A Machine Learning Method for Genome Engineering Design Tool Attribution.

As the ability to engineer biological systems improves with increasingly advanced technology, the risk of accidental or intentional release of a dangerous genetically modified organism becomes greater. It is important that authorities can carry out attribution for the source of a genetically modified biological agent release. In the absence of evidence that ties a release directly to the individuals responsible, attribution can be carried out in part by discovering the in silico tools used to design the engineered genetic components, which can leave a signature in the DNA of the organism. Previous attribution methods have focused on identifying the laboratory of origin of an engineered organism using machine learning on plasmid signatures. The next logical step is to address attribution using signatures from the tools that are used to create the engineered modifications. A random forest classifier was developed that discriminates between design tools used to optimize coding regions for incorporation into the genome of another organism. To this end, tens of thousands of genes were optimized with 4 different codon optimization methods and relevant features from these sequences were generated for a machine learning classifier. This method achieves more than 97% accuracy in predicting which tools were used to design codon optimized genes for expression in other organisms. The methods presented here lay the groundwork for the creation of effective organism engineering attribution techniques. Such methods can act both as deterrents for future attempts at creating dangerous organisms as well as tools for forensic science.

An in vitro platform for study of the human gut microbiome under an oxygen gradient.

The complex, dynamic environment of the human lower gastrointestinal tract is colonized by hundreds of bacterial species that impact health and performance. Ex vivo study of the functional interactions between microbial community members in conditions representative of those in the gut is an ongoing challenge. We have developed an in vitro 40-plex platform that provides an oxygen gradient to support simultaneous maintenance of microaerobic and anaerobic microbes from the gut microbiome that can aid in rapid characterization of microbial interactions and direct comparison of individual microbiome samples. In this report, we demonstrate that the platform more closely maintained the microbial diversity and composition of human donor fecal microbiome samples than strict anaerobic conditions. The oxygen gradient established in the platform allowed the stratification and subsequent sampling of diverse microbial subpopulations that colonize microaerobic and anaerobic micro-environments. With the ability to run forty samples in parallel, the platform has the potential to be used as a rapid screening tool to understand how the gut microbiome responds to environmental perturbations such as toxic compound exposure, dietary changes, or pharmaceutical treatments.

Geographic source estimation using airborne plant environmental DNA in dust.

Information obtained from the analysis of dust, particularly biological particles such as pollen, plant parts, and fungal spores, has great utility in forensic geolocation. As an alternative to manual microscopic analysis of dust components, we developed a pipeline that utilizes the airborne plant environmental DNA (eDNA) in settled dust to estimate geographic origin. Metabarcoding of settled airborne eDNA was used to identify plant species whose geographic distributions were then derived from occurrence records in the USGS Biodiversity in Service of Our Nation (BISON) database. The distributions for all plant species identified in a sample were used to generate a probabilistic estimate of the sample source. With settled dust collected at four U.S. sites over a 15-month period, we demonstrated positive regional geolocation (within 600 km2 of the collection point) with 47.6% (20 of 42) of the samples analyzed. Attribution accuracy and resolution was dependent on the number of plant species identified in a dust sample, which was greatly affected by the season of collection. In dust samples that yielded a minimum of 20 identified plant species, positive regional attribution was achieved with 66.7% (16 of 24 samples). For broader demonstration, citizen-collected dust samples collected from 31 diverse U.S. sites were analyzed, and trace plant eDNA provided relevant regional attribution information on provenance in 32.2% of samples. This showed that analysis of airborne plant eDNA in settled dust can provide an accurate estimate regional provenance within the U.S., and relevant forensic information, for a substantial fraction of samples analyzed.

Predicting Influenza A Tropism with End-to-End Learning of Deep Networks.

The type of host that a virus can infect, referred to as host specificity or tropism, influences infectivity and thus is important for disease diagnosis, epidemic response, and prevention. Advances in DNA sequencing technology have enabled rapid metagenomic analyses of viruses, but the prediction of virus phenotype from genome sequences is an active area of research. As such, automatic prediction of host tropism from analysis of genomic information is of considerable utility. Previous research has applied machine learning methods to accomplish this task, although deep learning (particularly deep convolutional neural network, CNN) techniques have not yet been applied. These techniques have the ability to learn how to recognize critical hierarchical structures within the genome in a data-driven manner. We designed deep CNN models to identify host tropism for human and avian influenza A viruses based on protein sequences and performed a detailed analysis of the results. Our findings show that deep CNN techniques work as well as existing approaches (with 99% mean accuracy on the binary prediction task) while performing end-to-end learning of the prediction model (without the need to specify handcrafted features). The findings also show that these models, combined with standard principal component analysis, can be used to quantify and visualize viral strain similarity.

Comprehensive evaluation of poly(I:C) induced inflammatory response in an airway epithelial model.

Respiratory viruses invade the upper airway of the lung, triggering a potent immune response that often exacerbates preexisting conditions such as asthma and COPD. Poly(I:C) is a synthetic analog of viral dsRNA that induces the characteristic inflammatory response associated with viral infection, such as loss of epithelial integrity, and increased production of mucus and inflammatory cytokines. Here, we explore the mechanistic responses to poly(I:C) in a well-defined primary normal human bronchial epithelial (NHBE) model that recapitulates in vivo functions and responses. We developed functional and quantifiable methods to evaluate the physiology of our model in both healthy and inflamed states. Through gene and protein expression, we validated the differentiation state and population of essential cell subtypes (i.e., ciliated, goblet, club, and basal cells) as compared to the human lung. Assays for total mucus production, cytokine secretion, and barrier function were used to evaluate in vitro physiology and response to viral insult. Cells were treated apically with poly(I:C) and evaluated 48 h after induction. Results revealed a dose-dependent increase in goblet cell differentiation, as well as, an increase in mucus production relative to controls. There was also a dose-dependent increase in secretion of IL-6, IL-8, TNF-α, and RANTES. Epithelial barrier function, as measured by TEER, was maintained at 1501 ± 355 Ω*cm² postdifferentiation, but dropped significantly when challenged with poly(I:C). This study provides first steps toward a well-characterized model with defined functional methods for understanding dsRNA stimulated inflammatory responses in a physiologically relevant manner.

Heparin octasaccharide decoy liposomes inhibit replication of multiple viruses.

Heparan sulfate (HS) is a ubiquitous glycosaminoglycan that serves as a cellular attachment site for a number of significant human pathogens, including respiratory syncytial virus (RSV), human parainfluenza virus 3 (hPIV3), and herpes simplex virus (HSV). Decoy receptors can target pathogens by binding to the receptor pocket on viral attachment proteins, acting as 'molecular sinks' and preventing the pathogen from binding to susceptible host cells. Decoy receptors functionalized with HS could bind to pathogens and prevent infection, so we generated decoy liposomes displaying HS-octasaccharide (HS-octa). These decoy liposomes significantly inhibited RSV, hPIV3, and HSV infectivity in vitro to a greater degree than the original HS-octa building block. The degree of inhibition correlated with the density of HS-octa displayed on the liposome surface. Decoy liposomes with HS-octa inhibited infection of viruses to a greater extent than either full-length heparin or HS-octa alone. Decoy liposomes were effective when added prior to infection or following the initial infection of cells in vitro. By targeting the well-conserved receptor-binding sites of HS-binding viruses, decoy liposomes functionalized with HS-octa are a promising therapeutic antiviral agent and illustrate the utility of the liposome delivery platform.

A breath fungal secondary metabolite signature to diagnose invasive aspergillosis.

BACKGROUND: Invasive aspergillosis (IA) remains a leading cause of mortality in immunocompromised patients, in part due to the difficulty of diagnosing this infection. METHODS: Using thermal desorption-gas chromatography/mass spectrometry, we characterized the in vitro volatile metabolite profile of Aspergillus fumigatus, the most common cause of IA, and other pathogenic aspergilli. We prospectively collected breath samples from patients with suspected invasive fungal pneumonia from 2011 to 2013, and assessed whether we could discriminate patients with proven or probable IA from patients without aspergillosis, as determined by European Organization for Research and Treatment of Cancer/Mycoses Study Group consensus definitions, by direct detection of fungal volatile metabolites in these breath samples. RESULTS: The monoterpenes camphene, α- and ß-pinene, and limonene, and the sesquiterpene compounds α- and ß-trans-bergamotene were distinctive volatile metabolites of A. fumigatus in vitro, distinguishing it from other pathogenic aspergilli. Of 64 patients with suspected invasive fungal pneumonia based on host risk factors, clinical symptoms, and radiologic findings, 34 were diagnosed with IA, whereas 30 were ultimately diagnosed with other causes of pneumonia, including other invasive mycoses. Detection of α-trans-bergamotene, ß-trans-bergamotene, a ß-vatirenene-like sesquiterpene, or trans-geranylacetone identified IA patients with 94% sensitivity (95% confidence interval [CI], 81%-98%) and 93% specificity (95% CI, 79%-98%). CONCLUSIONS: In patients with suspected fungal pneumonia, an Aspergillus secondary metabolite signature in breath can identify individuals with IA. These results provide proof-of-concept that direct detection of exogenous fungal metabolites in breath can be used as a novel, noninvasive, pathogen-specific approach to identifying the precise microbial cause of pneumonia.

Sialylneolacto-N-tetraose c (LSTc)-bearing liposomal decoys capture influenza A virus.

Influenza is a severe disease in humans and animals with few effective therapies available. All strains of influenza virus are prone to developing drug resistance due to the high mutation rate in the viral genome. A therapeutic agent that targets a highly conserved region of the virus could bypass resistance and also be effective against multiple strains of influenza. Influenza uses many individually weak ligand binding interactions for a high avidity multivalent attachment to sialic acid-bearing cells. Polymerized sialic acid analogs can form multivalent interactions with influenza but are not ideal therapeutics due to solubility and toxicity issues. We used liposomes as a novel means for delivery of the glycan sialylneolacto-N-tetraose c (LSTc). LSTc-bearing decoy liposomes form multivalent, polymer-like interactions with influenza virus. Decoy liposomes competitively bind influenza virus in hemagglutination inhibition assays and inhibit infection of target cells in a dose-dependent manner. Inhibition is specific for influenza virus, as inhibition of Sendai virus and respiratory syncytial virus is not observed. In contrast, monovalent LSTc does not bind influenza virus or inhibit infectivity. LSTc decoy liposomes prevent the spread of influenza virus during multiple rounds of replication in vitro and extend survival of mice challenged with a lethal dose of virus. LSTc decoy liposomes co-localize with fluorescently tagged influenza virus, whereas control liposomes do not. Considering the conservation of the hemagglutinin binding pocket and the ability of decoy liposomes to form high avidity interactions with influenza hemagglutinin, our decoy liposomes have potential as a new therapeutic agent against emerging influenza strains.

Pseudomonas aeruginosa fimL regulates multiple virulence functions by intersecting with Vfr-modulated pathways.

Virulence of Pseudomonas aeruginosa involves the co-ordinate expression of a range of factors including type IV pili (tfp), the type III secretion system (TTSS) and quorum sensing. Tfp are required for twitching motility, efficient biofilm formation, and for adhesion and type III secretion (TTS)-mediated damage to mammalian cells. We describe a novel gene (fimL) that is required for tfp biogenesis and function, for TTS and for normal biofilm development in P. aeruginosa. The predicted product of fimL is homologous to the N-terminal domain of ChpA, except that its putative histidine and threonine phosphotransfer sites have been replaced with glutamine. fimL mutants resemble vfr mutants in many aspects including increased autolysis, reduced levels of surface-assembled tfp and diminished production of type III secreted effectors. Expression of vfr in trans can complement fimL mutants. vfr transcription and production is reduced in fimL mutants whereas cAMP levels are unaffected. Deletion and insertion mutants of fimL frequently revert to wild-type phenotypes suggesting that an extragenic suppressor mutation is able to overcome the loss of fimL. vfr transcription and production, as well as cAMP levels, are elevated in these revertants, while Pseudomonas quinolone signal (PQS) production is reduced. These results suggest that the site(s) of spontaneous mutation is in a gene(s) which lies upstream of vfr transcription, cAMP, production, and PQS synthesis. Our studies indicate that Vfr and FimL are components of intersecting pathways that control twitching motility, TTSS and autolysis in P. aeruginosa.

The role of dor gene products in controlling the P2 promoter of the cytochrome c2 gene, cycA, in Rhodobacter sphaeroides.

This study explores the regulatory networks controlling anaerobic energy production by the facultative phototroph Rhodobacter sphaeroides. The specific aim was to determine why activity of the P2 promoter for the gene (cycA) encoding the essential photosynthetic electron carrier, cytochrome c(2), is decreased when the alternative electron acceptor DMSO is added to photosynthetically grown cells. The presence of DMSO is believed to activate the DorR response regulator, which controls expression of proteins required to reduce DMSO. A DorR(-) strain showed no change in cycA P2 promoter activity when DMSO was added to photosynthetic cells, indicating that DorR was required for the decreased expression in wild-type cells. To test if DorR acted directly at this promoter to change gene expression, recombinant DorR was purified and studied in vitro. Preparations of DorR that were active at other target promoters showed no detectable interaction with cycA P2, suggesting that this protein is not a direct regulator of this promoter. We also found that cycA P2 activity in a DorA(-) strain was not decreased by the addition of DMSO to photosynthetic cells. A model is presented to explain why the presence of a functional DMSO reductase (DorA) is required for DMSO to decrease cycA P2 expression under photosynthetic conditions.

Differences in two Pseudomonas aeruginosa cbb3 cytochrome oxidases.

Bacterial cytochrome cbb3 oxidases are members of the haeme-copper oxidase superfamily that are important for energy conservation by a variety of proteobacteria under oxygen-limiting conditions. The opportunistic pathogen Pseudomonas aeruginosa is unusual in possessing two operons that each potentially encode a cbb3 oxidase (cbb3-1 or cbb3-2). Our results demonstrate that, unlike typical enzymes of this class, the cbb3-1 oxidase has an important metabolic function at high oxygen tensions. In highly aerated cultures, cbb3-1 abundance and expression were greater than that of cbb3-2, and only loss of cbb3-1 influenced growth. Also, the activity of cbb3-1, not cbb3-2, inhibited expression of the alternative oxidase CioAB and thus influenced a signal transduction pathway much like that found in the alpha-proteobacterium Rhodobacter sphaeroides. Cbb3-2 appeared to play a more significant role under oxygen limitation by nature of its increased abundance and expression compared to highly aerated cultures, and the regulation of the cbb3-2 operon by the putative iron-sulphur protein Anr. These results indicate that each of the two P. aeruginosa cbb3 isoforms have assumed specialized energetic and regulatory roles.

A type-1 phosphoprotein phosphatase from a dinoflagellate as a possible component of the circadian mechanism.

Indicative of the importance of protein phosphorylation in the core circadian clock mechanism, chronically applied inhibitors of both protein kinases and phosphoprotein phosphatases have significant effects on the period, phase, and light-dependent regulation of circadian rhythms in the dinoflagellate Lingulodinium polyedrum. This study was aimed at identifying the presence of the affected phosphatase(s). Dephosphorylation of a PP1/PP2A-specific substrate by L. polyedrum extracts was inhibited by okadaic acid only at concentrations greater than 100 nM, as in vivo, by mammalian inhibitor-2 (I-2), and by an endogenous inhibitor with properties similar to I-2, indicating that a type-1 protein phosphatase (PP1) was predominant. A cDNA encoding a highly conserved PP1 was isolated, the 1st such signaling molecule identified in dinoflagellates. Antisera specific for this type of phosphatase recognized a 34 kDa protein in L. polyedrum extract, this being the same size as the PP1 encoded by the isolated cDNA. These findings are consistent with the suggestion that the L. polyedrum PP1 may be a part of the clock mechanism in this species.

Pseudomonas aeruginosa RoxR, a response regulator related to Rhodobacter sphaeroides PrrA, activates expression of the cyanide-insensitive terminal oxidase.

The facultative anaerobe Pseudomonas aeruginosa has multiple aerobic electron transport pathways, one of which is terminated by a cyanide-insensitive oxidase (CIO). This study characterizes a P. aeruginosa two-component system that regulates CIO production. The response regulator of this system (RoxR) has significant amino acid sequence similarity to PrrA of Rhodobacter sphaeroides and related proteins in other alpha-proteobacteria. In heterologous complementation analysis, R. sphaeroides PrrA rescued the growth defect of a P. aeruginosa mutant lacking RoxR, and RoxR enabled photosynthetic growth of an R. sphaeroides PrrA mutant. Also, RoxR could substitute for PrrA in activating transcription in vitro, demonstrating that these proteins are functional homologues. P. aeruginosa strains lacking RoxR or the sensor kinase (RoxS) were more sensitive than wild type to the respiratory inhibitors cyanide and azide. The phenotypes of these mutant strains correlated with reduced cyanide-insensitive O2 utilization and less cyanide-dependent expression of the locus encoding the CIO (cioAB). The ability of purified RoxR to bind to the cioAB promoter region also suggests that this protein acts directly to regulate cioAB transcription. Therefore, RoxR appears to play a role in regulating the transcription of loci for P. aeruginosa energy-generating enzymes similar to that of its homologues in alpha-proteobacteria.

Transcriptional activation of the Rhodobacter sphaeroides cytochrome c(2) gene P2 promoter by the response regulator PrrA.

Anoxygenic photosynthetic growth of Rhodobacter sphaeroides, a member of the alpha subclass of the class Proteobacteria, requires the response regulator PrrA. PrrA and the sensor kinase PrrB are part of a two-component signaling pathway that influences a wide range of processes under oxygen-limited conditions. In this work we characterized the pathway of transcription activation by PrrB and PrrA by purifying these proteins, analyzing them in vitro, and characterizing a mutant PrrA protein in vivo and in vitro. When purified, a soluble transmitter domain of PrrB (cPrrB) could autophosphorylate, rapidly transfer phosphate to PrrA, and stimulate dephosphorylation of phospho-PrrA. Unphosphorylated PrrA activated transcription from a target cytochrome c(2) gene (cycA) promoter, P2, which contained sequences from -73 to +22 relative to the transcription initiation site. However, phosphorylation of PrrA increased its activity since activation of cycA P2 was enhanced up to 15-fold by treatment with the low-molecular-weight phosphodonor acetyl phosphate. A mutant PrrA protein containing a single amino acid substitution in the presumed phosphoacceptor site (PrrA-D63A) was not phosphorylated in vitro but also was not able to stimulate cycA P2 transcription. PrrA-D63A also had no apparent in vivo activity, demonstrating that aspartate 63 is necessary both for the function of PrrA and for its phosphorylation-dependent activation. The cellular level of wild-type PrrA was negatively autoregulated so that less PrrA was present in the absence of oxygen, conditions in which the activities of many PrrA target genes increase. PrrA-D63A failed to repress expression of the prrA gene under anaerobic conditions, suggesting that this single amino acid change also eliminated PrrA function in vivo.