DeepFunc: A Deep Learning Framework for Accurate Prediction of Protein Functions from Protein Sequences and Interactions

Annotation of protein functions plays an important role in understanding life at the molecular level. High‐throughput sequencing produces massive numbers of raw proteins sequences and only about 1% of them have been manually annotated with functions. Experimental annotations of functions are expensive, time‐consuming and do not keep up with the rapid growth of the sequence numbers. This motivates the development of computational approaches that predict protein functions. A novel deep learning framework, DeepFunc, is proposed which accurately predicts protein functions from protein sequence‐ and network‐derived information. More precisely, DeepFunc uses a long and sparse binary vector to encode information concerning domains, families, and motifs collected from the InterPro tool that is associated with the input protein sequence. This vector is processed with two neural layers to obtain a low‐dimensional vector which is combined with topological information extracted from protein–protein interactions (PPIs) and functional linkages. The combined information is processed by a deep neural network that predicts protein functions. DeepFunc is empirically and comparatively tested on a benchmark testing dataset and the Critical Assessment of protein Function Annotation algorithms (CAFA) 3 dataset. The experimental results demonstrate that DeepFunc outperforms current methods on the testing dataset and that it secures the highest F_max = 0.54 and AUC = 0.94 on the CAFA3 dataset.     

On the Relation Between the Predicted Secondary Structure and the Protein Size

Accurately predicted protein secondary structure provides useful information for target selection, to analyze protein function and to predict higher dimensional structure. Existing research shows that more data + refined search = better prediction. We analyze relation between the prediction accuracy and another crucial factor, the protein size. Empirical tests performed with two secondary structure predictors on a large set of high-resolution, non-redundant proteins show that the average accuracies for small proteins (<100 residues) equal 73% and 54% for alpha-helices and beta-strands, respectively. The alpha-helix/beta-strand accuracies for very large proteins (>300 residues) equal 77%/68%, respectively. Similarly, the tests with three secondary structure content predictors show that the prediction errors for the small/very large proteins equal 0.13/0.09 and 0.09/0.06 for alpha-helix and beta-strand content, respectively. Our tests confirm that the secondary structure/content predictions for the very large proteins are characterized statistically significantly better quality than prediction for the small proteins. This is in contrast with the tertiary structure predictions in which higher accuracy is obtained for smaller proteins.     

Structures of a Bifunctional Cell Wall Hydrolase CwlT Containing a Novel Bacterial Lysozyme and an NlpC/P60 dl-Endopeptidase

Tn916-like conjugative transposons carrying antibiotic resistance genes are found in a diverse range of bacteria. Orf14 within the conjugation module encodes a bifunctional cell wall hydrolase CwlT that consists of an N-terminal bacterial lysozyme domain (N-acetylmuramidase, bLysG) and a C-terminal NlpC/P60 domain (γ-d-glutamyl-l-diamino acid endopeptidase) and is expected to play an important role in the spread of the transposons. We determined the crystal structures of CwlT from two pathogens, Staphylococcus aureus Mu50 (SaCwlT) and Clostridium difficile 630 (CdCwlT). These structures reveal that NlpC/P60 and LysG domains are compact and conserved modules, connected by a short flexible linker. The LysG domain represents a novel family of widely distributed bacterial lysozymes. The overall structure and the active site of bLysG bear significant similarity to other members of the glycoside hydrolase family 23 (GH23), such as the g-type lysozyme (LysG) and Escherichia coli lytic transglycosylase MltE. The active site of bLysG contains a unique structural and sequence signature (DxxQSSES + S) that is important for coordinating a catalytic water. Molecular modeling suggests that the bLysG domain may recognize glycan in a similar manner to MltE. The C-terminal NlpC/P60 domain contains a conserved active site (Cys-His-His-Tyr) that appears to be specific to murein tetrapeptide. Access to the active site is likely regulated by isomerism of a side chain atop the catalytic cysteine, allowing substrate entry or product release (open state), or catalysis (closed state).     

Gene silencing activity of siRNA molecules containing phosphorodithioate substitutions

Chemically synthesized small interfering RNAs (siRNAs) have been widely used to identify gene function and hold great potential in providing a new class of therapeutics. Chemical modifications are desired for therapeutic applications to improve siRNA efficacy. Appropriately protected ribonucleoside-3'-yl S-[β-(benzoylmercapto)ethyl]pyrrolidino-thiophosphoramidite monomers were prepared for the synthesis of siRNA containing phosphorodithioate (PS2) substitutions in which the two non-bridging oxygen atoms are replaced by sulfur atoms. A series of siRNAs containing PS2 substitutions have been strategically designed, synthesized, and evaluated for their gene silencing activities. These PS2-siRNA duplexes exhibit an A-form helical structure similar to unmodified siRNA. The effect of PS2 substitutions on gene silencing activity is position-dependent, with certain PS2-siRNAs showing activity significantly higher than that of unmodified siRNA. The relative gene silencing activities of siRNAs containing either PS2 or phosphoromonothioate (PS) linkages at identical positions are variable and depend on the sites of modification. 5'-Phosphorylation of PS2-siRNAs has little or no effect on gene silencing activity. Incorporation of PS2 substitutions into siRNA duplexes increases their serum stability. These results offer preliminary evidence of the potential value of PS2-modified siRNAs.     

Effects of the novel mitochondrial protein mimitin in insulin-secreting cells

Mimitin, a novel mitochondrial protein, has been shown to act as a molecular chaperone for the mitochondrial complex I and to regulate ATP synthesis. During Type 1 diabetes development, pro-inflammatory cytokines induce mitochondrial damage in pancreatic β-cells, inhibit ATP synthesis and reduce glucose-induced insulin secretion. Mimitin was expressed in rat pancreatic islets including β-cells and decreased by cytokines. In the ob/ob mouse, a model of insulin resistance and obesity, mimitin expression was down-regulated in liver and brain, up-regulated in heart and kidney, but not affected in islets. To further analyse the impact of mimitin on β-cell function, two β-cell lines, one with a low (INS1E) and another with a higher (MIN6) mimitin expression were studied. Mimitin overexpression protected INS1E cells against cytokine-induced caspase 3 activation, mitochondrial membrane potential reduction and ATP production inhibition, independently from the NF-κB (nuclear factor κB)-iNOS (inducible NO synthase) pathway. Mimitin overexpression increased basal and glucose-induced insulin secretion and prevented cytokine-mediated suppression of insulin secretion. Mimitin knockdown in MIN6 cells had opposite effects to those observed after overexpression. Thus mimitin has the capacity to modulate pancreatic islet function and to reduce cytokine toxicity.     

MQAPsingle: A quasi single‐model approach for estimation of the quality of individual protein structure models

We present a Model Quality Assessment Program (MQAP), called MQAPsingle, for ranking and assessing the absolute global quality of single protein models. MQAPsingle is quasi single‐model MQAP, a method that combines advantages of both “pure” single‐model MQAPs and clustering MQAPs. This approach results in higher accuracy compared to the state‐of‐the‐art single‐model MQAPs. Notably, the prediction for a given model is the same regardless if this model is submitted to our server alone or together with other models. Proteins 2016; 84:1021–1028. © 2015 Wiley Periodicals, Inc.     

Important role of matrix metalloproteinase 9 in epileptogenesis

Temporal lobe epilepsy (TLE) is a devastating disease in which aberrant synaptic plasticity plays a major role. We identify matrix metalloproteinase (MMP) 9 as a novel synaptic enzyme and a key pathogenic factor in two animal models of TLE: kainate-evoked epilepsy and pentylenetetrazole (PTZ) kindling-induced epilepsy. Notably, we show that the sensitivity to PTZ epileptogenesis is decreased in MMP-9 knockout mice but is increased in a novel line of transgenic rats overexpressing MMP-9. Immunoelectron microscopy reveals that MMP-9 associates with hippocampal dendritic spines bearing asymmetrical (excitatory) synapses, where both the MMP-9 protein levels and enzymatic activity become strongly increased upon seizures. Further, we find that MMP-9 deficiency diminishes seizure-evoked pruning of dendritic spines and decreases aberrant synaptogenesis after mossy fiber sprouting. The latter observation provides a possible mechanistic basis for the effect of MMP-9 on epileptogenesis. Our work suggests that a synaptic pool of MMP-9 is critical for the sequence of events that underlie the development of seizures in animal models of TLE.     

Human mitochondrial RNA turnover caught in flagranti: involvement of hSuv3p helicase in RNA surveillance

The mechanism of human mitochondrial RNA turnover and surveillance is still a matter of debate. We have obtained a cellular model for studying the role of hSuv3p helicase in human mitochondria. Expression of a dominant-negative mutant of the hSUV3 gene which encodes a protein with no ATPase or helicase activity results in perturbations of mtRNA metabolism and enables to study the processing and degradation intermediates which otherwise are difficult to detect because of their short half-lives. The hSuv3p activity was found to be necessary in the regulation of stability of mature, properly formed mRNAs and for removal of the noncoding processing intermediates transcribed from both H and L-strands, including mirror RNAs which represent antisense RNAs transcribed from the opposite DNA strand. Lack of hSuv3p function also resulted in accumulation of aberrant RNA species, molecules with extended poly(A) tails and degradation intermediates truncated predominantly at their 3′-ends. Moreover, we present data indicating that hSuv3p co-purifies with PNPase; this may suggest participation of both proteins in mtRNA metabolism.     

Expansion of the Protein Repertoire in Newly Explored Environments: Human Gut Microbiome Specific Protein Families

The microbes that inhabit particular environments must be able to perform molecular functions that provide them with a competitive advantage to thrive in those environments. As most molecular functions are performed by proteins and are conserved between related proteins, we can expect that organisms successful in a given environmental niche would contain protein families that are specific for functions that are important in that environment. For instance, the human gut is rich in polysaccharides from the diet or secreted by the host, and is dominated by Bacteroides, whose genomes contain highly expanded repertoire of protein families involved in carbohydrate metabolism. To identify other protein families that are specific to this environment, we investigated the distribution of protein families in the currently available human gut genomic and metagenomic data. Using an automated procedure, we identified a group of protein families strongly overrepresented in the human gut. These not only include many families described previously but also, interestingly, a large group of previously unrecognized protein families, which suggests that we still have much to discover about this environment. The identification and analysis of these families could provide us with new information about an environment critical to our health and well being.