Developing an in silico minimum inhibitory concentration panel test for Klebsiella pneumoniae

Antimicrobial resistant infections are a serious public health threat worldwide. Whole genome sequencing approaches to rapidly identify pathogens and predict antibiotic resistance phenotypes are becoming more feasible and may offer a way to reduce clinical test turnaround times compared to conventional culture-based methods, and in turn, improve patient outcomes. In this study, we use whole genome sequence data from 1668 clinical isolates of Klebsiella pneumoniae to develop a XGBoost-based machine learning model that accurately predicts minimum inhibitory concentrations (MICs) for 20 antibiotics. The overall accuracy of the model, within ±1 two-fold dilution factor, is 92%. Individual accuracies are ≥90% for 15/20 antibiotics. We show that the MICs predicted by the model correlate with known antimicrobial resistance genes. Importantly, the genome-wide approach described in this study offers a way to predict MICs for isolates without knowledge of the underlying gene content. This study shows that machine learning can be used to build a complete in silico MIC prediction panel for K. pneumoniae and provides a framework for building MIC prediction models for other pathogenic bacteria.

Improvements to PATRIC, the all-bacterial Bioinformatics Database and Analysis Resource Center

The Pathosystems Resource Integration Center (PATRIC) is the bacterial Bioinformatics Resource Center (https://www.patricbrc.org). Recent changes to PATRIC include a redesign of the web interface and some new services that provide users with a platform that takes them from raw reads to an integrated analysis experience. The redesigned interface allows researchers direct access to tools and data, and the emphasis has changed to user-created genome-groups, with detailed summaries and views of the data that researchers have selected. Perhaps the biggest change has been the enhanced capability for researchers to analyze their private data and compare it to the available public data. Researchers can assemble their raw sequence reads and annotate the contigs using RASTtk. PATRIC also provides services for RNA-Seq, variation, model reconstruction and differential expression analysis, all delivered through an updated private workspace. Private data can be compared by ‘virtual integration’ to any of PATRIC's public data. The number of genomes available for comparison in PATRIC has expanded to over 80 000, with a special emphasis on genomes with antimicrobial resistance data. PATRIC uses this data to improve both subsystem annotation and k-mer classification, and tags new genomes as having signatures that indicate susceptibility or resistance to specific antibiotics.

Computing and Applying Atomic Regulons to Understand Gene Expression and Regulation

Understanding gene function and regulation is essential for the interpretation, prediction, and ultimate design of cell responses to changes in the environment. An important step toward meeting the challenge of understanding gene function and regulation is the identification of sets of genes that are always co-expressed. These gene sets, Atomic Regulons (ARs), represent fundamental units of function within a cell and could be used to associate genes of unknown function with cellular processes and to enable rational genetic engineering of cellular systems. Here, we describe an approach for inferring ARs that leverages large-scale expression data sets, gene context, and functional relationships among genes. We computed ARs for Escherichia coli based on 907 gene expression experiments and compared our results with gene clusters produced by two prevalent data-driven methods: Hierarchical clustering and k-means clustering. We compared ARs and purely data-driven gene clusters to the curated set of regulatory interactions for E. coli found in RegulonDB, showing that ARs are more consistent with gold standard regulons than are data-driven gene clusters. We further examined the consistency of ARs and data-driven gene clusters in the context of gene interactions predicted by Context Likelihood of Relatedness (CLR) analysis, finding that the ARs show better agreement with CLR predicted interactions. We determined the impact of increasing amounts of expression data on AR construction and find that while more data improve ARs, it is not necessary to use the full set of gene expression experiments available for E. coli to produce high quality ARs. In order to explore the conservation of co-regulated gene sets across different organisms, we computed ARs for Shewanella oneidensis, Pseudomonas aeruginosa, Thermus thermophilus, and Staphylococcus aureus, each of which represents increasing degrees of phylogenetic distance from E. coli. Comparison of the organism-specific ARs showed that the consistency of AR gene membership correlates with phylogenetic distance, but there is clear variability in the regulatory networks of closely related organisms. As large scale expression data sets become increasingly common for model and non-model organisms, comparative analyses of atomic regulons will provide valuable insights into fundamental regulatory modules used across the bacterial domain.

Theranostic Application of Mixed Gold and Superparamagnetic Iron Oxide Nanoparticle Micelles in Glioblastoma Multiforme

The treatment of glioblastoma multiforme, the most prevalent and lethal form of brain cancer in humans, has been limited in part by poor delivery of drugs through the blood-brain barrier and by unclear delineation of the extent of infiltrating tumor margins. Nanoparticles, which selectively accumulate in tumor tissue due to their leaky vasculature and the enhanced permeability and retention effect, have shown promise as both therapeutic and diagnostic agents for brain tumors. In particular, superparamagnetic iron oxide nanoparticles (SPIONs) have been leveraged as T2-weighted MRI contrast agents for tumor detection and imaging; and gold nanoparticles (AuNP) have been demonstrated as radiosensitizers capable of propagating electron and free radical-induced radiation damage to tumor cells. In this study, we investigated the potential applications of novel gold and SPION-loaded micelles (GSMs) coated by polyethylene glycol-polycaprolactone (PEG-PCL) polymer. By quantifying gh2ax DNA damage foci in glioblastoma cell lines, we tested the radiosensitizing efficacy of these GSMs, and found that GSM administration in conjunction with radiation therapy (RT) led to ~2-fold increase in density of double-stranded DNA breaks. For imaging, we used GSMs as a contrast agent for both computed tomography (CT) and magnetic resonance imaging (MRI) studies of stereotactically implanted GBM tumors in a mouse model, and found that MRI but not CT was sufficiently sensitive to detect and delineate tumor borders after administration and accumulation of GSMs. These results suggest that with further development and testing, GSMs may potentially be integrated into both imaging and treatment of brain tumors, serving a theranostic purpose as both an MRI-based contrast agent and a radiosensitizer.

PATtyFams: Protein Families for the Microbial Genomes in the PATRIC Database

The ability to build accurate protein families is a fundamental operation in bioinformatics that influences comparative analyses, genome annotation, and metabolic modeling. For several years we have been maintaining protein families for all microbial genomes in the PATRIC database (Pathosystems Resource Integration Center, patricbrc.org) in order to drive many of the comparative analysis tools that are available through the PATRIC website. However, due to the burgeoning number of genomes, traditional approaches for generating protein families are becoming prohibitive. In this report, we describe a new approach for generating protein families, which we call PATtyFams. This method uses the k-mer-based function assignments available through RAST (Rapid Annotation using Subsystem Technology) to rapidly guide family formation, and then differentiates the function-based groups into families using a Markov Cluster algorithm (MCL). This new approach for generating protein families is rapid, scalable and has properties that are consistent with alignment-based methods.

RASTtk: a modular and extensible implementation of the RAST algorithm for building custom annotation pipelines and annotating batches of genomes

The RAST (Rapid Annotation using Subsystem Technology) annotation engine was built in 2008 to annotate bacterial and archaeal genomes. It works by offering a standard software pipeline for identifying genomic features (i.e., protein-encoding genes and RNA) and annotating their functions. Recently, in order to make RAST a more useful research tool and to keep pace with advancements in bioinformatics, it has become desirable to build a version of RAST that is both customizable and extensible. In this paper, we describe the RAST tool kit (RASTtk), a modular version of RAST that enables researchers to build custom annotation pipelines. RASTtk offers a choice of software for identifying and annotating genomic features as well as the ability to add custom features to an annotation job. RASTtk also accommodates the batch submission of genomes and the ability to customize annotation protocols for batch submissions. This is the first major software restructuring of RAST since its inception.

Enabling comparative modeling of closely related genomes: example genus Brucella

For many scientific applications, it is highly desirable to be able to compare metabolic models of closely related genomes. In this short report, we attempt to raise awareness to the fact that taking annotated genomes from public repositories and using them for metabolic model reconstructions is far from being trivial due to annotation inconsistencies. We are proposing a protocol for comparative analysis of metabolic models on closely related genomes, using fifteen strains of genus Brucella, which contains pathogens of both humans and livestock. This study lead to the identification and subsequent correction of inconsistent annotations in the SEED database, as well as the identification of 31 biochemical reactions that are common to Brucella, which are not originally identified by automated metabolic reconstructions. We are currently implementing this protocol for improving automated annotations within the SEED database and these improvements have been propagated into PATRIC, Model-SEED, KBase and RAST. This method is an enabling step for the future creation of consistent annotation systems and high-quality model reconstructions that will support in predicting accurate phenotypes such as pathogenicity, media requirements or type of respiration.

In search of genome annotation consistency: solid gene clusters and how to use them

Maintaining consistency in genome annotations is important for supporting many computational tasks, particularly metabolic modeling. The SEED project has implemented a process that improves annotation consistencies across microbial genomes for proteins with conserved sequences and genomic context. In this research report, we describe this process and show how this effort has resulted in improvements to microbial genome annotations in the SEED. We also compare SEED annotation consistencies with other commonly used resources such as IMG (the Joint Genome Institute’s Integrated Microbial Genomes system), RefSeq (the National Center for Biotechnology Information’s Reference Sequence Database), Swiss-Prot (the annotated protein sequence database of the Swiss Institute of Bioinformatics, European Molecular Biology Laboratory and the European Bioinformatics Institute) and TrEMBL (Translated European Molecular Biology Laboratory nucleotide sequence data Library). Our analysis indicates that manual and computational efforts are paying off for the databases where consistency is a major goal.

The SEED and the Rapid Annotation of microbial genomes using Subsystems Technology (RAST)

In 2004, the SEED (http://pubseed.theseed.org/) was created to provide consistent and accurate genome annotations across thousands of genomes and as a platform for discovering and developing de novo annotations. The SEED is a constantly updated integration of genomic data with a genome database, web front end, API and server scripts. It is used by many scientists for predicting gene functions and discovering new pathways. In addition to being a powerful database for bioinformatics research, the SEED also houses subsystems (collections of functionally related protein families) and their derived FIGfams (protein families), which represent the core of the RAST annotation engine (http://rast.nmpdr.org/). When a new genome is submitted to RAST, genes are called and their annotations are made by comparison to the FIGfam collection. If the genome is made public, it is then housed within the SEED and its proteins populate the FIGfam collection. This annotation cycle has proven to be a robust and scalable solution to the problem of annotating the exponentially increasing number of genomes. To date, >12 000 users worldwide have annotated >60 000 distinct genomes using RAST. Here we describe the interconnectedness of the SEED database and RAST, the RAST annotation pipeline and updates to both resources.

Selective targeting of brain tumors with gold nanoparticle-induced radiosensitization

Successful treatment of brain tumors such as glioblastoma multiforme (GBM) is limited in large part by the cumulative dose of Radiation Therapy (RT) that can be safely given and the blood-brain barrier (BBB), which limits the delivery of systemic anticancer agents into tumor tissue. Consequently, the overall prognosis remains grim. Herein, we report our pilot studies in cell culture experiments and in an animal model of GBM in which RT is complemented by PEGylated-gold nanoparticles (GNPs). GNPs significantly increased cellular DNA damage inflicted by ionizing radiation in human GBM-derived cell lines and resulted in reduced clonogenic survival (with dose-enhancement ratio of ∼1.3). Intriguingly, combined GNP and RT also resulted in markedly increased DNA damage to brain blood vessels. Follow-up in vitro experiments confirmed that the combination of GNP and RT resulted in considerably increased DNA damage in brain-derived endothelial cells. Finally, the combination of GNP and RT increased survival of mice with orthotopic GBM tumors. Prior treatment of mice with brain tumors resulted in increased extravasation and in-tumor deposition of GNP, suggesting that RT-induced BBB disruption can be leveraged to improve the tumor-tissue targeting of GNP and thus further optimize the radiosensitization of brain tumors by GNP. These exciting results together suggest that GNP may be usefully integrated into the RT treatment of brain tumors, with potential benefits resulting from increased tumor cell radiosensitization to preferential targeting of tumor-associated vasculature.

Genomes of the class Erysipelotrichia clarify the firmicute origin of the class Mollicutes

The tree of life is paramount for achieving an integrated understanding of microbial evolution and the relationships between physiology, genealogy and genomics. It provides the framework for interpreting environmental sequence data, whether applied to microbial ecology or to human health. However, there remain many instances where there is ambiguity in our understanding of the phylogeny of major lineages, and/or confounding nomenclature. Here we apply recent genomic sequence data to examine the evolutionary history of members of the classes Mollicutes (phylum Tenericutes) and Erysipelotrichia (phylum Firmicutes). Consistent with previous analyses, we find evidence of a specific relationship between them in molecular phylogenies and signatures of the 16S rRNA, 23S rRNA, ribosomal proteins and aminoacyl-tRNA synthetase proteins. Furthermore, by mapping functions over the phylogenetic tree we find that the erysipelotrichia lineages are involved in various stages of genomic reduction, having lost (often repeatedly) a variety of metabolic functions and the ability to form endospores. Although molecular phylogeny has driven numerous taxonomic revisions, we find it puzzling that the most recent taxonomic revision of the phyla Firmicutes and Tenericutes has further separated them into distinct phyla, rather than reflecting their common roots.

Characterizing the native codon usages of a genome: an axis projection approach

Codon usage can provide insights into the nature of the genes in a genome. Genes that are “native” to a genome (have not been recently acquired by horizontal transfer) range in codon usage from a low-bias “typical” usage to a more biased “high-expression” usage characteristic of genes encoding abundant proteins. Genes that differ from these native codon usages are candidates for foreign genes that have been recently acquired by horizontal gene transfer. In this study, we present a method for characterizing the codon usages of native genes—both typical and highly expressed—within a genome. Each gene is evaluated relative to a half line (or axis) in a 59D space of codon usage. The axis begins at the modal codon usage, the usage that matches the largest number of genes in the genome, and it passes through a point representing the codon usage of a set of genes with expression-related bias. A gene whose codon usage matches (does not significantly differ from) a point on this axis is a candidate native gene, and the location of its projection onto the axis provides a general estimate of its expression level. A gene that differs significantly from all points on the axis is a candidate foreign gene. This automated approach offers significant improvements over existing methods. We illustrate this by analyzing the genomes of Pseudomonas aeruginosa PAO1 and Bacillus anthracis A0248, which can be difficult to analyze with commonly used methods due to their biased base compositions. Finally, we use this approach to measure the proportion of candidate foreign genes in 923 bacterial and archaeal genomes. The organisms with the most homogeneous genomes (containing the fewest candidate foreign genes) are mostly endosymbionts and parasites, though with exceptions that include Pelagibacter ubique and Beutenbergia cavernae. The organisms with the most heterogeneous genomes (containing the most candidate foreign genes) include members of the genera Bacteroides, Corynebacterium, Desulfotalea, Neisseria, Xylella, and Thermobaculum.

Modal codon usage: assessing the typical codon usage of a genome

Most genomes are heterogeneous in codon usage, so a codon usage study should start by defining the codon usage that is typical to the genome. Although this is commonly taken to be the genomewide average, we propose that the mode-the codon usage that matches the most genes-provides a more useful approximation of the typical codon usage of a genome. We provide a method for estimating the modal codon usage, which utilizes a continuous approximation to the number of matching genes and a simplex optimization. In a survey of bacterial and archaeal genomes, as many as 20% more of the genes in a given genome match the modal codon usage than the average codon usage. We use the mode to examine the evolution of the multireplicon genomes of Agrobacterium tumefaciens C58 and Borrelia burgdorferi B31. In A. tumefaciens, the circular and linear chromosomes are characterized by a common "chromosome-like" codon usage, whereas both plasmids share a distinct "plasmid-like" codon usage. In B. burgdorferi, in addition to different codon-usage biases on the leading and lagging strands of DNA replication found by McInerney (McInerney JO. 1998. Replicational and transcriptional selection on codon usage in Borrelia burgdorferi. Proc Natl Acad Sci USA. 95:10698-10703), we also detect a codon-usage similarity between linear plasmid lp38 and the leading strand of the chromosome and a high similarity among the cp32 family of plasmids.

Isolated limb perfusion: a novel delivery system for wild-type p53 and fiber-modified oncolytic adenoviruses to extremity sarcoma

Isolated limb perfusion (ILP) is a limb salvage surgical modality used to deliver chemotherapy and biologic agents to locally advanced and recurrent extremity soft tissue sarcoma (STS), and may be readily tailored for delivery of gene therapy. We set out to test the feasibility of delivering AdFLAGp53 (replication incompetent adenovirus bearing FLAG-tagged wild-type p53) and Ad.hTC.GFP/E1a.RGD (a fiber-modified, replication selective oncolytic adenovirus) into human leiomyosarcoma xenografts by ILP. Nude rats bearing SKLMS-1 tumors in their hind limbs underwent ILP with escalating doses of AdLacZ or AdFLAGp53 (study 1), or with Ad.CMV.GFP.RGD or Ad.hTC.GFP/E1a.RGD (study 2) following in vitro confirmation of therapeutic potential in STS cell lines and strains. Seventy-two hours after delivery, reverse transcription-polymerase chain reaction confirmed FLAGp53 expression, and immunohistochemistry confirmed diffuse upregulation of p21CIP1/WAF1 in ILP-treated tumors. Ad.hTC.GFP/E1a.RGD perfused tumors demonstrated robust macroscopic transgene expression throughout their substance, but not in perfused normal tissues, 21 days after delivery. Intra-tumoral viral replication was confirmed by immunohistochemical staining for early (E1a) and late (hexon) viral protein expression. Terminal deoxynucleotidyl transferase-mediated-digoxigenin nick end-labeling staining identified foci of cell death within regions of viral replication. In conclusion, therapeutic adenoviral gene therapy against limb borne human STS can be successfully delivered by ILP and warrants further investigation.

Monitoring Interfacial Bioelectrochemistry Using a FRET Switch

Generation of functionally active biomolecular monolayers is important in both analytical science and biophysical analyses. Our ability to monitor the redox-active state of immobilized proteins or enzymes at a molecular level, from which stochastic and surface-induced variations would be apparent, is impeded by comparatively slow electron-transfer kinetics and associated signal:noise difficulties. We demonstrate herein that by covalently tethering an appropriate dye to the copper protein azurin a highly oxidation-state-sensitive FRET process can be established which enables redox switching to be optically monitored at protein levels down to the zeptomolar limit. The surface-potential-induced cycling of emission enables the redox potential of clusters of a few hundred molecules to be determined.

Scanning probe technology in metalloprotein and biomolecular electronics

The interfacing of man-made electronic components with specifically-folded biomacromolecules lies central not only to the development of sensory interfaces and potential new molecular-scale devices, but also enables us to analyse processes of great biological importance in a refined and controllable manner. Recent advances in both available technology, most notably optical and scanning probes in nature, and our understanding of suitable methodologies, have led us to the point where the characteristics of single biological molecules can be interrogated with good levels of reproducibility. We review here the application of scanning probe microscopy to the analysis of and experimentation on biological redox systems. Within this paper the tunnel transport characteristics, as assayed by both scanning tunnelling microscopy (STM) and conducting probe atomic force microscopy (AFM), of single metalloproteins are discussed. In a specific case study the electron transfer characteristics of the blue copper metalloprotein, azurin, are reported. The modulation of these properties under the influence of calibratable compressional force has also been examined in some detail. Work such as this enables one to reproducibly establish the conductance, barrier height, environmental sensitivity and electromechanical properties of these molecules.

Oncolytic virotherapy for cancer treatment: challenges and solutions

Advances in gene modification and viral therapy have led to the development of a variety of vectors in several viral families that are capable of replication specifically in tumor cells. Because of the nature of viral delivery, infection, and replication, this technology, oncolytic virotherapy, may prove valuable for treating cancer patients, especially those with inoperable tumors. Current limitations exist, however, for oncolytic virotherapy. They include the body's B and T cell responses, innate inflammatory reactions, host range, safety risks involved in using modified viruses as treatments, and the requirement that most currently available oncolytic viruses require local administration. Another important constraint is that genetically enhanced vectors may or may not adhere to their replication restrictions in long-term applications. Several solutions and strategies already exist, however, to minimize or circumvent many of these limitations, supporting viral oncolytic therapy as a viable option and powerful tool in the fight against cancer.

The tunnelling conductance of molecularly ordered metalloprotein arrays

Metalloproteins can be self-assembled in molecularly ordered, electrochemically addressable arrays. We report here on a study of the transport characteristics of the blue copper protein, azurin, from Pseudomonas aeruginosa, by a combination of electrochemical and scanning probe techniques (scanning tunnelling microscopy and conducting atomic force microscopy). Redox-switchable chemisorbed molecular arrays can be formed from both wild-type and mutant proteins using the strong affinity of cysteine residue thiolates for pristine gold surfaces. The molecular transconductance of single protein molecules within these arrays has been studied under controllable conditions where it has been additionally possible to resolve the effects of protein mechanical perturbation. Although tunnelling appears to be non-resonant and adequately explained through the use of a square barrier model, under some conditions the contribution of the redox-active copper centre to conductance is resolvable.

Molecular bioelectronics

Biological macromolecules have evolved over many millions of years into structures primed, in some cases, for both specific surface recognition and facile, directional electron tunnelling. The redox-active centres of metalloproteins play a central role in photosynthesis and respiration. The processes by which constructive man-made interfaces to these moieties can be generated have advanced greatly during the past two decades or so. Together with recent advances in molecular manipulation, analyses and lithographic fabrication, this knowledge has led to us to the point where bioelectronic devices can be designed and interrogated with good levels of reproducibility.

High resolution scanning force microscopy of cardiac myocytes

The advent of scanning probe microscopy has introduced a powerful new method of probing the structural features of biological specimens. In this study, high resolution atomic force microscopy micrographs of single, isolated, cardiac myocytes are presented. Significantly, our images show not only the features to be expected of the external sarcolemma, but also resolve sub-surface features, including the striated pattern of the contractile proteins and their associated sarcoplasmic reticulum and mitochondria.

A scanning tunnelling study of immobilised cytochrome P450cam

A site-specifically engineered surface cysteine residue, located in a region where the haem moiety is closest to the surface, is used to anchor cytochrome P450cam enzyme molecules covalently to a gold electrode. More reproducibly ordered adsorption, at high coverage, occurs with this K344C mutant than with the wild-type enzyme. The subsequently formed close-packed monolayer arrays have been probed by scanning tunnelling microscopy under ambient conditions and under aqueous (buffered) solution at high resolution. Initial indications suggest that the immobilised enzyme is both electrochemically addressable and catalytically active.

Ultrastructural organization of amyloid fibrils by atomic force microscopy

Atomic force microscopy has been employed to investigate the structural organization of amyloid fibrils produced in vitro from three very different polypeptide sequences. The systems investigated are a 10-residue peptide derived from the sequence of transthyretin, the 90-residue SH3 domain of bovine phosphatidylinositol-3'-kinase, and human wild-type lysozyme, a 130-residue protein containing four disulfide bridges. The results demonstrate distinct similarities between the structures formed by the different classes of fibrils despite the contrasting nature of the polypeptide species involved. SH3 and lysozyme fibrils consist typically of four protofilaments, exhibiting a left-handed twist along the fibril axis. The substructure of TTR(10-19) fibrils is not resolved by atomic force microscopy and their uniform appearance is suggestive of a regular self-association of very thin filaments. We propose that the exact number and orientation of protofilaments within amyloid fibrils is dictated by packing of the regions of the polypeptide chains that are not directly involved in formation of the cross-beta core of the fibrils. The results obtained for these proteins, none of which is directly associated with any human disease, are closely similar to those of disease-related amyloid fibrils, supporting the concept that amyloid is a generic structure of polypeptide chains. The detailed architecture of an individual fibril, however, depends on the manner in which the protofilaments assemble into the fibrillar structure, which in turn is dependent on the sequence of the polypeptide and the conditions under which the fibril is formed.

Riskier than we think? The relationship between risk statement completeness and perceptions of direct to consumer advertised prescription drugs

Direct to consumer (DTC) prescription drug advertising is one of the fastest growing categories of advertising. Expenditures have increased from about $25 million in 1992 to nearly $2 billion in 1999. Given strong evidence of consumer-driven demand for advertised prescription drugs, research was conducted to assess the extent to which DTC prescription drug advertising provides consumers with the information they need to make an informed evaluation of an advertised drug's relative benefits and risks. Two studies explored the relationship between the completeness of the statement describing drug-associated side effects (the "risk statement") and consumers' perceptions of a drug's safety and appeal. The research manipulated risk statement completeness with regard to the incidence levels of side effects mentioned in the statement (which in turn affected the number of side effects mentioned) and the presence or absence of a numeric indicator of side effect incidence. The research strongly suggests a direct relationship between risk statement completeness and consumers' perceptions of drug safety and appeal. Consumers rate the safety and appeal of drugs described with an incomplete risk statement significantly more positively than comparable drugs described with a more complete risk statement. Implications of the research for the regulation and presentation of DTC prescription drug advertising and advertiser communication practices are discussed.