The Study of Conditions for Bioleaching Heavy Metals from Polluted Soil and Metabonomics in Penicillium chrysogenum

The effects of the temperature, pH value and leaching time on the bioleaching ratio of heavy metals were studied using an orthogonal method in the paper. The metabonomics of Penicillium chrysogenum F1 were detected to deeply illustrate the bioleaching mechanism of Penicillium chrysogenumF1. The results showed that the bioleaching ratios of lead and zinc are the best at 30 °C on the 7th day, while those of cadmium and copper are the best at 20 °C on the 5th day. A pH value of 7.0 is best for heavy metal bioleaching ratios. The combination of 30 °C, pH 7.0 and 7 days is best for the total bioleaching ratios. The intensities of metabolites vary with the addition of heavy metal-polluted soil. The main catabolic pathways of glucose are not influenced by the heavy metal-polluted soil, though some metabolic enzymes are influenced, resulting in some metabolites undergoing upregulation or downregulation. Both organic acids and other metabolites containing functional groups (-NH2, -OH, -CHO, -CO) can extract heavy metals from soil.

Improved analysis of arabinoxylan-bound hydroxycinnamate conjugates in grass cell walls

Background Arabinoxylan in grass cell walls is acylated to varying extents by ferulate and p-coumarate at the 5-hydroxy position of arabinosyl residues branching off the xylan backbone. Some of these hydroxycinnamate units may then become involved in cell wall radical coupling reactions, resulting in ether and other linkages amongst themselves or to monolignols or oligolignols, thereby crosslinking arabinoxylan chains with each other and/or with lignin polymers. This crosslinking is assumed to increase the strength of the cell wall, and impedes the utilization of grass biomass in natural and industrial processes. A method for quantifying the degree of acylation in various grass tissues is, therefore, essential. We sought to reduce the incidence of hydroxycinnamate ester hydrolysis in our recently introduced method by utilizing more anhydrous conditions. Results The improved methanolysis method minimizes the undesirable ester-cleavage of arabinose from ferulate and p-coumarate esters, and from diferulate dehydrodimers, and produces more methanolysis vs. hydrolysis of xylan-arabinosides, improving the yields of the desired feruloylated and p-coumaroylated methyl arabinosides and their diferulate analogs. Free ferulate and p-coumarate produced by ester-cleavage were reduced by 78% and 68%, respectively, and 21% and 39% more feruloyl and p-coumaroyl methyl arabinosides were detected in the more anhydrous method. The new protocol resulted in an estimated 56% less combined diferulate isomers in which only one acylated arabinosyl unit remained, and 170% more combined diferulate isomers conjugated to two arabinosyl units. Conclusions Overall, the new protocol for mild acidolysis of grass cell walls is both recovering more ferulate- and p-coumarate-arabinose conjugates from the arabinoxylan and cleaving less of them down to free ferulic acid, p-coumaric acid, and dehydrodiferulates with just one arabinosyl ester. This cleaner method, especially when coupled with the orthogonal method for measuring monolignol hydroxycinnamate conjugates that have been incorporated into lignin, provides an enhanced tool to measure the extent of crosslinking in grass arabinoxylan chains, assisting in identification of useful grasses for biomass applications.

Unsupervised Scene Adaptation with Memory Regularization in vivo

This work focuses on the unsupervised scene adaptation problem of learning from both labeled source data and unlabeled target data. Existing approaches focus on minoring the inter-domain gap between the source and target domains. However, the intra-domain knowledge and inherent uncertainty learned by the network are under-explored. In this paper, we propose an orthogonal method, called memory regularization in vivo, to exploit the intra-domain knowledge and regularize the model training. Specifically, we refer to the segmentation model itself as the memory module, and minor the discrepancy of the two classifiers, i.e., the primary classifier and the auxiliary classifier, to reduce the prediction inconsistency. Without extra parameters, the proposed method is complementary to most existing domain adaptation methods and could generally improve the performance of existing methods. Albeit simple, we verify the effectiveness of memory regularization on two synthetic-to-real benchmarks: GTA5 → Cityscapes and SYNTHIA → Cityscapes, yielding +11.1% and +11.3% mIoU improvement over the baseline model, respectively. Besides, a similar +12.0% mIoU improvement is observed on the cross-city benchmark: Cityscapes → Oxford RobotCar.