Is the Membrane Lipid Matrix a Key Target for Action of Pharmacologically Active Plant Saponins?

This study was focused on the molecular mechanisms of action of saponins and related compounds (sapogenins and alkaloids) on model lipid membranes. Steroids and triterpenes were tested. A systematic analysis of the effects of these chemicals on the physicochemical properties of the lipid bilayers and on the formation and functionality of the reconstituted ion channels induced by antimicrobial agents was performed. It was found that digitonin, tribulosin, and dioscin substantially reduced the boundary potential of the phosphatidylcholine membranes. We concluded that saponins might affect the membrane boundary potential by restructuring the membrane hydration layer. Moreover, an increase in the conductance and lifetime of gramicidin A channels in the presence of tribulosin was due to an alteration in the membrane dipole potential. Differential scanning microcalorimetry data indicated the key role of the sapogenin core structure (steroid or triterpenic) in affecting lipid melting and disordering. We showed that an alteration in pore forming activity of syringomycin E by dioscin might be due to amendments in the lipid packing. We also found that the ability of saponins to disengage the fluorescent marker calcein from lipid vesicles might be also determined by their ability to induce a positive curvature stress.

New Boundary Constraint Loss to Facilitate Glands Segmentation

Histological assessment of glandular formation plays a crucial role in cancer grading, and automated glandular structures segmentation is the first step. To achieve an accurate boundary representation to separate packed glands, an extra branch is normally introduced to predict binary boundary mask in deep network. In this paper, we only use simple single stream networks, facilitated by our proposed boundary constraint loss, to locate the gland boundary and separate clustered glands, making our model efficient in terms of memory usage. To achieve this, we define a new boundary potential map from ground truth region mask. During training, edge representation is calculated from predicted region segmentation, from which a new regularization term is generated to force edge align well with the boundary potential. In addition, we also present a new differentiable version of argmax function to allow the model to be trained in an end-to-end manner. In evaluation, we employed three base model structures (suggestive annotation, U-net and FCN-32s) and two public dataset (GlaS15 and CRAG). Extensive experiments consistently demonstrated that it improves the boundaries segmentation performance over its counterparts (3% in object F1-score and 2% in object Dice).

Electronic structure and transport properties of graphene/h-BN controlled by boundary potential and magnetic field

We study the band structure of the lattice-matched graphene/[Formula: see text]-BN bilayer system in the most stable configuration. An effective way to individually manipulate the edge state by the boundary potentials is proposed. It is shown that the boundary potential can not only shift and deform the edge bands, but also modify the energy gap. We also explore the transport properties of graphene/[Formula: see text]-BN under a magnetic field. The boundary potential can change the distribution of the edge states, resulting in an interesting evolution of the quantized conductance.

A position-sensitive electronic skin based on boundary potential projection theory

Purpose This paper aims to report a flexible position-sensitive sensor that can be applied as large-area electronic skin over the stiff media. Design/methodology/approach The sensor uses a whole piezoresistive film as a touch sensing area. By alternately constructing two uniform electric fields with orthogonal directions in the piezoresistive film, the local changes in conductivity caused by touch can be projected to the boundary along the equipotential line under the constraint of electric field. Based on the change of boundary potential in the two uniform electric fields, it can be easy to determine the position of the contact area in the piezoresistive film. Findings Experiment results show the proposed tactile sensor is capable of detecting the contact position and classifying the contact force in real-time based on the changes of the potential differences on the boundary of the sensor. Practical implications The application example of using the sensor sample as a controller in shooting game is presented in this paper. It shows that the sensor has excellent touch sensing performance. Originality/value In this paper, a position-sensitive electronic skin is proposed. The experiment results show that the sensor has great application prospects in the field of interactive tactile sensing.