Ubiquitin-specific protease 19 (USP19) is one of the deubiquitinating enzymes (DUBs) involved in regulating the ubiquitination status of substrate proteins. There are two major isoforms of USP19 with distinct C-termini; the USP19_a isoform has a transmembrane domain for anchoring to the endoplasmic reticulum, while USP19_b contains an EEVD motif. Here, we report that the cytoplasmic isoform USP19_b up-regulates the protein levels of the polyglutamine (polyQ)-containing proteins, ataxin-3 (Atx3) and huntingtin (Htt), and thus promotes aggregation of their polyQ-expanded species in cell models. Our data demonstrate that USP19_b may orchestrate the stability, aggregation and degradation of the polyQ-expanded proteins through the heat shock protein 90 (HSP90) chaperone system. USP19_b directly interacts with HSP90 through its N-terminal CS (CHORD and SGT1)/P23 domains. In conjunction with HSP90, the cytoplasmic USP19 may play a key role in triage decision for the disease-related polyQ-expanded substrates, suggesting a function of USP19 in quality control of misfolded proteins by regulating their protein levels.

Homo sapiens J domain protein (HSJ1) is a J-domain containing co-chaperone that is known to stimulate ATPase activity of HSP70 chaperone, while it also harbors two ubiquitin (Ub)-interacting motifs (UIMs) that may bind with ubiquitinated substrates and potentially function in protein degradation. We studied the effects of HSJ1a on the protein levels of both normal and the disease–related polyQ-expanded forms of ataxin-3 (Atx3) in cells. The results demonstrate that the N-terminal J-domain and the C-terminal UIM domain of HSJ1a exert opposite functions in regulating the protein level of cellular overexpressed Atx3. This dual regulation is dependent on the binding of the J-domain with HSP70, and the UIM domain with polyUb chains. The J-domain down-regulates the protein level of Atx3 through HSP70 mediated proteasomal degradation, while the UIM domain may alleviate this process via maintaining the ubiquitinated Atx3. We propose that co-chaperone HSJ1a orchestrates the balance of substrates in stressed cells in a Yin-Yang manner.

Fusion expression is a promising strategy for the production of bioactive peptides in Escherichia coli. In this study, we constructed a new recombinant expression plasmid containing the coding sequence of 56-residue B1 domain of streptococcal protein G (GB1). For easy purification and cleavage of the recombinant proteins, except GB1, an engineered hexahistidine and tobacco etch virus (TEV) protease recognition sites were included in the fusion sequence. Next, we cloned the coding sequence of human epidermal growth factor (hEGF) into this new plasmid and produced the recombinant hEGF in E. coli. The bioactive hEGF is a 53-amino acid peptide and is stabilized by three intramolecular disulphide bonds. Compared with glutathione Stransferase, thioredoxin and small ubiquitin-related modifier, GB1 greatly improved the expression and solubility of hEGF. Moreover, the recombinant hEGF bound to the nickel nitrilotriacetic acid resin column, was easily cleaved by TEV protease and the free hEGF was released. The results showed that this new plasmid was appropriate for recombinant production of small bioactive peptides, such as hEGF, which contains a high proportion of hydrophobic residues and intramolecular disulphide linkages.

Colorectal cancer (CRC) is ranked as the second most common cause of cancer deaths and the third most common cancer globally. It has been described as a ‘silent disease’ which is often easily treatable if detected early—before progression to carcinoma. Colonoscopy, which is the gold standard for diagnosis is not only expensive but is also an invasive diagnostic procedure, thus, effective and non‐invasive diagnostic methods are urgently needed. Unfortunately, the current methods are not sensitive and specific enough in detecting adenomas and early colorectal neoplasia, hampering treatment and consequently, survival rates. Studies have shown that imbalances in such a relationship which renders the gut microbiota in a dysbiotic state are implicated in the development of adenomas ultimately resulting in CRC. The differences found in the makeup and diversity of the gut microbiota of healthy individuals relative to CRC patients have in recent times gained attention as potential biomarkers in early non‐invasive diagnosis of CRC, with promising sensitivity, specificity and even cost‐effectiveness. This review summarizes recent studies in the application of these microbiota biomarkers in early CRC diagnosis, limitations encountered in the area of the faecal microbiota studies as biomarkers for CRC, and future research exploits that address these limitations.

MicroRNAs (miRNAs) are involved in a diverse variety of biological processes through regulating the expression of target genes in the post-transcriptional level. So, it is of great importance to discover the targets of miRNAs in biological research. But, due to the short length of miRNAs and limited sequence complementarity to their gene targets in animals, it is challenging to develop algorithms to predict the targets of miRNA accurately. Here we developed a new miRNA target prediction algorithm using a multilayer convolutional neural network. Our model learned automatically the interaction patterns of the experiment-validated miRNA:target-site chimeras from the raw sequence, avoiding hand-craft selection of features by domain experts. The performance on test dataset is inspiring, indicating great generalization ability of our model. Moreover, considering the stability of miRNA:target-site duplexes, our method also showed good performance to predict the target transcripts of miRNAs.

Due to the biogenesis difference, miRNAs can be divided into canonical microRNAs and mirtrons. Compared to canonical microRNAs, mirtrons are less conserved and hard to be identified. Except stringent annotations based on experiments, many in silico computational methods have be developed to classify miRNAs. Although several machine learning classifiers delivered high classification performance, all the predictors depended heavily on the selection of calculated features. Here, we introduced nucleotide-level convolutional neural networks (CNNs) for pre-miRNAs classification. By using “one-hot” encoding and padding, pre-miRNAs were converted into matrixes with the same shape. The convolution and max-pooling operations can automatically extract features from pre-miRNAs sequences. Evaluation on test dataset showed that our models had a satisfactory performance. Our investigation showed that it was feasible to apply CNNs to extract features from biological sequences. Since there are many hyperparameters can be tuned in CNNs, we believe that the performance of nucleotide-level convolutional neural networks can be greatly improved in the future.

No abstract

No abstract