Next-Generation Sequencing for Biodefense: Biothreat Detection, Forensics, and the Clinic.

BACKGROUND: Next-generation sequencing (NGS) is revolutionizing a variety of molecular biology fields including bioforensics, biosurveillance, and infectious disease diagnostics. For pathogen detection, the ability to sequence all nucleic acids in a sample allows near limitless multiplexability, free from a priori knowledge regarding an etiologic agent as is typically required for targeted molecular assays such as real-time PCR. Furthermore, sequencing capabilities can generate in depth genomic information, allowing detailed molecular epidemiological studies and bioforensics analysis, which is critical for source agent identification in a biothreat outbreak. However, lack of analytical specificity, inherent to NGS, presents challenges for regulated applications such as clinical diagnostics and molecular attribution. CONTENT: Here, we discuss NGS applications in the context of preparedness and biothreat readiness. Specifically, we investigate current and future applications of NGS technologies to affect the fields of biosurveillance, bioforensics, and clinical diagnostics with specific focus on biodefense. SUMMARY: Overall, there are many advantages to the implementation of NGS for preparedness and readiness against biowarfare agents, from forensics to diagnostics. However, appropriate caveats must be associated with any technology. This includes NGS. While NGS is not the panacea replacing all molecular techniques, it will greatly enhance the ability to detect, characterize, and diagnose biowarfare agents, thus providing an excellent addition to the biodefense toolbox of biosurveillance, bioforensics, and biothreat diagnosis.

Crimean-Congo Hemorrhagic Fever Virus, Mongolia, 2013-2014.

During 2013-2014, we collected 1,926 serum samples from humans and 4,583 ticks (Hyalomma asiaticum or Dermacentor nuttalli) in select regions of Mongolia to determine the risk for Crimean-Congo hemorrhagic fever virus (CCHFV) infection among humans in this country. Testing of human serum samples by ELISA demonstrated an overall CCHFV antibody prevalence of 1.4%; Bayankhongor Province had the highest prevalence, 2.63%. We pooled and analyzed tick specimens by real-time reverse transcription PCR; 1 CCHFV-positive H. asiaticum tick pool from Ömnögovi was identified. In phylogenetic analyses, the virus's partial small segment clustered with CCHFV isolates from Central Asia, and the complete medium segment grouped with CCHFV isolates from Africa, Asia, and the Middle East. This study confirms CCHFV endemicity in Mongolia and provides information on risk for CCHFV infection. Further research is needed to better define the risk for CCHFV disease to improve risk mitigation, diagnostics, and surveillance.

Development of real-time reverse transcriptase qPCR assays for the detection of Punta Toro virus and Pichinde virus.

BACKGROUND: Research with high biocontainment pathogens such as Rift Valley fever virus (RVFV) and Lassa virus (LASV) is expensive, potentially hazardous, and limited to select institutions. Surrogate pathogens such as Punta Toro virus (PTV) for RVFV infection and Pichinde virus (PICV) for LASV infection allow research to be performed under more permissive BSL-2 conditions. Although used as infection models, PTV and PICV have no standard real-time RT-qPCR assays to detect and quantify pathogenesis. PTV is also a human pathogen, making a standardized detection assay essential for biosurveillance. Here, we developed and characterized two real-time RT-qPCR assays for PICV and PTV by optimizing assay conditions and measuring the limit of detection (LOD) and performance in multiple clinical matrices. METHODS: Total nucleic acid from virus-infected Vero E6 cells was used to optimize TaqMan-minor groove binder (MGB) real-time RT-qPCR assays. A 10-fold dilution series of nucleic acid was used to perform analytical experiments with 60 replicates used to confirm assay LODs. Serum and whole blood spiked with 10-fold dilutions of PTV and PICV virus were assessed as matrices in a mock clinical context. The Cq, or cycle at which the fluoresce of each sample first crosses a threshold line, was determined using the second derivative method using Roche LightCycler 480 software version 1.5.1. Digital droplet PCR (ddPCR) was utilized to quantitatively determine RNA target counts/µl for PTV and PICV. RESULTS: Optimized PTV and PICV assays had LODs of 1000 PFU/ml and 100 PFU/ml, respectively, and this LOD was confirmed in 60/60 (PTV) and 58/60 (PICV) positive replicates. Preliminary mock clinical LODs remained consistent in serum and whole blood for PTV and PICV at 1000 PFU/ml and 100 PFU/ml. An exclusivity panel showed no cross reaction with near neighbors. CONCLUSIONS: PTV and PICV Taq-man MGB based real-time RT-qPCR assays developed here showed relevant sensitivity and reproducibility in samples extracted from a variety of clinical matrices. These assays will be useful as a standard by researchers for future experiments utilizing PTV and PICV as infection models, offering the ability to track infection and viral replication kinetics during research studies.

Optimized microRNA purification from TRIzol-treated plasma.

BACKGROUND: MicroRNAs (miRNAs) represent new and potentially informative diagnostic targets for disease detection and prognosis. However, little work exists documenting the effect of TRIzol, a common viral inactivation and nucleic acid extraction reagent, on miRNA purification. Here, we developed an optimized protocol for miRNA extraction from plasma samples by evaluating five different RNA extraction kits, TRIzol phase separation, purification additives, and initial plasma sample volume. This method was then used for downstream profiling of plasma miRNAs found in archived samples from one nonhuman primate (NHP) experimentally challenged with Ebola virus by the aerosol route. RESULTS: Comparison of real-time RT-PCR results for spiked-in and endogenous miRNA sequences determined extraction efficiencies from five different RNA purification kits. These experiments showed that 50 µL plasma processed using the QIAGEN miRNeasy Mini Kit with 5 µg of glycogen as a co-precipitant yielded the highest recovery of endogenous miRNAs. Using this optimized protocol, miRNAs from archived plasma samples of one rhesus macaque challenged with aerosolized Ebola virus was profiled using a targeted real-time PCR array. A total of 519 of the 752 unique miRNAs assayed were present in the plasma samples at day 0 and day 7 (time of death) post-exposure. Statistical analyses revealed 25 sequences significantly up- or down-regulated between day 0 and day 7 post infection, validating the utility of the extraction method for plasma miRNA profiling. CONCLUSIONS: This study contributes to the knowledgebase of circulating miRNA extraction methods and expands on the potential applications of cell-free miRNA profiling for diagnostics and pathogenesis studies. Specifically, we optimized an extraction protocol for miRNAs from TRIzol-inactivated plasma samples that can be used for highly pathogenic viruses.

Monitoring of Ebola Virus Makona Evolution through Establishment of Advanced Genomic Capability in Liberia.

To support Liberia's response to the ongoing Ebola virus (EBOV) disease epidemic in Western Africa, we established in-country advanced genomic capabilities to monitor EBOV evolution. Twenty-five EBOV genomes were sequenced at the Liberian Institute for Biomedical Research, which provided an in-depth view of EBOV diversity in Liberia during September 2014-February 2015. These sequences were consistent with a single virus introduction to Liberia; however, shared ancestry with isolates from Mali indicated at least 1 additional instance of movement into or out of Liberia. The pace of change is generally consistent with previous estimates of mutation rate. We observed 23 nonsynonymous mutations and 1 nonsense mutation. Six of these changes are within known binding sites for sequence-based EBOV medical countermeasures; however, the diagnostic and therapeutic impact of EBOV evolution within Liberia appears to be low.

Demonstration of the Pre-Emergency Use Authorization Path Using 3 Minor Groove Binder-Hydrolysis Probe Assays to Detect Escherichia coli O104:H4.

BACKGROUND: The Department of Defense (DoD) and the Food and Drug Administration (FDA) have collaboratively worked on a pre-Emergency Use Authorization (pre-EUA) process for in vitro diagnostic (IVD) devices, using FDA's regulatory flexibilities under the EUA authorities. The pre-EUA process enables FDA review of data in anticipation of a request for an EUA, advancing US government public health emergency preparedness efforts. METHODS: The IVD device developed to detect Escherichia coli O104:H4, for which an EUA has not been issued, serves as an example to illustrate that process. Specifically, DoD designed real-time PCR assays to target the virulent E. coli strain O104:H4 (etiological agent of the 2011 German outbreak) including: fliC (flagellin), Agg3C (AAF), and rfb (wbwC) on the basis of the published sequences. RESULTS: After development and optimization of these 3 specific assays, a defined protocol was followed to determine and document the sensitivity and specificity of each assay analytically. CONCLUSIONS: FDA reviewed these data and returned commentary on additional required experiments to complete the pre-EUA process and expedite the use of the device should there be an emergency need for an IVD device to detect this virulent E. coli strain before such a test is cleared by FDA.

Real-time reverse transcriptase polymerase chain reaction assays for Middle East Respiratory Syndrome.

Here we designed and tested two highly specific quantitative TaqMan(®)-MGB-based reverse transcriptase quantitative polymerase chain reaction (RT-qPCR) assays for Middle East Respiratory Syndrome (MERS). The primers and probes for these assays were evaluated and found to have a limit of detection (LOD) of 0.005 plaque-forming units/PCR (pfu/PCR).

Development of real-time PCR assays for the detection of Moraxella macacae associated with bloody nose syndrome in rhesus (Macaca mulatta) and cynomolgus (Macaca fascicularis) macaques.

BACKGROUND: Moraxella macacae is a recently described bacterial pathogen that causes epistaxis or so-called bloody nose syndrome in captive macaques. The aim of this study was to develop specific molecular diagnostic assays for M. macacae and to determine their performance characteristics. METHODS: We developed six real-time PCR assays on the Roche LightCycler. The accuracy, precision, selectivity, and limit of detection (LOD) were determined for each assay, in addition to further validation by testing nasal swabs from macaques presenting with epistaxis at the Tulane National Primate Research Center. RESULTS: All assays exhibited 100% specificity and were highly sensitive with an LOD of 10 fg for chromosomal assays and 1 fg for the plasmid assay. Testing of nasal swabs from 10 symptomatic macaques confirmed the presence of M. macacae in these animals. CONCLUSIONS: We developed several accurate, sensitive, and species-specific real-time PCR assays for the detection of M. macacae in captive macaques.

Comparative assessment of automated nucleic acid sample extraction equipment for biothreat agents.

Magnetic beads offer superior impurity removal and nucleic acid selection over older extraction methods. The performances of nucleic acid extraction of biothreat agents in blood or buffer by easyMAG, MagNA Pure, EZ1 Advanced XL, and Nordiag Arrow were evaluated. All instruments showed excellent performance in blood; however, the easyMAG had the best precision and versatility.

Cross-institute evaluations of inhibitor-resistant PCR reagents for direct testing of aerosol and blood samples containing biological warfare agent DNA.

Rapid pathogen detection is crucial for the timely introduction of therapeutics. Two groups (one in the United Kingdom and one in the United States) independently evaluated inhibitor-resistant PCR reagents for the direct testing of substrates. In the United Kingdom, a multiplexed Bacillus anthracis (target) and Bacillus subtilis (internal-control) PCR was used to evaluate 4 reagents against 5 PCR inhibitors and down-selected the TaqMan Fast Virus 1-Step master mix (Life Technologies Inc.). In the United States, four real-time PCR assays (targeting B. anthracis, Brucella melitensis, Venezuelan equine encephalitis virus [VEEV], and Orthopoxvirus spp.) were used to evaluate 5 reagents (plus the Fast Virus master mix) against buffer, blood, and soil samples and down-selected the KAPA Blood Direct master mix (KAPA Biosystems Inc.) with added Platinum Taq (Life Technologies). The down-selected reagents underwent further testing. In the United Kingdom experiments, both reagents were tested against seven contrived aerosol collector samples containing B. anthracis Ames DNA and B. subtilis spores from a commercial formulation (BioBall). In PCR assays with reaction mixtures containing 40% crude sample, an airfield-collected sample induced inhibition of the B. subtilis PCR with the KAPA reagent and complete failure of both PCRs with the Fast Virus reagent. However, both reagents allowed successful PCR for all other samples-which inhibited PCRs with a non-inhibitor-resistant reagent. In the United States, a cross-assay limit-of-detection (LoD) study in blood was conducted. The KAPA Blood Direct reagent allowed the detection of agent DNA (by four PCRs) at higher concentrations of blood in the reaction mixture (2.5%) than the Fast Virus reagent (0.5%), although LoDs differed between assays and reagent combinations. Across both groups, the KAPA Blood Direct reagent was determined to be the optimal reagent for inhibition relief in PCR.

Development of real-time PCR assays for specific detection of hmsH, hmsF, hmsR, and irp2 located within the 102-kb pgm locus of Yersinia pestis.

Virulent isolates of three pathogenic Yersinia species (Yersinia pestis, Yersinia pseudotuberculosis, and Yersinia enterocolitica) harbor a 102-kb chromosomal region which encodes elements critical for virulence. A 35-kb high pathogenicity island is contained in this region, is a known virulence determinant, contains irp1 and irp2 iron-regulating genes. An additional segment, the 68-kb high pathogenicity island, contains genetic elements responsible for conferring the Y. pestis pigmentation phenotype on Congo red agar at 28 °C. Collectively, these contiguous segments are referred to as the pigmentation (pgm) locus, the absence of which results in strain attenuation and exemption from CDC Select Agent status. In this study, we developed a set of four real-time PCR assays to detect the presence or absence of multiple virulence genes located within this region. Specifically, we designed TaqMan(®) PCR assays to individually detect three hemin storage genes (hmsH, hmsF, and hmsR) which are genetic elements that confer the pigmentation phenotype, as well as the iron-regulating status of 25 Y. pestis isolates (representing 23 different strains), thus establishing a molecular based assay capable of determining the pgm status of candidate Y. pestis isolates. Included in the validation process, was a comparison of these real-time PCR assays and newly developed conventional PCR assays targeting much larger areas of the 102-kb region (including one assay spanning hmsR and hmsF, one spanning hmsH and hsmF, one targeting hmsF, and one targeting irp2). There was high concordance between the conventional and real-time PCR assays for all Y. pestis strains tested. The results from the comparative analysis document the specificity and sensitivity of the real-time PCR assays and further solidify the ostensible benefits of real-time PCR over conventional PCR.

Relative quantification of the recA gene for antimicrobial susceptibility testing in response to ciprofloxacin for pathogens of concern.

Antimicrobial resistance (AR) is one of the greatest threats to global health and is associated with higher treatment costs, longer hospital stays, and increased mortality. Current gold standard antimicrobial susceptibility tests (AST) rely on organism growth rates that result in prolonged time-to-answer for slow growing organisms. Changes in the cellular transcriptome can be rapid in the presence of stressors such as antibiotic pressure, providing the opportunity to develop AST towards transcriptomic signatures. Here, we show that relative quantification of the recA gene is an indicator of pathogen susceptibly when select species are challenged with relevant concentrations of ciprofloxacin. We demonstrate that ciprofloxacin susceptible strains of Y. pestis and B. anthracis have significant increases in relative recA gene expression after 15 min of exposure while resistant strains show no significant differences. Building upon this data, we designed and optimized seven duplex RT-qPCR assays targeting the recA and 16S rRNA gene, response and housekeeping genes, respectively, for multiple biothreat and ESKAPE pathogens. Final evaluation of all seven duplex assays tested against 124 ciprofloxacin susceptible and resistant strains, including Tier 1 pathogens, demonstrated an overall categorical agreement compared to microbroth dilution of 97% using a defined cutoff. Testing pathogen strains commonly associated with urinary tract infections in contrived mock sample sets demonstrated an overall categorical agreement of 96%. These data indicate relative quantification of a single highly conserved gene accurately determines susceptibility for multiple bacterial species in response to ciprofloxacin.

Sequence optimized diagnostic assay for Ebola virus detection.

Rapid pathogen identification is a critical first step in patient isolation, treatment, and controlling an outbreak. Real-time PCR is a highly sensitive and specific approach commonly used for infectious disease diagnostics. However, mismatches in the primer or probe sequence and the target organism can cause decreased sensitivity, assay failure, and false negative results. Limited genomic sequences for rare pathogens such as Ebola virus (EBOV) can negatively impact assay performance due to undiscovered genetic diversity. We previously developed and validated several EBOV assays prior to the 2013-2016 EBOV outbreak in West Africa, and sequencing EBOV Makona identified sequence variants that could impact assay performance. Here, we assessed the impact sequence mismatches have on EBOV assay performance, finding one or two primer or probe mismatches resulted in a range of impact from minimal to almost two log sensitivity reduction. Redesigning this assay improved detection of all EBOV variants tested. Comparing the performance of the new assay with the previous assays across a panel of human EBOV samples confirmed increased assay sensitivity as reflected in decreased Cq values with detection of three positive that tested negative with the original assay.

Host-response transcriptional biomarkers accurately discriminate bacterial and viral infections of global relevance.

Diagnostic limitations challenge management of clinically indistinguishable acute infectious illness globally. Gene expression classification models show great promise distinguishing causes of fever. We generated transcriptional data for a 294-participant (USA, Sri Lanka) discovery cohort with adjudicated viral or bacterial infections of diverse etiology or non-infectious disease mimics. We then derived and cross-validated gene expression classifiers including: 1) a single model to distinguish bacterial vs. viral (Global Fever-Bacterial/Viral [GF-B/V]) and 2) a two-model system to discriminate bacterial and viral in the context of noninfection (Global Fever-Bacterial/Viral/Non-infectious [GF-B/V/N]). We then translated to a multiplex RT-PCR assay and independent validation involved 101 participants (USA, Sri Lanka, Australia, Cambodia, Tanzania). The GF-B/V model discriminated bacterial from viral infection in the discovery cohort an area under the receiver operator curve (AUROC) of 0.93. Validation in an independent cohort demonstrated the GF-B/V model had an AUROC of 0.84 (95% CI 0.76-0.90) with overall accuracy of 81.6% (95% CI 72.7-88.5). Performance did not vary with age, demographics, or site. Host transcriptional response diagnostics distinguish bacterial and viral illness across global sites with diverse endemic pathogens.

Comparison of Illumina and Oxford Nanopore Sequencing Technologies for Pathogen Detection from Clinical Matrices Using Molecular Inversion Probes.

Next-generation sequencing is rapidly finding footholds in numerous microbiological fields, including infectious disease diagnostics. Here, we describe a molecular inversion probe panel for the identification of bacterial, viral, and parasitic pathogens. We describe the ability of Illumina and Oxford Nanopore Technologies (ONT) to sequence small amplicons originating from this panel for the identification of pathogens in complex matrices. The panel correctly classified 31 bacterial pathogens directly from positive blood culture bottles with a genus-level concordance of 96.7% and 90.3% on the Illumina and ONT platforms, respectively. Both sequencing platforms detected 18 viral and parasitic organisms directly from mock clinical samples of plasma and whole blood at concentrations of 104 PFU/mL with few exceptions. In general, Illumina sequencing exhibited greater read counts with lower percent mapped reads; however, this resulted in no effect on limits of detection compared with ONT sequencing. Mock clinical evaluation of the probe panel on the Illumina and ONT platforms resulted in positive predictive values of 0.91 and 0.88 and negative predictive values of 1 and 1 from de-identified human chikungunya virus samples compared with gold standard quantitative RT-PCR. Overall, these data show that molecular inversion probes are an adaptable technology capable of pathogen detection from complex sample matrices on current next-generation sequencing platforms.

Exposure Route Influences Disease Severity in the COVID-19 Cynomolgus Macaque Model.

The emergence of SARS-CoV-2 and the subsequent pandemic has highlighted the need for animal models that faithfully replicate the salient features of COVID-19 disease in humans. These models are necessary for the rapid selection, testing, and evaluation of potential medical countermeasures. Here, we performed a direct comparison of two distinct routes of SARS-CoV-2 exposure-combined intratracheal/intranasal and small particle aerosol-in two nonhuman primate species, rhesus and cynomolgus macaques. While all four experimental groups displayed very few outward clinical signs, evidence of mild to moderate respiratory disease was present on radiographs and at necropsy. Cynomolgus macaques exposed via the aerosol route also developed the most consistent fever responses and had the most severe respiratory disease and pathology. This study demonstrates that while all four models produced suitable representations of mild COVID-like illness, aerosol exposure of cynomolgus macaques to SARS-CoV-2 produced the most severe disease, which may provide additional clinical endpoints for evaluating therapeutics and vaccines.

Comparison of transcriptional responses between pathogenic and nonpathogenic hantavirus infections in Syrian hamsters using NanoString.

BACKGROUND: Syrian hamsters infected with Andes virus (ANDV) develop a disease that recapitulates many of the salient features of human hantavirus pulmonary syndrome (HPS), including lethality. Infection of hamsters with Hantaan virus (HTNV) results in an asymptomatic, disseminated infection. In order to explore this dichotomy, we examined the transcriptome of ANDV- and HTNV-infected hamsters. RESULTS: Using NanoString technology, we examined kinetic transcriptional responses in whole blood collected from ANDV- and HTNV-infected hamsters. Of the 770 genes analyzed, key differences were noted in the kinetics of type I interferon sensing and signaling responses, complement activation, and apoptosis pathways between ANDV- and HTNV-infected hamsters. CONCLUSIONS: Delayed activation of type I interferon responses in ANDV-infected hamsters represents a potential mechanism that ANDV uses to subvert host immune responses and enhance disease. This is the first genome-wide analysis of hantavirus-infected hamsters and provides insight into potential avenues for therapeutics to hantavirus disease.

Host response transcriptomic analysis of Crimean-Congo hemorrhagic fever pathogenesis in the cynomolgus macaque model.

Crimean-Congo hemorrhagic fever virus (CCHFV) is a highly pathogenic tick-borne RNA virus prevalent in Asia, Europe, and Africa, and can cause a hemorrhagic disease (CCHF) in humans with mortality rates as high as 60%. A general lack of both effective medical countermeasures and a comprehensive understanding of disease pathogenesis is partly driven by an historical lack of viable CCHF animal models. Recently, a cynomolgous macaque model of CCHF disease was developed. Here, we document the targeted transcriptomic response of non-human primates (NHP) to two different CCHFV strains; Afghan09-2990 and Kosova Hoti that both yielded a mild CCHF disease state. We utilized a targeted gene panel to elucidate the transcriptomic changes occurring in NHP whole blood during CCHFV infection; a first for any primate species. We show numerous upregulated genes starting at 1 day post-challenge through 14 days post-challenge. Early gene changes fell predominantly in the interferon stimulated gene family with later gene changes coinciding with an adaptive immune response to the virus. There are subtle differences between viral strains, namely duration of the differentially expressed gene response and biological pathways enriched. After recovery, NHPs showed no lasting transcriptomic changes at the end of sample collection.

Transcriptomic Analysis Reveals Host miRNAs Correlated with Immune Gene Dysregulation during Fatal Disease Progression in the Ebola Virus Cynomolgus Macaque Disease Model.

Ebola virus is a continuing threat to human populations, causing a virulent hemorrhagic fever disease characterized by dysregulation of both the innate and adaptive host immune responses. Severe cases are distinguished by an early, elevated pro-inflammatory response followed by a pronounced lymphopenia with B and T cells unable to mount an effective anti-viral response. The precise mechanisms underlying the dysregulation of the host immune system are poorly understood. In recent years, focus on host-derived miRNAs showed these molecules to play an important role in the host gene regulation arsenal. Here, we describe an investigation of RNA biomarkers in the fatal Ebola virus disease (EVD) cynomolgus macaque model. We monitored both host mRNA and miRNA responses in whole blood longitudinally over the disease course in these non-human primates (NHPs). Analysis of the interactions between these classes of RNAs revealed several miRNA markers significantly correlated with downregulation of genes; specifically, the analysis revealed those involved in dysregulated immune pathways associated with EVD. In particular, we noted strong interactions between the miRNAs hsa-miR-122-5p and hsa-miR-125b-5p with immunological genes regulating both B and T-cell activation. This promising set of biomarkers will be useful in future studies of severe EVD pathogenesis in both NHPs and humans and may serve as potential prognostic targets.