Machine learning approaches for biomarker discovery to predict large-artery atherosclerosis.

Large-artery atherosclerosis (LAA) is a leading cause of cerebrovascular disease. However, LAA diagnosis is costly and needs professional identification. Many metabolites have been identified as biomarkers of specific traits. However, there are inconsistent findings regarding suitable biomarkers for the prediction of LAA. In this study, we propose a new method integrates multiple machine learning algorithms and feature selection method to handle multidimensional data. Among the six machine learning models, logistic regression (LR) model exhibited the best prediction performance. The value of area under the receiver operating characteristic curve (AUC) was 0.92 when 62 features were incorporated in the external validation set for the LR model. In this model, LAA could be well predicted by clinical risk factors including body mass index, smoking, and medications for controlling diabetes, hypertension, and hyperlipidemia as well as metabolites involved in aminoacyl-tRNA biosynthesis and lipid metabolism. In addition, we found that 27 features were present among the five adopted models that could provide good results. If these 27 features were used in the LR model, an AUC value of 0.93 could be achieved. Our study has demonstrated the effectiveness of combining machine learning algorithms with recursive feature elimination and cross-validation methods for biomarker identification. Moreover, we have shown that using shared features can yield more reliable correlations than either model, which can be valuable for future identification of LAA.

A plasmid-free Zymomonas mobilis mutant strain reducing reactive oxygen species for efficient bioethanol production using industrial effluent of xylose mother liquor.

Genome minimization is an effective way for industrial chassis development. In this study, Zymomonas mobilis ZMNP, a plasmid-free mutant strain of Z. mobilis ZM4 with four native plasmids deleted, was constructed using native type I-F CRISPR-Cas system. Cell growth of ZMNP under different temperatures and industrial effluent of xylose mother liquor were examined to investigate the impact of native plasmid removal. Despite ZMNP grew similarly as ZM4 under different temperatures, ZMNP had better xylose mother liquor utilization than ZM4. In addition, genomic, transcriptomic, and proteomic analyses were applied to unravel the molecular changes between ZM4 and ZMNP. Whole-genome resequencing result indicated that an S267P mutation in the C-terminal of OxyR, a peroxide-sensing transcriptional regulator, probably alters the transcription initiation of antioxidant genes for stress responses. Transcriptomic and proteomic studies illustrated that the reason that ZMNP utilized the toxic xylose mother liquor better than ZM4 was probably due to the upregulation of genes in ZMNP involving in stress responses as well as cysteine biosynthesis to accelerate the intracellular ROS detoxification and nucleic acid damage repair. This was further confirmed by lower ROS levels in ZMNP compared to ZM4 in different media supplemented with furfural or ethanol. The upregulation of stress response genes due to the OxyR mutation to accelerate ROS detoxification and DNA/RNA repair not only illustrates the underlying mechanism of the robustness of ZMNP in the toxic xylose mother liquor, but also provides an idea for the rational design of synthetic inhibitor-tolerant microorganisms for economic lignocellulosic biochemical production.

Molecular mechanism of enhanced ethanol tolerance associated with hfq overexpression in Zymomonas mobilis.

Zymomonas mobilis is a promising microorganism for industrial bioethanol production. However, ethanol produced during fermentation is toxic to Z. mobilis and affects its growth and bioethanol production. Although several reports demonstrated that the RNA-binding protein Hfq in Z. mobilis contributes to the tolerance against multiple lignocellulosic hydrolysate inhibitors, the role of Hfq on ethanol tolerance has not been investigated. In this study, hfq in Z. mobilis was either deleted or overexpressed and their effects on cell growth and ethanol tolerance were examined. Our results demonstrated that hfq overexpression improved ethanol tolerance of Z. mobilis, which is probably due to energy saving by downregulating flagellar biosynthesis and heat stress response proteins, as well as reducing the reactive oxygen species induced by ethanol stress via upregulating the sulfate assimilation and cysteine biosynthesis. To explore proteins potentially interacted with Hfq, the TEV protease mediated Yeast Endoplasmic Reticulum Sequestration Screening system (YESS) was established in Z. mobilis. YESS results suggested that Hfq may modulate the cytoplasmic heat shock response by interacting with the heat shock proteins DnaK and DnaJ to deal with the ethanol inhibition. This study thus not only revealed the underlying mechanism of enhanced ethanol tolerance by hfq overexpression, but also provided an alternative approach to investigate protein-protein interactions in Z. mobilis.

Identification and Characterization of a Trillin Rhamnosyltransferase From Dioscorea zingiberensis.

Dioscorea zingiberensis accumulates abundant steroidal saponins, such as dioscin, which is the principal bioactive ingredient displaying a wide range of pharmacological activities. Diosgenin is the aglycone of dioscin, and recently, genes encoding cytochrome P450 enzymes in the late steps of diosgenin biosynthesis have been isolated. Diosgenin was successfully synthesized in the cholesterol-producing yeasts. From diosgenin to dioscin, one glucose and two rhamnose groups need to be added. Although genes encoding UDP-glucosyltransferases converting diosgenin to trillin were isolated, genes encoding UDP-rhamnosyltransferases involved in dioscin biosynthesis remain unknown. In this study, we isolated the cDNA encoding the trillin rhamnosyltransferase (designated DzGT1) from D. zingiberensis. Heterologous expression of DzGT1 in Escherichia coli cells showed that the gene product exhibits an enzyme activity that glycosylates the trillin to form prosapogenin A of dioscin (PSA). The transcript level of DzGT1 is in accord with PSA accumulation in different organs of D. zingiberensis. Integration of the biochemical, metabolic, and transcriptional data supported the function of DzGT1 in dioscin biosynthesis. The identification and characterization of DzGT1 will help understand the metabolism of steroidal saponins in D. zingiberensis and provide candidate UDP-rhamnosyltransferase for efficient production of PSA, dioscin, and relevant steroidal saponins in microbial hosts.