The Correlation Between Glycation Gap and Renal Complications in Patients with Type 2 Diabetes Mellitus.

Purpose: The aim of this study was to investigate the correlations between the glycation gap (GG) and renal complications such as urinary albumin-creatinine ratio (UACR) and estimated glomerular filtration rate (eGFR) in type 2 diabetes mellitus patients. Materials and Methods: A cross-sectional study was conducted on 104 individuals (52 males and 52 females), aged 36-93 years old. Fasting blood glucose (FBG), HbA1c, and serum fructosamine were measured simultaneously. GG was calculated as the difference between the measured and fructosamine-based predicted HbA1c levels (FHbA1c). Results: There was a moderately positive correlation between HbA1c and fructosamine concentration (r = 0.488; p < 0.001). GG was positively correlated with UACR (r = 0.3275; p = 0.0007), negatively correlated with eGFR (r = -0.3400; p = 0.0004). HbA1c was positively correlated with UACR (r = 0.2437; p = 0.0127) but not correlated with eGFR (r = -0.444; p = 0.6542). Fructosamine has a positive correlation with eGFR (r = 0.2426; p = 0.0131) but not with UACR (r = -0.1021; p = 0.3025). Conclusion: GG was positively correlated with UACR and inversely correlated with eGFR in type 2 Diabetes mellitus patients. This suggests that GG is a valuable index for predicting kidney complications due to diabetes.

Effects of Catalyst Pretreatment on Carbon Nanotube Synthesis from Methane Using Thin Stainless-Steel Foil as Catalyst by Chemical Vapor Deposition Method.

Synthesis of carbon nanotubes (CNTs) was carried out using methane as a carbon source via the chemical vapor deposition (CVD) method. A thin stainless-steel foil was used as catalyst for CNT growth. Our results revealed that pretreatment step of the stainless-steel foil as a catalyst plays an important role in CNT formation. In our experiments, a catalyst pretreatment temperature of 850 °C or 950 °C was found to facilitate the creation of Fe- and Cr-rich particles are active sites on the foil surface, leading to CNT formation. It is noted that the size of metallic particles after pretreatment is closely related to the diameter of the synthesized CNTs. It is interesting that a shorter catalyst pretreatment brings the growth of semiconducting typed CNTs while a longer pretreatment creates metallic CNTs. This finding might lead to a process for improving the quality of CNTs grown on steel foil as catalyst.

Novel mechanism of voltage-gated N-type (Cav2.2) calcium channel inhibition revealed through α-conotoxin Vc1.1 activation of the GABA(B) receptor.

Neuronal voltage-gated N-type (Cav2.2) calcium channels are expressed throughout the nervous system and regulate neurotransmitter release and hence synaptic transmission. They are predominantly modulated via G protein-coupled receptor activated pathways, and the well characterized Gßγ subunits inhibit Cav2.2 currents. Analgesic α-conotoxin Vc1.1, a peptide from predatory marine cone snail venom, inhibits Cav2.2 channels by activating pertussis toxin-sensitive Gi/o proteins via the GABAB receptor (GABA(B)R) and potently suppresses pain in rat models. Using a heterologous GABA(B)R expression system, electrophysiology, and mutagenesis, we showed α-conotoxin Vc1.1 modulates Cav2.2 via a different pathway from that of the GABA(B)R agonists GABA and baclofen. In contrast to GABA and baclofen, Vc1.1 changes Cav2.2 channel kinetics by increasing the rate of activation and shifting its half-maximum inactivation to a more hyperpolarized potential. We then systematically truncated the GABA(B)(1a) C terminus and discovered that removing the proximal carboxyl terminus of the GABA(B)(1a) subunit significantly reduced Vc1.1 inhibition of Cav2.2 currents. We propose a novel mechanism by which Vc1.1 activates GABA(B)R and requires the GABA(B)(1a) proximal carboxyl terminus domain to inhibit Cav2.2 channels. These findings provide important insights into how GABA(B)Rs mediate Cav2.2 channel inhibition and alter nociceptive transmission.

γ-Aminobutyric acid type B (GABAB) receptor expression is needed for inhibition of N-type (Cav2.2) calcium channels by analgesic α-conotoxins.

α-Conotoxins Vc1.1 and RgIA are small peptides isolated from the venom of marine cone snails. They have effective anti-nociceptive actions in rat models of neuropathic pain. Pharmacological studies in rodent dorsal root ganglion (DRG) show their analgesic effect is mediated by inhibition of N-type (Ca(v)2.2) calcium channels via a pathway involving γ-aminobutyric acid type B (GABA(B)) receptor. However, there is no direct demonstration that functional GABA(B) receptors are needed for inhibition of the Ca(v)2.2 channel by analgesic α-conotoxins. This study examined the effect of the GABA(B) agonist baclofen and α-conotoxins Vc1.1 and RgIA on calcium channel currents after transient knockdown of the GABA(B) receptor using RNA interference. Isolated rat DRG neurons were transfected with small interfering RNAs (siRNA) targeting GABA(B) subunits R1 and R2. Efficient knockdown of GABA(B) receptor expression at mRNA and protein levels was confirmed by quantitative real time PCR (qRT-PCR) and immunocytochemical analysis, respectively. Whole-cell patch clamp recordings conducted 2-4 days after transfection showed that inhibition of N-type calcium channels in response to baclofen, Vc1.1 and RgIA was significantly reduced in GABA(B) receptor knockdown DRG neurons. In contrast, neurons transfected with a scrambled nontargeting siRNA were indistinguishable from untransfected neurons. In the HEK 293 cell heterologous expression system, Vc1.1 and RgIA inhibition of Ca(v)2.2 channels needed functional expression of both human GABA(B) receptor subunits. Together, these results confirm that GABA(B) receptors must be activated for the modulation of N-type (Ca(v)2.2) calcium channels by analgesic α-conotoxins Vc1.1 and RgIA.

Symbolic representation of recurrent neural network dynamics.

Simple recurrent error backpropagation networks have been widely used to learn temporal sequence data, including regular and context-free languages. However, the production of relatively large and opaque weight matrices during learning has inspired substantial research on how to extract symbolic human-readable interpretations from trained networks. Unlike feedforward networks, where research has focused mainly on rule extraction, most past work with recurrent networks has viewed them as dynamical systems that can be approximated symbolically by finite-state machine (FSMs). With this approach, the network's hidden layer activation space is typically divided into a finite number of regions. Past research has mainly focused on better techniques for dividing up this activation space. In contrast, very little work has tried to influence the network training process to produce a better representation in hidden layer activation space, and that which has been done has had only limited success. Here we propose a powerful general technique to bias the error backpropagation training process so that it learns an activation space representation from which it is easier to extract FSMs. Using four publicly available data sets that are based on regular and context-free languages, we show via computational experiments that the modified learning method helps to extract FSMs with substantially fewer states and less variance than unmodified backpropagation learning, without decreasing the neural networks' accuracy. We conclude that modifying error backpropagation so that it more effectively separates learned pattern encodings in the hidden layer is an effective way to improve contemporary FSM extraction methods.

Guiding hidden layer representations for improved rule extraction from neural networks.

The production of relatively large and opaque weight matrices by error backpropagation learning has inspired substantial research on how to extract symbolic human-readable rules from trained networks. While considerable progress has been made, the results at present are still relatively limited, in part due to the large numbers of symbolic rules that can be generated. Most past work to address this issue has focused on progressively more powerful methods for rule extraction (RE) that try to minimize the number of weights and/or improve rule expressiveness. In contrast, here we take a different approach in which we modify the error backpropagation training process so that it learns a different hidden layer representation of input patterns than would normally occur. Using five publicly available datasets, we show via computational experiments that the modified learning method helps to extract fewer rules without increasing individual rule complexity and without decreasing classification accuracy. We conclude that modifying error backpropagation so that it more effectively separates learned pattern encodings in the hidden layer is an effective way to improve contemporary RE methods.

Studying mechanosensitive ion channels using liposomes.

Mechanosensitive (MS) ion channels are the primary molecular transducers of mechanical force into electrical and/or chemical intracellular signals in living cells. They have been implicated in innumerable mechanosensory physiological processes including touch and pain sensation, hearing, blood pressure control, micturition, cell volume regulation, tissue growth, or cellular turgor control. Much of what we know about the basic physical principles underlying the conversion of mechanical force acting upon membranes of living cells into conformational changes of MS channels comes from studies of MS channels reconstituted into artificial liposomes. Using bacterial MS channels as a model, we have shown by reconstituting these channels into liposomes that there is a close relationship between the physico-chemical properties of the lipid bilayer and structural dynamics bringing about the function of these channels.

Are calcifying matrix vesicles in atherosclerotic lesions of cellular origin?

Over recent years, the role of matrix vesicles in the initial stages of arterial calcification has been recognized. Matrix calcifying vesicles have been isolated from atherosclerotic arteries and the biochemical composition of calcified vesicles has been studied. No studies have yet been carried out to examine the fine structure of matrix vesicles in order to visualize the features of the consequent stages of their calcification in arteries. In the present work, a high resolution ultrastructural analysis has been employed and the study revealed that matrix vesicles in human atherosclerotic lesions are heterogeneous with two main types which we classified. Type I calcified vesicles were presented by vesicles surrounded by two electron-dense layers and these vesicles were found to be resistant to the calcification process in atherosclerotic lesions in situ. Type II matrix vesicles were presented by vesicles surrounded by several electron-dense layers and these vesicles were found to represent calcifying vesicles in atherosclerotic lesions. To test the hypothesis that calcification of matrix vesicles surrounded by multilayer sheets may occur simply as a physicochemical process, independently from the cell regulation, we produced multilamellar liposomes and induced their calcification in vitro in a manner similar to that occurring in matrix vesicles in atherosclerotic lesions in situ.