Biology subjects

Du, D.

Publications and source records attributed to Du, D..

23 records · Page 2Linked to original sources

bioRxiv2021

Structure and efflux mechanism of the yeast pleiotropic drug resistance transporter Pdr5

Pdr5, a member of the extensive ABC transporter superfamily, is representative of a clinically relevant subgroup involved in pleiotropic drug resistance. Through the coupling of nucleotide hydrolysis with drug efflux, Pdr5 homologues enable pathogenic species to survive in the presence of chemically diverse antifungal agents. Our structural and functional results reveal details of an ATP-driven conformational cycle, which mechanically drives drug translocation through an amphipathic channel, and a clamping switch within a conserved linker loop that acts as a nucleotide sensor. One half of the transporter remains nearly invariant throughout the cycle, while its partner undergoes changes that are transmitted across interdomain interfaces to support a peristaltic motion of the pumped molecule. The efflux model proposed here rationalises the pleiotropic impact of Pdr5 and opens avenues for the development of effective antifungal compounds.

microbiology↗

bioRxiv2019

Mapping biology from mouse to man using transfer learning

Biomedical research often involves conducting experiments on model organisms in the anticipation that the biology learnt from these experiments will transfer to the human. Yet, it is commonly the case that biology does not transfer effectively, often for unknown reasons. Despite its importance to translational research this transfer process is not currently rigorously quantified. Here, we show that transfer learning - the branch of machine learning that concerns passing information from one domain to another - can be used to efficiently map biology from mouse to man, using the bone marrow (BM) as a representative example of a complex tissue. We first trained an artificial neural network (ANN) to accurately recognize various different cell types in mouse BM using data obtained from single-cell RNA-sequencing (scRNA-Seq) experiments. We found that this ANN, trained exclusively on mouse data, was able to identify individual human cells obtained from comparable scRNA-Seq experiments of human BM with 83% overall accuracy. However, while some human cell types were easily identified, others were not, indicating important differences in biology. To obtain a more accurate map of the human BM we then retrained the mouse ANN using scRNA-Seq data from a limited sample of human BM cells. Typically, less than 10 human cells of a given type were needed to accurately learn its representation in the updated model. In some cases, human cell identities could be inferred directly from the mouse ANN without retraining, via a process of biologically-guided zero-shot learning. These results show how machine learning can be used to reconstruct complex biology from limited data and have broad implications for biomedical research.

systems biology↗

bioRxiv2019

Interactions of a bacterial RND transporter with a transmembrane small protein in a lipid environment

The small protein AcrZ in Escherichia coli interacts with the transmembrane portion of the multidrug efflux pump AcrB and increases the resistance of the bacterium to a subset of the antibiotic substrates of that transporter. It is not clear how the physical association of the two proteins selectively changes activity of the pump for defined substrates. Here, we report cryo-EM structures of AcrB and the AcrBZ complex in lipid environments, and comparisons suggest that conformational changes occur in the drug binding pocket as a result of AcrZ binding. Simulations indicate that cardiolipin preferentially interacts with the AcrBZ complex, due to increased contact surface, and we observe that the drug sensitivity of bacteria lacking AcrZ is exacerbated when combined with cardiolipin deficiency. Taken together, the data suggest that AcrZ and lipid cooperate to allosterically modulate the activity of AcrB. This mode of regulation by a small protein and lipid may occur for other membrane proteins.

microbiology↗

bioRxiv2018

Generating a sequence map of the pig pan-genome

Pigs (Sus scrofa) exhibit diverse phenotypes in different breeds shaped by the combined effects of various local adaptation and artificial selection. To comprehensively characterize the genetic diversity of pigs, we construct a pig pan-genome by comparing genome assemblies of 11 representative pig breeds with the reference genome (Sscrofa11.1). Approximately 72.5 Mb non-redundant sequences were identified as pan-sequences which were absent from the Sscrofa11.1. On average, 41.7 kb of spurious heterozygous SNPs per individual are removed and 12.9 kb novel SNPs per individual are recovered using pan-genome as the reference for SNP calling, thereby providing enhanced resolution for genetic diversity in pigs. Homolog annotation and analysis using RNA-seq and Hi-C data indicate that these pan-sequences contain protein-coding regions and regulatory elements. These pan-sequences can further improve the interpretation of local 3D structure. The pan-genome as well as the accompanied web-based database will serve as a primary resource for exploration of genetic diversity and promote pig breeding and biomedical research.

genomics↗

bioRxiv2018

Construction of high-density genetic maps and detection of QTLs associated with Huanglongbing infection in citrus

No true resistance to Huanglongbing (HLB), a citrus disease associated with infection of Candidatus Liberibacter asiaticus (CLas), is found within commercial citrus cultivars, though trifoliate orange (Poncirus trifoliata) has been described as resistant or tolerant. Through genotyping an intergeneric F1 population by Genotyping-by-Sequencing, high-density SNP-based genetic maps were constructed separately for trifoliate orange and sweet orange (Citrus sinensis). Both genetic maps exhibited high synteny and high coverage of citrus genome. After exposure to intense HLB pressure for two years, Ct value of qPCR for CLas detection in leaves throughout ten time points during the next three years was above 35 in trifoliate oranges, under 28 in sweet oranges, and ranged from 24 to 38 and exhibited obvious segregation among progenies. Phenotypic data of percentage of healthy trees showed high correlation with the Ct value. By mapping the two traits at all time points, a total of nine clusters of QTLs were detected, of which five, respectively located on LG-t7 and LG-t8 of trifoliate orange map and LG-s3, LG-s5 and LG-s9 of sweet orange map, collectively explained a major part of the phenotypic variation. This study provides a starting point for citrus breeding to support long-term control of this devastating disease.\n\nHighlight1). Constructed the first high-density genetic map for trifoliate orange (Poncirus trifoliata)\n\n2). The first report on identification of QTLs related to Huanglongbing in citrus.\n\nAbbreviations

genetics↗