Interpretable Multimodal Fusion Networks Reveal Mechanisms of Brain Cognition

The combination of multimodal imaging and genomics provides a more comprehensive way for the study of mental illnesses and brain functions. Deep network-based data fusion models have been developed to capture their complex associations, resulting in improved diagnosis of diseases. However, deep learning models are often difficult to interpret, bringing about challenges for uncovering biological mechanisms using these models. In this work, we develop an interpretable multimodal fusion model to perform automated diagnosis and result interpretation simultaneously. We name it Grad-CAM guided convolutional collaborative learning (gCAM-CCL), which is achieved by combining intermediate feature maps with gradient-based weights. The gCAM-CCL model can generate interpretable activation maps to quantify pixel-level contributions of the input features. Moreover, the estimated activation maps are class-specific, which can therefore facilitate the identification of biomarkers underlying different groups. We validate the gCAM-CCL model on a brain imaging-genetic study, and demonstrate its applications to both the classification of cognitive function groups and the discovery of underlying biological mechanisms. Specifically, our analysis results suggest that during task-fMRI scans, several object recognition related regions of interests (ROIs) are activated followed by several downstream encoding ROIs. In addition, the high cognitive group may have stronger neurotransmission signaling while the low cognitive group may have problems in brain/neuron development due to genetic variations.

Functional network estimation using multigraph learning with application to brain maturation study

Although most dramatic structural changes occur in the perinatal period, a growing body of evidences demonstrates that adolescence and early adulthood are also important for substantial neurodevelopment. We were thus motivated to explore brain development during puberty by evaluating functional connectivity network (FCN) differences between childhood and young adulthood using multi-paradigm task-based functional magnetic resonance imaging (fMRI) measurements. Different from conventional multigraph based FCN construction methods where the graph network was built independently for each modality/paradigm, we proposed a multigraph learning model in this work. It promises a better fitting to FCN construction by jointly estimating brain network from multi-paradigm fMRI time series, which may share common graph structures. To investigate the hub regions of the brain, we further conducted graph Fourier transform (GFT) to divide the fMRI BOLD time series of a node within the brain network into a range of frequencies. Then we identified the hub regions characterizing brain maturity through eigen-analysis of the low frequency components, which were believed to represent the organized structures shared by a large population. The proposed method was evaluated using both synthetic and real data, which demonstrated its effectiveness in extracting informative brain connectivity patterns. We detected 14 hub regions from the child group and 12 hub regions from the young adult group. We show the significance of these findings with a discussion of their functions and activation patterns as a function of age. In summary, our proposed method can extract brain connectivity network more accurately by considering the latent common structures between different fMRI paradigms, which are significant for both understanding brain development and recognizing population groups of different ages.

Recombinant lactobacillin PlnK adjusts the gut microbiome distribution in broilers

1. The heterologous expression and biological function of the Lactobacillus bacteriocin plantaricin K (PlnK) remain largely unknown. 2. In this study, PlnK was efficiently expressed in competent E. coli BL21 (used in transformation and protein expression) after 12 h, at 37°C and in 0.4 mmol/l isopropyl β- d-1-thiogalactopyranoside (IPTG). 3. The inhibitory bacterial spectrum of recombinant PlnK was investigated and indicated that levels of PlnK above 0.10 mg/ml produced an obvious inhibitory effect on gram-positive bacteria and gram-negative bacteria in vitro. 4. The effects of PlnK on intestinal immune function and the gut microbiome distribution in broilers were studied. The results revealed that, after consuming 2.50 × 10^-3 mg/ml of PlnK in water for one week, at the phylum level, the abundance of Firmicutes was increased and the abundance of Bacleroidetes was decreased. At the family level, the abundance of Lachnospiraceae, Ruminococcaceae and Streptococcaceae were significantly improved, but the abundance of Bacteroidaceae was reduced. At the genus level, the abundances of Lachnoclostridium, Streptococcus and Ruminococcaceae-UCG-013, were significantly up-regulated, and the abundance of Bacteroides was down-regulated. 5. After oral liquid intake of PlnK for one week, levels of secretory immunoglobulin A (sIgA) in the duodenal mucus were not significantly increased, but the mRNA levels of TLR3, MDA5, IFN-α, IFN-β, IFITM3 and IFITM10 in the duodenum were significantly reduced. 6. This study demonstrated that the recombinant PlnK could adjust the intestinal microbiome distribution and downregulate the IFN pathway.

Graph temporal ensembling based semi-supervised convolutional neural network with noisy labels for histopathology image analysis

Although convolutional neural networks have achieved tremendous success on histopathology image classification, they usually require large-scale clean annotated data and are sensitive to noisy labels. Unfortunately, labeling large-scale images is laborious, expensive and lowly reliable for pathologists. To address these problems, in this paper, we propose a novel self-ensembling based deep architecture to leverage the semantic information of annotated images and explore the information hidden in unlabeled data, and meanwhile being robust to noisy labels. Specifically, the proposed architecture first creates ensemble targets for feature and label predictions of training samples, by using exponential moving average (EMA) to aggregate feature and label predictions within multiple previous training epochs. Then, the ensemble targets within the same class are mapped into a cluster so that they are further enhanced. Next, a consistency cost is utilized to form consensus predictions under different configurations. Finally, we validate the proposed method with extensive experiments on lung and breast cancer datasets that contain thousands of images. It can achieve 90.5% and 89.5% image classification accuracy using only 20% labeled patients on the two datasets, respectively. This performance is comparable to that of the baseline method with all labeled patients. Experiments also demonstrate its robustness to small percentage of noisy labels.

Association between caregivers' anxiety and depression symptoms and feeding difficulties of preschool children: A cross-sectional study in rural China

PURPOSE

To investigate the prevalence of feeding difficulties in preschool children and explore the association between caregivers' anxiety and depression symptoms and preschool children's feeding difficulties.

METHODS

This cross-sectional study was conducted between June 2017 and January 2018 in rural areas of Anhui province, China. A total of 2231 preschool children and their caregivers were interviewed. Feeding difficulties of preschool children were reported by caregivers using the adapted Identification and Management of Feeding Difficulties (IMFeD) tool. Anxiety and depression symptoms of caregivers were evaluated via the Self-Rating Anxiety Scale (SAS) and the Self-Rating Depression Scale (SDS).

RESULTS

In total, 54.1% of preschool children were reported to have feeding difficulties by their caregivers. Among all children, when the caregivers had symptoms of anxiety or depression, the children had a higher risk of feeding difficulties. Specifically, for caregivers' anxiety symptoms, the odds ratios (ORs) of feeding difficulties in all children, left-behind children (LBC), and non-LBC were 1.91 (95% confidence interval [CI]: 1.42-2.57), 2.04 (95% CI: 1.34-3.09), and 1.86 (95% CI: 1.21-2.87), respectively; for caregivers' depression symptoms, the ORs of feeding difficulties in all children, LBC, and non-LBC were 1.86 (95% CI: 1.46-2.39), 1.76 (95% CI: 1.24-2.51), and 2.08 (95% CI: 1.45-2.97), respectively. In addition, when caregivers who were parents or grandparents had anxiety or depression symptoms, their children had a higher risk of feeding difficulties. Specifically, for parents and grandparents with anxiety symptoms, the ORs of feeding difficulties were 1.84 (95% CI: 1.14-2.98) and 2.17 (95% CI: 1.46-3.22), respectively; for parents and grandparents with depression symptoms, the ORs of feeding difficulties were 2.03 (95% CI: 1.40-2.95) and 1.93 (95% CI: 1.37-2.73), respectively.

CONCLUSION

Caregivers' anxiety or depression symptoms are positively associated with feeding difficulties in children.

Ultrastretchable and Wireless Bioelectronics Based on All-Hydrogel Microfluidics

Hydrogel bioelectronics that can interface biological tissues and flexible electronics is at the core of the growing field of healthcare monitoring, smart drug systems, and wearable and implantable devices. Here, a simple strategy is demonstrated to prototype all-hydrogel bioelectronics with embedded arbitrary conductive networks using tough hydrogels and liquid metal. Due to their excellent stretchability, the resultant all-hydrogel bioelectronics exhibits stable electrochemical properties at large tensile stretch and various modes of deformation. The potential of fabricated all-hydrogel bioelectronics is demonstrated as wearable strain sensors, cardiac patches, and near-field communication (NFC) devices for monitoring various physiological conditions wirelessly. The presented simple platform paves the way of implantable hydrogel electronics for Internet-of-Things and tissue-machine interfacing applications.

LncRNA WWOX‑AS1 inhibits the proliferation, migration and invasion of osteosarcoma cells

Recently, numerous long non-coding (lnc)RNAs have been revealed as serving important roles in human gene regulation. Previous studies have suggested that aberrant expression of lncRNAs is associated with cancer progression and metastasis. Previous studies have also demonstrated that decreased expression of WW domain-containing oxidoreductase (WWOX) is associated with poor prognosis in numerous cancer types. However, the effect of WWOX antisense RNA 1 (WWOX-AS1) in the development of cancer remains unknown. The aim of the present study was to investigate the role of WWOX-AS1 in osteosarcoma. The expression levels of WWOX-AS1 in human osteosarcoma cell lines and a normal osteoblastic cell line were investigated using reverse transcription-quantitative polymerase chain reaction (RT-qPCR). The results revealed that WWOX-AS1 expression was downregulated in osteosarcoma tissues. Furthermore, the association between WWOX-AS1 and the prognosis of patients with osteosarcoma was investigated using Kaplan-Meier and log-rank tests. The results suggested that patients exhibiting high WWOX-AS1 expression demonstrated a greater overall survival compared with patients exhibiting low WWOX-AS1 expression. In addition, overexpression and knockdown of WWOX-AS1 was performed using transfection experiments and confirmed by RT-qPCR in MG63 and SAOS2 cells, respectively. The results demonstrated that WWOX-AS1 and WWOX expression were positively correlated. Furthermore, the results of the knockdown and overexpression functional experiments suggested that WWOX-AS1 overexpression inhibited the proliferation, migration and invasion of MG63 cells, and knockdown of WWOX-AS1 enhanced the proliferation, migration and invasion of MG63 cells in SAOS2 cells. In conclusion, the results of the present study suggested that WWOX-AS1 may represent a potential biomarker and therapeutic target for the treatment of osteosarcoma.

Scalable and Automated Fabrication of Conductive Tough-Hydrogel Microfibers with Ultrastretchability, 3D Printability, and Stress Sensitivity

Creating complex three-dimensional structures from soft yet durable materials enables advances in fields such as flexible electronics, regenerating tissue engineering, and soft robotics. Tough hydrogels that mimic the human skin can bear enormous mechanical loads. By employing a spider-inspired biomimetic microfluidic nozzle, we successfully achieve continuous printing of tough hydrogels into fibers, two-dimensional networks, and even three-dimensional structures without compromising their extreme mechanical properties. The resultant thin fibers demonstrate a stretch up to 21 times of their original length at a water content of 52%, and are intrinsically transparent, biocompatible, and conductive at high stretches. Moreover, the printed robust tough-hydrogel networks can sense strain that are orders of magnitude lower than stretchable conductors by percolations of conductive particles. To demonstrate their potential application, we use printed tough-hydrogel fiber networks as wearable sensors for detecting human motions. The capability to shape tough hydrogels into complex structures by scalable continuous printing opens opportunities for new areas of applications such as tissue scaffolds, large-area soft electronics, and smart textiles.

4‑Phenylbutyrate protects rat skin flaps against ischemia‑reperfusion injury and apoptosis by inhibiting endoplasmic reticulum stress

4-phenylbutyrate (4-PBA) is a low molecular weight fatty acid, which has been demonstrated to regulate endoplasmic reticulum (ER) stress. ER stress-induced cell apoptosis has an important role in skin flap ischemia; however, a pharmacological approach for treating ischemia-induced ER dysfunction has yet to be reported. In the present study, the effects of 4-PBA-induced ER stress inhibition on ischemia-reperfusion injury were investigated in the skin flap of rats, and transcriptional regulation was examined. 4-PBA attenuated ischemia-reperfusion injury and inhibited cell apoptosis in the skin flap. Furthermore, 4-PBA reversed the increased expression levels of two ER stress markers: CCAAT/enhancer-binding protein-homologous protein and glucose-regulated protein 78. These results suggested that 4-PBA was able to protect rat skin flaps against ischemia-reperfusion injury and apoptosis by inhibiting ER stress marker expression and ER stress-mediated apoptosis. The beneficial effects of 4-PBA may prove useful in the treatment of skin flap ischemia-reperfusion injury.

Role of Glycol Chitosan-incorporated Ursolic Acid Nanoparticles in the Treatment of Osteosarcoma

Purpose: To investigate the effect of ursolic acid (UA)-incorporated glycol chitosan (GC) nanoparticles on inhibition of human osteosarcoma.Methods: U2OS and Saos-2 osteosarcoma cells were transfected with ursolic acid (UA) incorporated glycol chitosan (GC) nanoparticles. Ultraviolet (UV)  spectrophotometry was used to measure drug contents in nanoparticles at 365 nm with empty GC vehicles as blank. Bicinchoninic acid assay (BCA) method was employed to determine protein concentration. Identification of apoptosis and necrosis in osteosarcoma cells was performed by propidium iodide and FITC-annexin V reagents, respectively. FAC Scan flow cytometry was used to analyse apoptotic cells.Results: Among the range of UA concentrations tested, the minimum effective concentration was 10 μM with half inhibitory concentration IC50 of 25 μM. In U2OS cells, treatment with 10 and 25 μM UAinduced apoptosis in 5.89 ± 3.90 and 60.54 ± 5.40 % cells, respectively, compared to 2.05 ± 1.01 % cells for control. In Saos-2 cells, exposure to 10 and 25 μM UA induced apoptosis in 9.86 ± 8.89 and 47.54 ± 14.5 % cells, respectively, compared to 1.79 ± 0.23 % for control cells. Western blot analysis revealed  translocation of Bax and Bcl-2 proteins from mitochondria to cell cytosol. Increase in UA concentration  from 10 μM to 25 μM led to increase in the proportion of cells in G0/G1 phase and decrease in the number of cells in S and G2/M phases. These results confirm that UA transfection arrests cell cycle in G0/G1 phase in human osteosarcoma cell lines.Conclusion: UA transfection resulted in the inhibition of cell proliferation, Ezh2 expression inhibition, and apoptosis via mitochondrial pathway due to decrease in membrane potential and release of cytochrome C, as well as cell cycle arrest in G0/G1 phase.Keywords: Osteosarcoma, Cell cycle arrest, Palliation, Glycol chitosan, Ursolic acid

Increased expression of phospholipase Dα1 in guard cells decreases water loss with improved seed production under drought in Brassica napus

The activation of phospholipase Dα1 (PLDα1) produces lipid messenger phosphatidic acid and promotes stomatal closure in Arabidopsis. To explore the use of the PLDα1-mediated signalling towards decreasing water loss in crop plants, we introduced Arabidopsis PLDα1 under the control of a guard cell-specific promoter AtKatIpro into two canola (Brassica napus) cultivars. Multiple AtKatIpro ::PLDα1 lines in each cultivar displayed decreased water loss and improved biomass accumulation under hyperosmotic stress conditions, including drought and high salinity. Moreover, AtKatIpro ::PLDα1 plants produced more seeds than did WT plants in fields under drought. The results indicate that the guard cell-specific expression of PLDα1 has the potential to improve crop yield by enhancing drought tolerance.

Global identification of peptidase specificity by multiplex substrate profiling

We developed a simple and rapid multiplex substrate-profiling method to reveal the substrate specificity of any endo- or exopeptidase using liquid chromatography-tandem mass spectrometry sequencing. We generated a physicochemically diverse library of peptides by incorporating all combinations of neighbor and near-neighbor amino acid pairs into decapeptide sequences that are flanked by unique dipeptides at each terminus. Addition of a panel of evolutionarily diverse peptidases to a mixture of these tetradecapeptides generated information on prime and nonprime sites as well as on substrate specificity that matched or expanded upon known substrate motifs. This method biochemically confirmed the activity of the klassevirus 3C protein responsible for polypeptide processing and allowed granzyme B substrates to be ranked by enzymatic turnover efficiency using label-free quantitation of precursor-ion abundance. Additionally, the proteolytic secretions from schistosome parasitic flatworm larvae and a pancreatic cancer cell line were deconvoluted in a subtractive strategy using class-specific peptidase inhibitors.

Soil quality indicators response to application of hydrophilic polymers to a soil from a sulfide mine

In soils impacted by mining activities a vegetal cover is required to protect the site from the erosive forces of water and wind. The success of this objective depends on plant establishment and canopy closure. Polyacryalate polymers aid the growth of crops and indigenous plants in soils from sulfide mines. Soil characteristics change as a consequence of polymer application, but indicators that pinpoint these changes have not been identified yet. Our objectives were to (1) identify the sensitive indicators of changes in soil quality following polymer application, (2) relate these with assessment based on plant growth and soil cover. A mine soil was left unamended or received a characterized polyacrylate, a polyacrylate removed from diapers, or shredded diapers. Biomass of Spergularia purpurea was measured and proportion of soil cover evaluated. Soil enzymes, microbial activity, and respiration were analyzed. Availability of potentially toxic trace elements was estimated by their concentration in shoots. Factor analysis identified three factors that accounted for 94% of the variation in parameters, and the scores separated the four treatments. The indicators with greatest communality were correlated with plant growth and soil cover. The best soil quality indicators were As and Zn in shoots, protease, β-glucosidase, and fructose-induced respiration. It seems that the most important indicators to be used to assess the restoration of sulfide mine soils are those related with bioavailability of trace elements and soil enzymatic activities.

Use of hydrophilic polymers from diapers to aid the establishment of Spergularia purpurea in a mine soil

We used hydrophilic polymers from diapers to aid the establishment of an indigenous plant (Spergularia purpurea (Persoon) G. Don fil.) in a soil from a pyrite mine. Lysimeters were filled with the mine soil with no amendment (control), with a polyacrylate polymer, with a polymer removed from diapers, and with shredded diapers. The establishment of a plant cover was faster in soil amended with polymer from diapers, and 85 days after sowing the soil was completely covered in all treatments except control. The concentrations of trace elements in plant shoots decreased in amended soil. The activities of soil acid phosphatase, beta-glucosidase, protease and cellulase were greatest in soil amended with the polyacrylate polymer or with polymer removed from diapers, while the application of shredded diapers leads to values that were in general intermediate between these treatments and unamended control. Basal- and substrate-induced respirations, and dehydrogenase were greatest in soil amended with polymers, but the presence of a plastic film and fibrous materials from shredded diapers prevented any improvement in these parameters compared with unamended soil. In the second experiment, we evaluated the risk of downward movement of polymers in columns of a sandy soil. Polymer from diapers, with or without Cu, was placed at a 10 cm-depth. Five leaching cycles with artificial rain took place and leachates were analyzed for organic matter and Cu. At the end of the experiment, the soil columns were sliced and each layer was analyzed separately. Some repacking of soil and polymer particles took place, but there was no indication that polymers moved to any great depth in soil columns.

Use of insoluble polyacrylate polymers to aid phytostabilization of mine soils: effects on plant growth and soil characteristics

We evaluated the use of polyacrylate polymers to aid phytostabilization of mine soils. In a pot experiment, perennial ryegrass was grown in a mine soil and in uncontaminated soil. Growth was stimulated in the polymer-amended mine soil compared with an unamended control, and water-extractable levels of soil Cu and Zn decreased after polymer application. In an experiment performed in six 60-cm-diameter cylinders filled with fertilized mine soil, polymers were applied to three cylinders, with the remainder used as unamended control. Total biomass produced by indigenous plant species sown in polymer-amended soil was 1.8 (Spring-Summer) or 2.4 times (Fall-Winter) greater than that of plants from unamended soil. The application of polymers to the mine soil led to the greatest activity of soil enzymes. Soil pH, biomass of Spergularia purpurea and Chaetopogon fasciculatus, and activities of protease and cellulase had large loadings on principal component (PC)1, whereas growth of Briza maxima and the activities of urease, acid phosphatase, and beta-glucosidase had large loadings on PC2. The treatments corresponding to controls were located on the negative side of PC1 and PC2. Amended treatments were on the positive side of PC2 (Spring-Summer) or on the positive side of PC1 (Fall-Winter), demonstrating differential responses of plants and soil parameters in the two growth cycles.

Beta-sitosterol inhibits cell growth and induces apoptosis in SGC-7901 human stomach cancer cells

Beta-sitosterol is an important phytosterol found in plant food. It has been shown to have antiproliferative effects on cancers of the colon, breast, and prostate, but its effect on stomach cancer cells in vitro is unknown. Proliferation, cytotoxicity, and apoptosis in SGC-7901 human stomach cancer cells were examined by 3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyltetrazolium bromide (MTT) assay, clone formation, lactate dehydrogenase (LDH) leakage assay, acridine orange (AO)/ethidium bromide (EB) double staining, 4',6-diamidine-2'-phenylindole dihydrochloride (DAPI) staining, comet assay, and Western blotting. The results showed that beta-sitosterol suppresses the proliferation and induces the cell cytotoxicity of SGC-7901 stomach cancer cells in a time- and dose-dependent manner. Cells treated with different concentrations of beta-sitosterol also showed changes typical of apoptosis: morphological changes, DNA damage, increased expression of pro-caspase-3 and bax (p < 0.05), and activation of pro-caspase-3 and suppression of bcl-2 expression (p < 0.05). This study therefore revealed that beta-sitosterol significantly inhibits the growth and induces the apoptosis of SGC-7901 human stomach cancer cells in vitro. The decrease of the bcl-2/bax ratio and DNA damage may be the critical mechanisms of apoptosis induced by beta-sitosterol in SGC-7901 human stomach cancer cells.

The osteogenic potential of pseudoarthrosis tissue and bone from human scaphoid non-unions

Scaphoid fractures have the highest prevalence of non-union in the human body, but little is known about the osteogenic potential of cells at the pseudoarthrosis. It was our goal to determine whether cells isolated from non-unions could be stimulated to differentiate into osteoblasts and produce bone in vitro. Fifteen human scaphoid non-unions were excised during surgery and bone from either side of the non-union and the fibrocartilagenous central regions were harvested. Osteoblastic populations were subcultured from these. The number of bone nodules (colonies of osteoblast cells that produced bone) from all three regions was similar to the number of nodules derived from iliac bone cultures from the same patients. Treatment of cells with rhBMP-2 resulted in a 3- to 10-fold increase in bone nodule formation in vitro from cells derived from the non-unions. These data demonstrate that cells at the pseudoarthrosis have osteogenic capability and can be stimulated by rhBMP-2, possibly increasing the ability to heal.