The mechanism for the lysis pathway of double-stranded DNA bacteriophages involves a small hole-forming class of membrane proteins, the holins. This study focuses on a poorly characterized class of holins, the pinholin, of which the S 21 protein of phage ϕ21 is the prototype. Here we report the first in vitro synthesis of the wildtype form of the S 21 pinholin, S 21 68, and negative-dominant mutant form, S 21 IRS, both prepared using solid phase peptide synthesis and studied using biophysical techniques. Both forms of the pinholin were labeled with a nitroxide spin label and successfully incorporated into both bicelles and multilamellar vesicles which are membrane mimetic systems. Circular dichroism revealed the two forms were both >80% alpha helical, in agreement with the predictions based on the literature. The molar ellipticity ratio [θ] 222 /[θ] 208 for both forms of the pinholin was 1.4, suggesting a coiled-coil tertiary structure in the bilayer consistent with the proposed oligomerization step in models for the mechanism of hole formation. 31 P solid-state NMR spectroscopic data on pinholin indicate a strong interaction of both forms of the pinholin with the membrane headgroups. The 31 P NMR data has an axially symmetric line shape which is consistent with lamellar phase proteoliposomes lipid mimetics.
Failures in prospective memory (PM) - that is, the failure to remember intended future actions - can have adverse consequences. It is therefore important to study those processes that may help to minimize such cognitive failures. Although multisensory integration has been shown to enhance a wide variety of behaviors, including perception, learning, and memory, its effect on prospective memory, in particular, is largely unknown. In the present study, we investigated the effects of multisensory processing on two simultaneously-performed memory tasks: An ongoing 2- or 3-back working memory (WM) task (20% target ratio), and a PM task in which the participants had to respond to a rare predefined letter (8% target ratio). For PM trials, multisensory enhancement was observed for congruent multisensory signals; however, this effect did not generalize to the ongoing WM task. Participants were less likely to make errors for PM than for WM trials, thus suggesting that they may have biased their attention toward the PM task. Multisensory advantages on memory tasks, such as PM and WM, may be dependent on how attention resources are allocated across dual tasks.
A lack of sufficient labeled data often limits the applicability of advanced machine learning algorithms to real life problems. However, the efficient use of transfer learning (TL) has been shown to be very useful across domains. TL make use of valuable knowledge learned in one task (source task), where sufficient data is available, in order to improve performance on the task of interest (target task). In the biomedical and clinical domain, a lack of sufficient training data means that machine learning models cannot be fully exploited. In this work, we present two unified recurrent neural models leading to three transfer learning frameworks for relation classification tasks. We systematically investigate the effectiveness of the proposed frameworks in transferring knowledge from a source task to a target task when the characteristics of the source data vary, such as similarity or relatedness between the source and target tasks, and the size of training data for the source task. Our empirical results show that the proposed frameworks, in general, improve the model performance. However, these improvements do depend on characteristics of source and target tasks. This dependence then finally determine the choice of a particular TL framework.
Electron spin echo envelope modulation (ESEEM) spectroscopy in combination with site-directed spin labeling (SDSL) has been established as a valuable biophysical technique to provide site-specific local secondary structure of membrane proteins. This pulsed electron paramagnetic resonance (EPR) method can successfully distinguish between α-helices, β-sheets, and 3 10 -helices by strategically using 2 H-labeled amino acids and SDSL. In this study, we have explored the use of 13 C-labeled residues as the NMR active nuclei for this approach for the first time. 13 C-labeled d 5 -valine (Val) or 13 C-labeled d 6 -leucine (Leu) were substituted at a specific Val or Leu residue (i), and a nitroxide spin label was positioned 2 or 3 residues away (denoted i-2 and i-3) on the acetylcholine receptor M2δ (AChR M2δ) in a lipid bilayer. The 13 C ESEEM peaks in the FT frequency domain data were observed for the i-3 samples, and no 13 C peaks were observed in the i-2 samples. The resulting spectra were indicative of the α-helical local secondary structure of AChR M2δ in bicelles. This study provides more versatility and alternative options when using this ESEEM approach to study the more challenging recombinant membrane protein secondary structures.