show that amino acids in the ␤1-␤2 linker determine constitutive opening of ASIC1b from shark. The most crucial residue within the ␤1-␤2 linker (Asp 110 ), when mutated from aspartate to cysteine, can be altered by cysteine-modifying reagents much more readily when channels are closed than when they are desensitized. Finally, engineering of a cysteine at position 110 and at an adjacent position in the ␤11-␤12 linker leads to spontaneous formation of a disulfide bond that traps the channel in the desensitized conformation. Collectively, our results suggest that the ␤1-␤2 and ␤11-␤12 linkers are dynamic during gating and tightly appose to each other during desensitization gating. Hindrance of this tight apposition leads to reopening of the channel. It follows that the ␤1-␤2 and ␤11-␤12 linkers modulate gating movements of ASIC1 and may thus be drug targets to modulate ASIC activity.

Guided Policy Search enables robots to learn control policies for complex manipulation tasks efficiently. Therein, the control policies are represented as high-dimensional neural networks which derive robot actions based on states. However, due to the small number of real-world trajectory samples in Guided Policy Search, the resulting neural networks are only robust in the neighbourhood of the trajectory distribution explored by real-world interactions. In this paper, we present a new policy representation called Generative Motor Reflexes, which is able to generate robust actions over a broader state space compared to previous methods. In contrast to prior state-action policies, Generative Motor Reflexes map states to parameters for a state-dependent motor reflex, which is then used to derive actions. Robustness is achieved by generating similar motor reflexes for many states. We evaluate the presented method in simulated and real-world manipulation tasks, including contact-rich peg-in-hole tasks. Using these evaluation tasks, we show that policies represented as Generative Motor Reflexes lead to robust manipulation skills also outside the explored trajectory distribution with less training needs compared to previous methods.1 Philipp Ennen, Pia Bresenitz, Rene Vossen, Frank Hees are with the

No abstract

No abstract