While the majority of proteins fold rapidly and spontaneously to their native states, the extracellular bacterial protease alpha-lytic protease (alphaLP) has a t(1/2) for folding of approximately 2,000 years, corresponding to a folding barrier of 30 kcal mol(-1). AlphaLP is synthesized as a pro-enzyme where its pro region (Pro) acts as a foldase to stabilize the transition state for the folding reaction. Pro also functions as a potent folding catalyst when supplied as a separate polypeptide chain, accelerating the rate of alphaLP folding by a factor of 3 x 10(9). In the absence of Pro, alphaLP folds only partially to a stable molten globule-like intermediate state. Addition of Pro to this intermediate leads to rapid formation of native alphaLP. Here we report the crystal structures of Pro and of the non-covalent inhibitory complex between Pro and native alphaLP. The C-shaped Pro surrounds the C-terminal beta-barrel domain of the folded protease, forming a large complementary interface. Regions of extensive hydration in the interface explain how Pro binds tightly to the native state, yet even more tightly to the folding transition state. Based on structural and functional data we propose that a specific structural element in alphaLP is largely responsible for the folding barrier and suggest how Pro can overcome this barrier.

Intrinsic elasticity of the septal cartilage, the mucoperichondrial flap, and overlap with the bony vault all contribute to the stability of the L-strut, which is enhanced by preserving a small segment of cartilage at the bony-cartilaginous junction of the dorsal L-strut.

Silent speech recognition (SSR) converts non-audio information such as articulatory movements into text. SSR has the potential to enable persons with laryngectomy to communicate through natural spoken expression. Current SSR systems have largely relied on speaker-dependent recognition models. The high degree of variability in articulatory patterns across different speakers has been a barrier for developing effective speaker-independent SSR approaches. Speaker-independent SSR approaches, however, are critical for reducing the amount of training data required from each speaker. In this paper, we investigate speaker-independent SSR from the movements of flesh points on tongue and lip with articulatory normalization methods that reduce the inter-speaker variation. To minimize the across-speaker physiological differences of the articulators, we propose Procrustes matching-based articulatory normalization by removing locational, rotational, and scaling differences. To further normalize the articulatory data, we apply feature-space maximum likelihood linear regression and i-vector. In this paper, we adopt a bidirectional long short term memory recurrent neural network (BLSTM) as an articulatory model to effectively model the articulatory movements with long-range articulatory history. A silent speech data set with flesh points was collected using an electromagnetic articulograph (EMA) from twelve healthy and two laryngectomized English speakers. Experimental results showed the effectiveness of our speaker-independent SSR approaches on healthy as well as laryngectomy speakers. In addition, BLSTM outperformed standard deep neural network. The best performance was obtained by BLSTM with all the three normalization approaches combined.

No abstract