Novel Heterogeneous Hydrogel with Dual-Responsive Shape Programmability and Good Biocompatibility.

Shape memory polymers (SMPs) responsive to various external stimuli can realize a complex shape transformation process and have attracted extensive attention. However, integrating multiple stimulus-responsive mechanisms in one material often requires a complex molecular design and synthesis procedure. In this work, we designed a novel dual-responsive heterogeneous hydrogel (PU-PAM/Alg/PDA), which was manufactured through in situ free radical polymerization of acrylamide (AM) in the presence of alginate (Alg) and polydopamine (PDA) in a porous polycaprolactone-based polyurethane foam (PU-foam). The PU-PAM/Alg/PDA hydrogel could achieve thermal responsiveness through melting-crystallization transformation of polycaprolactone (PCL), while the metallo-supramolecular interactions between Alg and Fe3+ could provide ion responsiveness for this hydrogel. This dual-programmable feature endowed the heterogeneous hydrogel with a complex shape-morphing behavior and also a reconfiguration ability for the permanent shape. Meanwhile, the strong hydrogen bondings between PDA and polyurethane chains enhanced the interfacial adhesions, resulting in the structural integrity and excellent mechanical property of PU-PAM/Alg/PDA. The in vitro and in vivo tests revealed the good biocompatibility of the heterogeneous hydrogel, and the potential of the heterogeneous hydrogel as an esophageal stent was evaluated in vitro as conceptual proof.

Polydopamine Nanoparticle-Reinforced Near-Infrared Light-Triggered Shape Memory Polycaprolactone-Polydopamine Polyurethane for Biomedical Implant Applications.

Near-infrared (NIR) light-triggered shape memory polymers are expected to have a more promising prospect in biomedical applications compared with traditional heat-triggered shape memory polymers. In this work, a new kind of polyurethane with NIR light-triggered shape memory property was prepared by using polycaprolactone (PCL), polydopamine nanoparticles (PDANPs), hexamethylene diisocyanate (HDI), and 1,4-butanediol (BDO). The synthesized PCL-PDA polyurethanes, especially when the weight content of PDANPs was 0.17%, showed excellent mechanical properties because the PDANPs were well-dispersed in polyurethanes by the chain extension reaction. Moreover, it also showed an NIR light-triggered rapid shape recovery because of the photothermal effect of polydopamine. The in vitro and in vivo tests showed that the PCL-PDA polyurethane would not inhibit cell proliferation nor induce a strong host inflammatory response, revealing the non-cytotoxicity and good biocompatibility of the material. In addition, the PCL-PDA polyurethane exhibited excellent in vivo NIR light-triggered shape memory performance under an 808 nm laser with low intensity (0.33 W cm-2), which was harmless to the human skin. These results demonstrated the potential of the PCL-PDA polyurethane in biomedical implant applications.

The influence of surface charge on the tumor-targeting behavior of Fe3O4 nanoparticles for MRI.

Nanomedicine-based tumor-targeted therapy has emerged as a promising strategy to overcome the lack of specificity of conventional chemotherapeutic agents. "Passive" targeting caused by the tumor EPR effect and "active" targeting endowed by the tumor-targeting moieties provide promising biomedical utilities and cancer therapy strategies for nanomedicine. However, as the nanoparticles are exposed to biological fluids, a large number of protein molecules will be adsorbed on their surface, known as protein corona, which may alter the targeting ability of the nanoparticles. The impact of different protein corona on the "passive" and "active" targeting behaviors is still ambiguous. Herein, three kinds of aqueous soluble Fe3O4 nanoparticles with different surface modifications were synthesized and applied to explore the correlation between their protein corona and passive/active tumor-targeting abilities. In the in vitro and in vivo studies, the protein corona exhibited completely different effects on the active and passive cancer-targeting capability of the particles. The particles presented active cancer-targeting ability if there was enough interaction time between the particles and cells. This was mainly due to the dynamic evolution of the protein corona, the proteins of which may be outcompeted by the cancer cell membrane and determine the targeting abilities. Unfortunately, the protein corona also inevitably accelerated RES/MPS uptake after the particles were injected into the body, which almost completely disabled the active targeting abilities of the particles. We believe that this in-depth understanding of protein corona will provide new ideas on the tumor-targeting mechanisms of nanoparticles and present a feasible approach to designing targeted drugs in the future.

GrantExtractor: Accurate Grant Support Information Extraction from Biomedical Fulltext Based on Bi-LSTM-CRF.

Grant support (GS) in the MEDLINE database refers to funding agencies and contract numbers. It is important for funding organizations to track their funding outcomes from the GS information. As such, how to accurately and automatically extract funding information from biomedical literature is challenging. In this paper, we present a pipeline system called GrantExtractor that is able to accurately extract GS information from fulltext biomedical literature. GrantExtractor effectively integrates several advanced machine learning techniques. In particular, we use a sentence classifier to identify funding sentences from articles first. A bi-directional LSTM and the CRF layer (BiLSTM-CRF), and pattern matching are then used to extract entities of grant numbers and agencies from these identified funding sentences. After removing noisy numbers by a multi-class model, we finally match each grant number with its corresponding agency. Experimental results on benchmark datasets have demonstrated that GrantExtractor clearly outperforms all baseline methods. It is further evident that GrantExtractor won the first place in Task 5C of 2017 BioASQ challenge, with achieving the Micro-recall of 0.9526 for 22,610 articles. Moreover, GrantExtractor has achieved the Micro F-measure score as high as 0.90 in extracting grant pairs.

Crystallization and Alkaline Degradation Behaviors of Poly(l-Lactide)/4-Armed Poly(ε-Caprolactone)-Block-Poly(d-Lactide) Blends with Different Poly(d-Lactide) Block Lengths.

Four-armed poly(ε-caprolactone)-block-poly(d-lactide) (4-C-D) copolymers with different poly(d-lactide) (PDLA) block lengths (Mn,PDLAs) were synthesized by sequential ring-opening polymerization (ROP). The formation of stereocomplex (SC) crystallites in the 80/20 poly(l-lactide) (PLLA)/4-C-D blends were investigated with the change of Mn,PDLA from 0.5 to 1.5 kg/mol. It was found that the crystallization and alkaline degradation of the blends were profoundly affected by the formed SC crystallites. The PLLA/4-C-D0.5 blend had the lowest crystallization rate of the three blends, and it was difficult to see spherulites in this blend by polarized optical microscopy (POM) observation after isothermal crystallization at 140 °C for 4 h. Meanwhile, when Mn,PDLA was 1 kg/mol or 1.5 kg/mol, SC crystallites could be formed in the PLLA/4-C-D blend and acted as nucleators for the crystallization of PLLA homo-crystals. However, the overall crystallization rates of the two blends were still lower than that of the neat PLLA. In the PLLA/4-C-D1.5 blend, the Raman results showed that small isolated SC spherulites were trapped inside the big PLLA homo-spherulites during isothermal crystallization. The degradation rate of the PLLA/4-C-D blend decreased when Mn,PDLA increased from 0.5 to 1.5 kg/mol, and the degradation morphologies had a close relationship with the crystallization state of the blends. This work revealed the gradual formation of SC crystallites with the increase in Mn,PDLA in the PLLA/4-C-D blends and its significant effect on the crystallization and degradation behaviors of the blend films.

FullMeSH: improving large-scale MeSH indexing with full text.

MOTIVATION: With the rapidly growing biomedical literature, automatically indexing biomedical articles by Medical Subject Heading (MeSH), namely MeSH indexing, has become increasingly important for facilitating hypothesis generation and knowledge discovery. Over the past years, many large-scale MeSH indexing approaches have been proposed, such as Medical Text Indexer, MeSHLabeler, DeepMeSH and MeSHProbeNet. However, the performance of these methods is hampered by using limited information, i.e. only the title and abstract of biomedical articles. RESULTS: We propose FullMeSH, a large-scale MeSH indexing method taking advantage of the recent increase in the availability of full text articles. Compared to DeepMeSH and other state-of-the-art methods, FullMeSH has three novelties: (i) Instead of using a full text as a whole, FullMeSH segments it into several sections with their normalized titles in order to distinguish their contributions to the overall performance. (ii) FullMeSH integrates the evidence from different sections in a 'learning to rank' framework by combining the sparse and deep semantic representations. (iii) FullMeSH trains an Attention-based Convolutional Neural Network for each section, which achieves better performance on infrequent MeSH headings. FullMeSH has been developed and empirically trained on the entire set of 1.4 million full-text articles in the PubMed Central Open Access subset. It achieved a Micro F-measure of 66.76% on a test set of 10 000 articles, which was 3.3% and 6.4% higher than DeepMeSH and MeSHLabeler, respectively. Furthermore, FullMeSH demonstrated an average improvement of 4.7% over DeepMeSH for indexing Check Tags, a set of most frequently indexed MeSH headings. AVAILABILITY AND IMPLEMENTATION: The software is available upon request. SUPPLEMENTARY INFORMATION: Supplementary data are available at Bioinformatics online.