Skip to main content
PMC home page
EPA Author Manuscripts logoLink to EPA Author Manuscripts
. Author manuscript; available in PMC: 2020 Jul 1.
Published in final edited form as: J Microbiol Methods. 2019 May 13;162:21–27. doi: 10.1016/j.mimet.2019.05.005

Development of a rapid viability polymerase chain reaction method for detection of Yersinia pestis

Staci R Kane a, Sanjiv R Shah b,*, Teneile M Alfaro a
PMCID: PMC6555679  NIHMSID: NIHMS1530056  PMID: 31095987

Abstract

Due to the occurrence of natural plague outbreaks and its historical usage as a biological weapon, Yersinia pestis is considered one of the high-priority biological threat agents. It can remain viable in certain environments including water for > 100 days. Because of its slow-growth characteristic, it usually takes three or more days to detect and confirm the identity of viable Y. pestis cells by PCR, serological, or biochemical assays when using the traditional microbiological plate-culture-based analysis, and that too, assuming faster growing microbes present in a water sample do not mask the Y. pestis colonies and interfere with analysis. Therefore, a rapid-viability Polymerase Chain Reaction (RV-PCR) method was developed for detection of Y. pestis. The RV-PCR method combines 24 h-incubation broth culture in a 48-well plate, and pre- and post-incubation differential PCR analyses, thereby allowing for rapid and high-throughput sample analysis compared with the current plate culture method. One chromosomal and two plasmid gene target-based real-time PCR assays were down-selected, showing ca. 10 genome equivalent detection; the chromosomal assay was then used for RV-PCR method development. A 101-cell level (10–99 cells) sensitivity of detection was demonstrated even with complex sample backgrounds including known PCR inhibitors (ferrous sulfate and humic acid), as well as metal oxides and microbes present in Arizona Test Dust (ATD). The method sensitivity was maintained in the presence of dead Y. pestis cells up to 104 cells per sample. While affording high-throughput and rapid sample analysis, the 48-well plate format used in this method for sample enrichment significantly reduced labor requirements and generation of BioSafety Level-3 (BSL-3) laboratory waste as compared to the usual microbiological plate-culture-based methods. This method may serve as a model for other vegetative bacterial pathogens.

Keywords: Yersinia pestis, Detection, Viability, RV-PCR, Water contamination, Bioterrorism, Plague

1. Introduction

Yersinia pestis, the causative bacterium of plague, was responsible for the catastrophic loss of human life during the pandemics of 541 CE, 1346, and 1855 (Perry and Fetherston, 1997; Slack, 1989). Plague outbreaks continue to be a threat in many parts of the world, especially in Africa, and in particular Madagascar, which reported many fatal cases in 2017 (World Health Organization, 2017). Its high lethality made Y. pestis an attractive biological warfare agent (Inglesby et al., 2000; Derbes, 1966) and a potential agent of bioterrorism (Khan et al., 2001; Riedel, 2005). It is a Category A biological select agent in the CDC Critical Agents List generated in conjunction with medical, public health, and intelligence agencies (Centers for Disease Control and Prevention, 2000; Khan et al., 2000). Y. pestis, a Gram-negative, non-motile, capsule-forming, non-spore-forming coccobacillus bacterium causes three primary forms of plague disease in humans depending upon the route of infection: bubonic, septicemic, and pneumonic plague. Although rare, it can cause gastroenteritis via ingestion of infected animal meat (Bin Saeed et al., 2005; Leslie et al., 2011). The primary reservoir for Y. pestis is small rodents, with fleas typically responsible for animal-to-human transmission (Stenseth et al., 2008). In addition, recent studies have suggested that amoeba in soil or water may be competent environmental reservoirs thus enhancing its survival and transmission (Markman et al., 2018).

While Y. pestis is not a water-borne pathogen, it is a potential bio-threat agent for water (Khan et al., 2001). Y. pestis in water could cause septicemic and bubonic plague upon entry through open cuts or wounds. Additionally, it is possible that pneumonic plague could occur from inhalation of aerosols derived from water intentionally contaminated with Y. pestis (Tang et al., 2006). It has also been shown that a fatal systemic disease occurred when mice were administered Y. pestis-contaminated drinking water (Butler et al., 1982). Long-term survival of Y. pestis in water can pose a potential problem of prolonged disease transmission. Y. pestis has been shown to remain culturable from 2 to 21 days (Wilm, 1897; Pawlowski et al., 2011; Gilbert and Rose, 2012) in tap water, and > 100 days in bottled drinking water (Torosian et al., 2009). Drinking water can be accidentally or intentionally contaminated at its source, during treatment, in the distribution system, or even in bottled form (Khan et al., 2001). Although the implementation of drinking water standards has played a significant role in reducing water-borne illnesses, a serious vulnerability still remains from bioterrorism (Khan et al., 2001).

Current methods for detection of viable Y. pestis in water involving direct plating onto solid growth medium are labor-intensive and low throughput, and also require confirmatory analysis via PCR, serological and/or biochemical tests. For example, each sample is processed individually and requires two dilution tubes, nine growth medium plates for plating dilutions, two Microfunnel filters with a plate per filter, and one enrichment culture tube; thus, for 48 samples, one would generate a total of 96 dilution tubes, 48 enrichment cultures, 96 Microfunnel filters, and 528 plates (or more if restreaking from enrichment culture is needed), followed by confirmatory testing from presumptive Y. pestis colonies or turbid growth in enrichment cultures for each sample. Here we report development and optimization of a Rapid Viability Polymerase Chain Reaction (RV-PCR) method for rapid detection of viable Y. pestis. The RV-PCR method combines shorter sample incubation in liquid culture (compared to plate culture) with real-time PCR analysis before and after incubation and uses the change in real-time PCR response to specifically detect low concentrations of viable Y. pestis. In contrast to the plate culture method, for the RV-PCR method, 48 samples are processed together in one 48-well plate, with two DNA extracts generated per sample, and three PCR analyses conducted per extract, for a total of 288 PCR analyses (less than four 96-well PCR plates) per 48 samples, and no confirmatory testing required since PCR analysis is specific to Y. pestis. For two laboratorians, each working an 8-h shift per 24-h period, it would take about 54 h from sample receipt to results reporting for 48 samples using RV-PCR analysis; whereas, for the traditional culture method, it would take from about 88 h, if isolated Y. pestis colonies could be detected from all 48 samples, to > 160 h if sample enrichment cultures needed to be analyzed to confirm the absence of viable Y. pestis cells.

Previously, an RV-PCR method was developed for detection of viable Bacillus anthracis spores in environmental samples (Létant et al., 2011); however, the sample processing procedure for spores was not conducive to maintaining vegetative cell viability. Therefore, the RV-PCR method was developed for detection of Y. pestis cells, which may serve as a model for other vegetative bacterial pathogens, especially in water samples. In case of an incident, both pre- and post-disinfection water samples could be concentrated by ultrafiltration with or without secondary filtration (Kahler et al., 2015; Holowecky et al., 2009) prior to RV-PCR analysis to further improve the sensitivity of detection. The method described here was shown to enable detection of Y. pestis cells even in backgrounds of high levels of debris, potential inhibitors, non-target microbial cells/spores, and dead Y. pestis cells.

2. Materials and methods

2.1. Bacterial strains, growth conditions, and cell suspension preparation

The pathogenic Y. pestis CO92 strain and the attenuated CO92 pgm− strain lacking the 102-kilobase (kb) pgm locus (Buchrieser et al., 1999) were used. The strains were grown on Brain Heart Infusion (BHI) broth (Becton Dickinson BBL™ Cat. No. 237500) or Y. pestis Enrichment Broth (YPEB) (T. Doran, D. Hanes, S. Weagant, S. Torosian, D. Burr, K. Yoshitomi, K. Jinneman, R. Penev, O. Adeyemo, D. Williams-Hill, P. Morin [Food and Drug Administration], personal communication). The YPEB medium was used for RV-PCR experiments because it produced higher cell yields for shorter incubation times. Tryptose Blood Agar (TBA) plates without blood prepared from Difco™ Tryptose Blood Agar Base powder (Becton Dickinson, Cat. No. 223220) were used to isolate colonies for liquid culturing. The YPEB medium consisted of the following components (per/L): 25 g Bacto Heart Infusion Broth powder; 6 g Yeast extract; 3 g Soytone; 0.5 g Ferric Ammonium Sulfate; and 8.77 g 3-(N-Morpholino) propanesulfonic acid (MOPS). The solution was sterilized by passage through a 0.22-μm filter. Y. pestis cell suspensions were prepared from 2 to 3 colonies on TBA plates (started from −80 °C stocks), inoculated into 5-mL YPEB in 50-mL conical tubes. After overnight incubation (18–26 h) at 28 or 30 °C at 180 rpm, cells were harvested (3100 ×g at 4 °C for 15 min) and resuspended in phosphate-buffered saline (PBS, Teknova, Cat. No. P0261). Cells were then adjusted to an optical density at 600nm (OD600) of ca. 0.1 with PBS, corresponding to ca. 6–7 × 106 CFU/mL. Ten-fold serial dilutions were performed in PBS to achieve the desired starting cell density (colony-forming units, CFU/mL) in 2.7 mL sample per well of a 48-well, 5-mL rectangular well plate (E&K Scientific, Cat. No. EK-2044). One part 10 × (10-fold concentrated) YPEB medium (0.3 mL) was added to nine parts cell suspension in PBS (2.7 mL) in 48-well plates, yielding 1 × YPEB. Culture data are shown as either CFU/mL or CFU/2.7 mL sample (corrected for dilution) based on the average (Avg) and standard deviation (SD) of triplicate plates with colony counts within the range of 25–250 CFU/plate. Spent cultures and consummables were autoclaved at 15 p.s.i. for 60 min at 121 °C.

2.2. Sample types and preparation

PBS was used as a substitute for water samples because it maintained cell viability and represented a reproducible matrix in terms of pH and chemical composition to facilitate consistent experimental results during the RV-PCR method development. Materials were added to PBS including i) iron sulfate (heptahydrate; Sigma-Aldrich, Cat. No. 215422) and humic acid (HA; Sigma-Aldrich, Cat. No. 53680-10G) to represent chemical interferences, ii) Arizona Test Dust (ATD; ISO 12103-1, A3 Medium Test Dust; Powder Technology, Arden Hills, MN) to represent chemical, biological (live, non-target microorganisms), and physical challenges (particulates), and iii) dead Y. pestis cells to assess the background effect for post-disinfection applications or natural degradation. Iron sulfate and HA solutions were prepared in sterile distilled, deionized (DD) water and added to 2.7 mL water samples at a final concentration of 10 μg Fe2+/mL (27 μg Fe2+/sample) and 50 μg HA/mL (135 μg HA/sample), respectively. These concentrations were at the upper end of the range of values expected for drinking water samples (National Research Council, 1979; World Health Organization, 1996; US EPA, 2005). The dust was previously shown to contain ca. 5 × 104 CFU background microbes including fungi and bacterial spores per 10 mg (Rose et al., 2011). Dust was non-sterilized, made into a slurry in DD water, and added to samples at a final concentration of 4 mg/mL (10.8 mg/sample), which is within the range for total suspended solids typically measured in water samples.

2.3. Preparation of killed Y. pestis cell suspensions

Isopropyl alcohol (IPA) exposure was used to generate dead cells for evaluating the RV-PCR method since it has been used to generate dead cells as controls for viable cell staining kits (Molecular Probes, 2004). An overnight culture (100 mL) of Y. pestis CO92 pgm− was diluted to OD600 ca. 0.01 in YPEB and incubated with shaking at 30 °C and 180 rpm until an OD600 ca. 0.3–0.4 was achieved. The culture was then split into four 20-mL aliquots, harvested by centrifugation (3100 ×g at 4 °C for 15 min), the supernatant was removed, and the cell pellets were suspended in 6 mL PBS. For IPA-treated cells, 14 mL 99 + % IPA were added to yield ca. 70% IPA, and 14 mL PBS were added for the control treatment. The cell suspensions were incubated for 2 h at room temperature with gentle mixing every 30 min. The suspensions were then centrifuged at 3100 ×g at 4 °C for 15 min. The supernatant was removed and the pellets were washed in 20 mL PBS followed by centrifugation, removal of supernatant, and final suspension in PBS to 20 mL. The IPA-killed cell suspensions were determined from untreated controls to be ca. 4.1 × 107 cells per mL. Suspensions were divided into aliquots, stored at 4 °C, and used within 40 days of generation. Before use, the IPA-treated cell suspensions as well as cell pellets and super-natants (after centrifugation) were heat-lysed and tested by PCR to ensure that DNA was not being lost/degraded over time. Heat lysis was conducted at 95 °C for 5 min followed by placement on ice for 2 min, centrifugation (20,800 ×g at 4 °C for 5 min), and removal of liquid for PCR analysis (leaving the cell debris pellet in the tube). The YC2 assay (Table 1) was used for PCR analysis for the different fractions. A comparison of real-time PCR results from heat lysates of cell suspensions, pellets and supernatants from IPA-treated and untreated cells showed similar DNA contents suggesting little to no loss of DNA due to IPA treatment (data not shown). For experiments with different concentrations of dead cells, the IPA-treated suspension was diluted with PBS to achieve the desired dead cell level based on plate counts from the control processed in parallel.

Table 1.

Nucleotide sequences of the primer/probe sets used for Y. pestis RV-PCR analysis.

Assay Forward primera Reverse primera Probea,b Amplicon length (bp)
YC2 CAACGACTAGCCAGGCGAC CATTGTTCGCACGAAACGTAA TTTTATAACGATGCCTACAACGGCTCTGCAA 78
YpP1 TGGGTTCGGGCACATGATA CCAGCGTTAATTACGGTACCATAA CTTACTTTCCGTGAGAAGACATCCGGCTC 101
YpMT1 GGTAACAGATTCGTGGTTGAAGG CCCCACGGCAGTATAGGATG TCCCTTCTACCCAACAAACCTTTAAAGGACCA 99
Open in a new tab
a

Sequences are listed in the 5′ to 3′ orientation.

b

Probes were labeled at the 5′ end with FAM (6-carboxyfluorescein) and at the 3′ end with Black Hole Quencher® (BHQ-1); bp = base pair.

From reference plating, the actual live cells were 460 ± 100 per 2.7-mL water sample for the 102 live cell level (100–999 cells) and 46 ± 10 per 2.7 mL water sample for the 101 live cell level (10–99 cells), and dead cell levels ranged from 104 to 106 per sample. Control treatments without dead cells were processed in parallel. Aliquots were processed for DNA recovery at T0 and T24 and analyzed using the YC2 chromosomal assay with undiluted and 10-fold diluted DNA extracts to check for PCR inhibition.

2.4. Rapid-viability PCR method

The RV-PCR method for Bacillus anthracis (Létant et al., 2011) employed multiple vacuum filtration steps to first concentrate the spores in a filter cup and then wash twice with different buffers. Unlike ultra-filtration, this filtration resulted in complete drying of the filter and could not be used for vegetative cells while maintaining viability; therefore, the cell suspension in PBS was mixed with 10×-concentrated growth medium. Specifically, a 2.7 mL sample aliquot was added to 0.3 mL 10× YPEB in 5-mL wells of a 48-well plate. After mixing, a 500-μL aliquot was removed from each well before incubation (T0 aliquot), transferred to a 2-mL Eppendorf tube, and centrifuged at 20,800 ×g for 10 min at 4 °C, after which 300-μL supernatant were removed and discarded. The cell pellets in the remaining 200-μL were frozen prior to DNA extraction and PCR analysis. Additionally, 250-μL aliquots were also evaluated with processing as described except only 50-μL super-natant was removed and discarded, and the cell pellets in remaining 200-μL were frozen prior to DNA extraction and PCR analysis. The 48-well plate was then sealed with a sterile AeraSeal™ breathable adhesive seal (Excel Scientific, Cat. No. BS-25), incubated for different time periods from 12 to 40 h at 28 or 30 °C with shaking at 180 rpm, and removal of 500-μL aliquots for the different time points. Aliquots were processed as described and either stored at −20 °C prior to DNA extraction or processed immediately.

The MagneSil® Blood Genomic, Max Yield System (Promega, Cat. No. MD1360) was used for DNA extraction and purification. This kit enables DNA recovery from multiple complex samples simultaneously using a magnetic bead-based cleanup method. The procedure was modified from that used for B. anthracis cells (Létant et al., 2011). Briefly, the cell pellet in the remaining 200-μL aliquot was thawed and 800-μL Lysis Buffer were added. The mixture was vortex mixed and incubated for 5 min. Next, 600-μL of paramagnetic particle (PMP) mix were added and mixed by vortexing. The liquid was then removed after placing tubes on the magnetic rack. One lysis wash step with 360-μL of Lysis Buffer was included, followed by vortex mixing, placing on the magnetic rack, and subsequent liquid removal. Two washes with 360-μL of Salt Wash were then performed, in each case followed by mixing by vortexing and removal of the liquid when on the magnet. Finally, two washes with 500-μL of Alcohol Wash solution were performed with mixing by vortexing and liquid removal. A final wash with 70% ethanol was included to enhance PMP drying. PMPs were air-dried for 2 min and then dried at 80 °C for 20 min. DNA was then eluted by addition of 200-μL Elution Buffer followed by five cycles of vortexing (5–10 s) and heating (1 min) at 80 °C. Each sample DNA extract was kept at room temperature for 5 min prior to mixing and transferring to the magnetic rack. While on the magnet, the DNA extract was recovered and transferred into a clean 1.5-mL Eppendorf tube. If particles remained, the sample DNA extract was centrifuged at 20,800 ×g for 5 min at 4 °C, and the supernatant was transferred to a clean Eppendorf tube. The DNA extract was stored at −20 °C until real-time PCR analysis.

2.5. Y. pestis CO92 real-time PCR analysis

Y. pestis CO92 DNA standards were generated from harvested 5-mL YPEB cultures. A Master Pure™ Complete DNA and RNA (ribonucleic acid) Purification Kit (Epicentre® Biotechnologies Inc. Cat. No. MC85200) was used to extract DNA from pure culture following the manufacturer’s protocol. DNA was measured using the high sensitivity Quant-iT™ DNA assay (Invitrogen, Cat. No. Q32854) with a Qubit™ fluorometer (Cat. No. Q33216). Standard DNA concentrations prepared in PCR-grade water ranged from 1 fg/μL to 1 ng/μL. Each PCR plate contained seven 10-fold dilutions, ranging from 5 fg per 25-μL PCR to 5 ng per 25-μL PCR.

Candidate real-time PCR primer-probe sets (72 total) were analyzed in silico using GenBank and other sequence databases for predicted specificity to virulent Y. pestis strains and lack of cross-reactivity with near neighbors. Assays were then tested against 12 target DNA templates (Y. pestis strains Pestoides B, F and G; Nairobi; Shasta; A1122; Java 9; Nicholisk 41; Harbin 35; KIM 27; Antigua; CO92) and two near neighbor DNA templates (Y. pseudotuberculosis strains Yp III and PB1/+) to evaluate specificity and sensitivity of assays for the chromosome (YC2 assay), the pMT1 plasmid (YpMT1 assay), and the pPCP1 plasmid (YpP1 assay) (Table 1). The down-selected YpP1 assay targeted the plasminogen activator/outer membrane protease (Pla) gene, while the YpMT1 assay targeted the caf1R gene, a positive regulator of the F1 operon (encoding the F1 capsule antigen) involved in virulence. The YC2 assay targeted a gene encoding an outer membrane auto-transporter barrel domain protein, with similarity to Type V secretory pathway adhesin AidA. Autotransporter proteins of a type V secretion system and these systems have been linked to virulence in Gram-negative bacteria (Derbise et al., 2010).

The PCR mix contained TaqMan® 2× Universal PCR Master Mix (ThermoFisher, Cat. No. 4304437) including deoxynucleotide triphosphates (dNTPs), a 6-Carboxyl-X-Rhodamine (ROX), AmpErase® UNG (uracil-N-glycosylase), as well as forward and reverse primers and a probe labeled at the 5′ end with FAM (6-carboxyfluorescein) and at the 3′ end with Black Hole Quencher® (BHQ-1) (Table 1). PCR-grade water was used to make the mix volume up to 20-μL per reaction and 5-μL of sample DNA extract or DNA standard were added. The following thermal cycling conditions were used on an Applied Biosystems® (ABI) 7500 Fast Real-Time PCR instrument: 2 min at 50 °C for UNG incubation, 10 min at 95 °C for DNA polymerase activation, and 45 amplification cycles (5 s at 95 °C for denaturation and 20 s at 60 °C for annealing/extension). Both undiluted and 10-fold diluted sample DNA extracts (prepared in PCR-grade water) for both T0 and later time points were analyzed to check for PCR inhibition. Three replicate samples were analyzed per experimental condition, and three replicate PCR analyses were conducted per sample replicate. DNA extracts from different time points from the same samples were analyzed on the same plate to minimize variability. The ROX dye in the ABI Universal Master Mix was used to normalize the fluorescent reporter signal. Automatic baseline and threshold settings were used throughout after reviewing the response curves to confirm their appropriateness. Spent PCR plates and consumables were disinfected in a permitted autoclave operated at 15 p.s.i. for 60 min at 121 °C.

2.6. RV-PCR data interpretation

The RV-PCR criteria for positive detection of Y. pestis cells was evaluated, namely ΔCT (CT [T0] − CT [Tf]) ≥ 6, (where f = final incubation time, h). In most cases a 24-h incubation was used, such that Tf = T24. For cases where no PCR response was obtained (non-detect results), the CT values were set to 45 to calculate ΔCT (since 45 PCR cycles were used). A ΔCT ≥ 6 represented an increase in DNA concentration of approximately 2-log, due to the presence of viable cells in the original sample that propagated during incubation. For individual sample replicates within an experiment, the RV-PCR result was considered positive when the average of at least 2 of 3 PCR replicates for T0 and Tf had ΔCT ≥ 6. If a single PCR replicate was positive and the other two replicates were non-detect, the T0 or Tf aliquot was considered negative or non-detect (NDT) and the CT was set to 45, in order to calculate ΔCT. The RV-PCR method sensitivity of detection was equivalent to the Y. pestis cell level where 100% of the samples had a ΔCT ≥ 6; however, this did not account for losses from sampling and sample handling.

2.7. Biosafety

All manipulations with Y. pestis cultures were done under Biosafety Level 3 (BSL-3) conditions in an CDC-permitted facility, including use of a certified Class II biosafety cabinet with thimble connection and ducted exhaust and the following personal protective equipment (PPE): Hood Powered Air Purifying Respirator (PAPR), Tyvek coverall with hood and boots, shoe covers, and double latex or nitrile gloves. Aerosolization risk was mitigated by use of aerosol barrier tips during pipetting and use of gasketed safety cups for centrifugation. Secondary containment was used for capped culture tubes and sealed 48-well plates during incubation. Waste was subjected to two rounds of sterilization using a permitted autoclave, documented at 15 p.s.i. and 121 °C for ≥60 min prior to disposal.

3. Results

3.1. RV-PCR method development for detection of viable Y. pestis

The RV-PCR method developed here uses a differential PCR response before and after sample enrichment to detect viable Y. pestis cells based on a change in cycle threshold (CT) over time or ΔCT (de-fined by CT [T0] − CT [T24]) ≥ 6. This represents an increase in cells containing DNA of ca. 2 log10.

For real-time PCR assay evaluation for RV-PCR, the YpP1, YpMT1, and YC2 assays showed that 50-fg genomic Y. pestis CO92 DNA (ca. 10 genome copies) was detected 100% of the time (n = 21 runs per assay) and the 5-fg level was detected for about 75% of PCR runs (n = 21 per assay). Since the plasmid copy numbers can vary, the YC2 assay (showing similar detection sensitivity) was used for method development. For Y. pestis growth, Brain Heart Infusion (BHI) broth and Y. pestis Enrichment Broth (YPEB) were evaluated in 48-well plates. The BHI broth yielded poor growth (data not shown). There was a 3.6 to 4-log increase after 24-h incubation for the different Y. pestis cell levels on YPEB (Table 2). Similar Y. pestis log increases were only observed with BHI broth if inoculated from three sequential overnight cultures and the incubation period was increased to 40 h. Therefore, the YPEB medium was selected for Y. pestis RV-PCR method development.

Table 2.

Growth of Y. pestis cells in 48-well plates (1× YPEB).

Starting cell level (CFU/mL) Actual CFU/mL/timepoint Log increase
Avg (SD)a
0 h 24 h
100 2.8 (0.1) × 100 2.4 (1.2) × 104 4.0
101 2.8 (0.1) × 101 1.7 (0.6) × 105 3.9
102 2.8 (0.1) × 102 1.5 (0.1) × 106 3.8
103 2.8 (0.1) × 103 8.5 (2.0) × 106 3.6
Open in a new tab
a

Data points show the average (Avg) and standard deviation (SD) from triplicate analyses for inoculum reference plating (0 h) and after 24 h incubation, corrected for dilution.

Two aliquot volumes from 48-well cultures were evaluated for DNA extraction, 250-μL and 500-μL, in order to achieve sufficient detection levels for RV-PCR analysis. In addition, two incubation periods, 12 h and 24 h, were tested for both aliquot volumes (for time points T0 and T12 or T0 and T24). Culture data showed an average ca. 4.2–4.7 log increase over a 24-h period and ca. 1.7–2.0-log increase during 12-h incubation (Table 3). As expected, RV-PCR analysis showed that the 500-μL aliquot volume for DNA extraction gave higher average ΔCT values than the 250-μL aliquot volume, although differences were not statistically significant (p-values ranged from 0.1 to 0.9). Based on these data, the 500-μL aliquot volume was used for DNA extraction for subsequent RV-PCR method evaluation. Although the average ΔCT values were approximately 6 or greater after 12-h incubation with 10× YPEB diluted to 1× with phosphate-buffered saline (PBS), a longer incubation period was selected for water samples containing challenge material.

Table 3.

Effect of aliquot volume on Y. pestis cell growth and ΔCT for RV-PCR analysis.

Starting CFU level/samplea Aliquot volume μL YC2 assay ΔCTb Avg (SD) Avg change in log cell density
T0 - T12 T0 - T24 T0 - T12 T0 - T24
103 250 6.1 (0.8) 13.2 (2.3) 2.1 4.4
500 6.3 (0.6) 15.3 (0.6) 2.0 4.2
102 250 5.9 (1.9) 16.0 (0.7) 1.9 4.6
500 6.9 (1.1) 16.8 (0.9) 1.7 4.6
101 250 7.3 (1.1) 17.7 (0.7) 1.6 4.7
500 9.6 (0.8) 19.8 (0.6) 1.8 4.4
Open in a new tab
a

Per 2.7-mL water sample, there were ca. 38, 380, and 3800 cells, from reference plating, for the 101-, 102-, and 103-cell levels, respectively.

b

Average (Avg) and standard deviation (SD) were based on triplicate samples.

3.2. RV-PCR method performance with complex samples

Throughout the course of method development and evaluation, PBS was used as a substitute for actual water samples in order to maintain cell viability and reduce sample variability to obtain reproducible results. As per EPA protocol (US EPA, 2017), a large volume water sample (1–2 L) is typically collected and concentrated onto filter media, after which bacterial contaminants are recovered from the filter by washing with PBS for subsequent analysis. Results for RV-PCR method evaluation with PBS containing Arizona Test Dust (ATD) (10.8 mg/sample) or humic acid (HA; 135 μg/sample) plus iron (27 μg Fe2+ as FeSO4 per sample) are shown in Table 4. For the 102-cell level, there was little to no PCR inhibition for the control treatment with similar ΔCT values (CT [T0] − CT [T24]) for undiluted and 10-fold diluted DNA extracts; however, the treatment with iron and HA showed inhibition for 1 of the 3 replicates for undiluted extracts, while the 10-fold diluted extract had all three replicates with ΔCT > 6. The ATD treatment showed similar ΔCT values to those of the control treatment with all values > 6, for either undiluted or 10-fold diluted extracts. For the 101-cell level, there were lower ΔCT values for the ATD treatment compared to the control treatment, especially when comparing the 10-fold diluted DNA extracts. The ΔCT values for the ATD treatment met the criterion for positive detection (ΔCT ≥ 6) for all but one sample replicate for 10-fold diluted extracts (ΔCT = 4.1, while the undiluted extract value was 12.2). For the Fe/HA treatment, 1 of 3 sample replicates showed PCR inhibition (i.e., non-detect), although this was resolved with 10-fold dilution of the sample DNA extract. The data suggested that debris such as the reference test dust could cause growth inhibition, resulting in higher T24 CT values, likely due to the presence of indigenous organisms. The test dust was reported to contain fungal spores as well as Bacillus spores and other bacteria (Rose et al., 2011) that may be faster growing than Y. pestis. However, at these low cell levels, the RV-PCR method still showed the ability to accurately detect live Y. pestis cells in complex backgrounds, with consistent detection at the 101-cell level. The negative controls showed non-detect results for all replicates at both time points (data not shown).

Table 4.

RV-PCR results for 101 or 102 Y. pestis cell level per sample with and without chemical or biological backgrounds.

Treatment/Starting CFU levela Sample replicate YC2 assay CT Avg (SD)b ΔCT (T0 - T24)c
Undiluted 10-fold diluted Undiluted 10-fold diluted
T0 T24 T0 T24
Control/102 1 38.8 (0.7) 23.6 (0.5) NDT 28.2 (0.2) 15.2 16.8
2 37.0 (0.2) 24.3 (0.8) 39.4 (0.4)d 23.2 (1.8) 12.7 16.2
3 37.9 (0.8) 23.0 (0.4) 39.5 (2.1)d 26.6 (0.7) 14.9 12.9
Avg (SD) 37.9 (0.9) 23.6 (0.8) 39.4 (1.4) 26.0 (2.4) 14.3 (1.4) 15.3 (2.1)
Fe/HA 102 1 NDT 27.7 (0.6) NDT 30.4 (0.5) 17.3 14.6
2 43.6 (1.0)d 31.1 (3.3)d 41.4 (0.4)d 28.3 (2.8) 12.7 13.1
3 41.8 (0.3) 38.5 (6.5) 40.2 (0.1)d 26.8 (1.3) 3.3 13.4
Avg (SD) 42.8 (1.4) 32.4 (6.0) 40.8 (0.7) 28.5 (2.2) 11.1 (7.1) 13.7 (0.8)
ATD/102 1 NDT 31.3 (0.3) NDT 32.8 (0.4) 13.7 12.2
2 42.9 (0.3)d 27.0 (0.4) NDT 29.7 (0.3) 15.9 15.3
3 NDT 29.5 (0.5) NDT 32.4 (0.2) 15.5 12.6
Avg (SD) NA 29.2 (1.9) NDT 31.6 (1.5) 15.0 (1.2) 13.4 (1.7)
Control/101 1 39.1 (0.3) 24.4 (0.2) NDT 29.8 (0.7) 14.7 15.2
2 40.6 (0.3) 32.7 (0.5) NDT 33.9 (3.3) 7.9 11.1
3 NDT 23.6 (0.3) NDT 28.0 (0.2) 21.4 17.0
Avg (SD) 39.9 (0.9) 26.9 (4.4) NDT 30.6 (3.1) 14.7 (6.8) 14.4 (3.0)
Fe/HA 101 1 42.3 (0.1)d 33.2 (1.4) NDT 31.1 (0.1) 9.1 13.9
2 NDT 33.9 (0.2) NDT 30.3 (0.1) 11.1 14.7
3 NDT NDT 42.3 (1.7)d 28.3 (0.1) 0.0 14.0
Avg (SD) NA 37.3 (5.8) NA 29.9 (1.3) 6.7 (5.9) 14.2 (0.4)
ATD/101 1 NDT 33.8 (0.4) 40.0 (0.03)d 35.9 (0.5) 12.2 4.1
2 NDT 34.2 (0.2) NDT 37.6 (0.8) 10.8 7.4
3 40.5 (1.3) 32.7 (0.4) NDT 35.5 (0.3) 7.8 9.5
Avg (SD) NA 33.6 (0.7) NA 36.3 (1.1) 10.3 (2.2) 7.0 (2.7)
Open in a new tab
a

Per 2.7-mL water sample, there were ca. 29 cells and 290 cells, from reference plating, for the 101- and 102-cell levels, respectively.

b

Average (Avg) and standard deviation (SD) are from triplicate PCR analyses. Average CT values were only calculated if two or three sample replicates had positive PCR results.

c

Avg. ΔCT was calculated as the average of the individual sample ΔCT values rather than based on the difference of average T0 CT and T24 CT values.

d

Values are from two PCR replicates; the third replicate was non-detect. NDT = Non-detect. NDT set to 45 to calculate ΔCT. ATD = Arizona Test Dust. HA = Humic acid. NA = Not applicable; the average CT was only calculated from two or three samples with positive PCR results.

3.3. RV-PCR method performance in a dead Y. pestis cell background

For application to post-disinfection or natural degradation scenarios with high levels of dead cell backgrounds, the RV-PCR method was evaluated with low levels of live Y. pestis cells in the presence of different levels of isopropanol (IPA)-killed Y. pestis cells; this disinfection method effectively killed cells without DNA damage or loss, thus providing the most challenging test case. The RV-PCR method results for 101 and 102 live cell levels with 104–106 dead cells in 2.7 mL water samples are shown in Table 5, as both undiluted and 10-fold dilutions of T0 and T24 DNA extracts. A comparison of undiluted and diluted extracts from each sample showed an average 3.4 ± 0.9 CT difference (i.e., no PCR inhibition) and demonstrated similar trends in ΔCT values for different live and killed cell level combinations. For undiluted DNA extracts, the 102 live cell level with up to 105 dead cells and the 101 live cell level with up to 104 dead cells were both consistently detected with a 24-h incubation period (ΔCT values ≥ 6), whereas, higher dead cell backgrounds produced generally negative results (avg. ΔCT values of 4.6 ± 0.8 for the 102 live/106 dead cell level treatment and 5.7 ± 0.5 for the 101 live/105 dead cell level treatment). For 10-fold diluted DNA extracts, the 102 live cell level treatment had average ΔCT values ≥ 6 for all dead cell backgrounds; however, the individual ΔCT values were 7.5, 4.9, and 6.8 for 105 dead cells, showing 2 of 3 were positive. The replicate ΔCT values for the 106 dead cell/102 live cell treatment were 11.0, 5.5, and 3.7, showing 1 of 3 were positive. For the 101-live cell level treatment, 3 of 3 were positive with a 104 dead cell background (Avg. ΔCT = 7.8 ± 1.0), while none were positive with 105 dead cells (Avg. ΔCT = 4.9 ± 0.5). For these levels of dead cells (104–106), the control treatments with no live cells showed non-detect results for triplicate analyses as expected.

Table 5.

RV-PCR results for 101 and 102 live Y. pestis cell level/sample with different killed target cell concentrations.

Sample type Undiluted extracts YC2 assay CT
Avg ± SDb 		ΔCT Avg ± SD (Positive Replicates) 10-fold diluted extracts YC2 assay CT
Avg ± SDb 		ΔCT Avg ± SD (Positive Replicates)
Live cell levela Dead cell level T0 T24 T0 T24
102 0 38.3 ± 1.8 21.6 ± 0.6 16.7 ± 1.2 (3 of 3) 39.3, NDT, NDT 26.5 ± 1.6 16.6 ± 1.5 (3 of 3)
104 33.4 ± 0.3 22.5 ± 0.7 10.9 ± 1.1 (3 of 3) 37.2 ± 1.3 26.7 ± 1.8 10.5 ± 1.6 (3 of 3)
105 30.6 ± 0.9 23.2 ± 0.6 7.4 ± 0.8 (3 of 3) 32.7 ± 2.2 26.3 ± 1.1 6.4 ± 1.4d (2 of 3)
106 26.4 ± 0.3 21.8 ± 0.9 4.6 ± 0.8 (0 of 3) 31.8 ± 2.7 25.1 ± 1.1 6.7 ± 3.8d (1 of 3)
101 0 NDT, 40.4, NDT 24.3 ± 0.8 19.2 ± 2.5 (3 of 3) NDT, NDT, NDT 27.7 ± 0.7 17.3 ± 0.8 (3 of 3)
104 33.8 ± 0.4 26.3 ± 0.7 7.5 ± 0.8 (3 of 3) 37.1 ± 0.8 29.3 ± 1.0 7.8 ± 1.0 (3 of 3)
105 30.3 ± 0.5 24.6 ± 0.6 5.7 ± 0.5 (1 of 3) 33.2 ± 0.4 28.3 ± 0.6 4.9 ± 0.5 (0 of 3)
0 0 NDT, NDT NDT, NDT 0 ± 0c (0 of 2) NDT, NDT NDT, NDT 0 ± 0c (0 of 2)
104 34.5 ± 0.3 33.4 ± 0.4 1.1 ± 0.3c (0 of 2) 36.0 ± 1.6 36.5 ± 0.4 − 0.5 ± 2.1c (0 of 2)
105 30.2 ± 0.6 29.9 ± 0.5 0.4 ± 0.6 (0 of 3) 33.8 ± 0.7 33.1 ± 0.7 0.7 ± 0.4 (0 of 3)
106 26.2 ± 0.9 26.3 ± 0.5 −0.1 ± 0.5 (0 of 3) 29.5 ± 0.7 29.2 ± 0.1 0.3 ± 0.7 (0 of 3)
Open in a new tab
a

Per 2.7-mL water sample, there were ca. 46 cells and 460 cells, from reference plating, for the 101- and 102-cell levels, respectively.

b

Average (Avg) and standard deviation (SD) based on triplicate PCR analyses of undiluted and 10-fold diluted extracts from triplicate samples except where noted. NDT = Non-detect. NDT set to 45 to calculate ΔCT. The Avg ± SD for T0 or T24 was not calculated if some of the replicates were NDT.

c

Avg ± SD based on two sample replicates.

d

Denotes case where average ΔCT value exceeds the criterion for positive detection, although individual replicate ΔCT values were not all ≥6.

4. Discussion

A RV-PCR method which was developed for detection of viable Y. pestis cells, employs PCR analysis before and after incubation of a water sample mixed with concentrated growth medium. With optimized procedures for high throughput culturing and DNA extraction/purification, the method showed good accuracy and sensitivity of detection even with samples containing potential inhibitors. Consistent growth was observed with YPEB in place of BHI broth (≥4-log cell growth over 24 h) using 10× YPEB diluted to 1× concentration with a water sample. Presence of iron and HA did not impact the 101-cell level sensitivity of detection with a 24-h sample incubation (especially with 10-fold DNA extract dilution). While the presence of reference dust (ca. 10 mg/sample) apparently showed growth inhibition, the 101-cell level sensitivity of detection was still maintained.

In case of Y. pestis water contamination, large volume samples could be concentrated prior to analysis (Kahler et al., 2015; Holowecky et al., 2009), potentially also concentrating insoluble growth and PCR inhibitors, as well as dead Y. pestis cells (from disinfection or natural degradation). The RV-PCR method with 24-h sample incubation showed detection of 101- to 102-cell levels in killed cell backgrounds up to 104 and 105, respectively, while for higher killed cell backgrounds, detection was inconsistent; this could potentially be mitigated with a longer sample incubation period (i.e., 30 h). While these dead cell levels were quite high, they could be present upon sample concentration, especially for post-disinfection scenarios. In addition to a longer incubation for concentrated water samples, both undiluted and 10-fold diluted DNA extracts could be used to address inhibition and minimize false positive/false negative results. The method did not produce false positive results for high concentrations of dead cells alone.

This effort served to bracket the conditions where RV-PCR analysis could be used for detection of live Y. pestis cells in pre- and post-disinfection, or natural degradation scenarios. Furthermore, the method could be a model for vegetative cells of other bacterial pathogens including both bioterrorism and public health threats. The RV-PCR method is expected to have an advantage over traditional culture methods since isolated Y. pestis colonies (or other fastidious bacterial pathogens) may be difficult to detect in samples containing high concentration of non-target microbes. Furthermore, since Y. pestis has been shown to become viable but not culturable (VBNC) in water (Pawlowski et al., 2011; Gilbert and Rose, 2012; Suchkov et al., 1997), it is likely that VBNC cells could be more readily detected from liquid culture used in RV-PCR analysis than from solid media used in traditional plate culture (Wai et al., 2000; Miller and Davey, 1965). The former method could provide better conditions for cell resuscitation, and thus, avoid false negative results. In addition, the RV-PCR method could provide results in less than half the time of the traditional culture method, which requires 72 h or longer for confirmed results (Riedel, 2005). Integration of automated DNA extraction procedures with RV-PCR analysis could further reduce the labor and time-to-results. Finally, RV-PCR uses a single multi-well plate for 48 samples and controls, thereby generating less waste and comprising a smaller laboratory footprint for analysis relative to the traditional culture method that uses numerous petri plates, and dilution and enrichment culture tubes per sample. While the reported Y. pestis real-time PCR assays demonstrated ca. 10 genome-equivalent detection sensitivity, other Y. pestis real-time PCR assays could be integrated into the RV-PCR method as well as other quantitative PCR platforms. Future work to evaluate the RV-PCR method with actual water samples including tap/drinking water, both small (50–100 mL) and large volume (100 L concentrated using ultra-filtration) samples; ground water; source water; and waste water will significantly expand the method utility. Additionally, testing of other variables including sample holding time at different temperatures will guide real-world sample analysis requirements. Due to the shortened time to results, this viability detection method would enhance response capabilities for bioattacks or natural plague outbreak scenarios. More rapid results with the same or improved accuracy compared to plating methods will aid decision-makers in planning disinfection/decontamination efforts and determining their efficacy, thereby enabling safe, timely restoration.

Acknowledgments

We thank Worth Calfee (EPA) and Latisha Mapp (EPA) for technical review of this manuscript.

The United States Environmental Protection Agency through its Office of Research and Development funded the research described here under an Interagency Agreement (EPA IA DW-89-92328201-0). The DOE contractor role did not include establishing Agency policy. It has been subjected to the Agency’s review and has been approved for publication. Note that approval does not signify that the contents necessarily reflect the views of the Agency. Mention of trade names, products, or services does not convey official EPA approval, endorsement, or recommendation.

Lawrence Livermore National Laboratory is operated by Lawrence Livermore National Security (LLNS), LLC, for the U.S. Department of Energy, National Nuclear Security Administration under Contract DE-AC52-07NA27344.

Footnotes

Declarations of interest

None.

References

  1. Bin Saeed AA, Al-Hamdan NA, Fontaine RE, 2005. Plague from eating raw camel liver. Emerg. Infect. Dis 11, 1456–1457. 10.3201/eid1109.050081. [DOI] [PMC free article] [PubMed] [Google Scholar]
  2. Buchrieser C, Rusniok C, Frangeul L, Couve E, Billault A, Kunst F, Carniel E, Glaser P, 1999. The 102-kilobase pgm locus of Yersinia pestis: sequence analysis and comparison of selected regions among different Yersinia pestis and Yersinia pseudotuberculosis strains. Infect. Immun 67 (9), 4851–4861. https://www.ncbi.nlm.nih.gov/pmc/articles/PMC96819/pdf/ii004851.pdf. [DOI] [PMC free article] [PubMed] [Google Scholar]
  3. Butler T, Fu Y, Furman L, Almeida C, Almeida A, 1982. Experimental Yersinia pestis infection in rodents after intragastric inoculation and ingestion of bacteria. Infect. Immun 36, 1160–1167. https://iai.asm.org/content/iai/36/3/1160.full.pdf. [DOI] [PMC free article] [PubMed] [Google Scholar]
  4. Centers for Disease Control and Prevention, 2000. Biological and chemical terrorism: strategic plan for preparedness and response, recommendations of the CDC Strategic Planning Workgroup. MMWR Morb. Mortal. Wkly. Rep 49 (RR-4), 1–14. https://www.cdc.gov/mmwr/PDF/rr/rr4904.pdf. [PubMed] [Google Scholar]
  5. Derbes VJ, 1966. De Mussis and the great plague of 1348. A forgotten episode of bacteriological warfare. J. Am. Med. Assoc 196, 59–62. 10.1001/jama.1966.03100140113030. [DOI] [PubMed] [Google Scholar]
  6. Derbise A, Chenal-Francisque V, Huon C, Fayolle C, Demeure CE, Chane-Woon-Ming B, Médigue C, Hinnebusch BJ, Carniel E, 2010. Delineation and analysis of chromosomal regions specifiying Yersinia pestis. Infect. Immun 78 (9), 3930–3941. 10.1128/IAI.00281-10. [DOI] [PMC free article] [PubMed] [Google Scholar]
  7. Gilbert SE, Rose LJ, 2012. Survival and persistence of nonspore-forming biothreat agents in water. Lett. Appl. Microbiol 55, 189–194. 10.1111/j.1472-765X.2012.03277.x. [DOI] [PubMed] [Google Scholar]
  8. Holowecky PM, James RR, Lorch DP, Straka SE, Lindquist HAD, 2009. Evaluation of ultrafiltration cartridges for a water sampling apparatus. J. Appl. Microbiol 106, 738–747. 10.1111/j.1365-2672.2008.04019.x. [DOI] [PubMed] [Google Scholar]
  9. Inglesby TV, Dennis DT, Henderson DA, Bartlett JG, Ascher MS, Eitzen E, Fine AD, Friedlander AM, Hauer J, Koerner JF, Layton M, McDade J, Osterholm MT, O’Toole T, Parker G, Perl TM, Russell PK, Schoch-Spana M, Tonat K, Working Group on Civilian Biodefense, 2000. Plague as a biological weapon: medical and public health management. J. Am. Med. Assoc 283, 2281–2290. 10.1001/jama.283.17.2281. [DOI] [PubMed] [Google Scholar]
  10. Kahler AM, Johnson TB, Hahn D, Narayanan J, Derado G, Hill VR, 2015Evaluation of an ultrafiltration-based procedure for simultaneous recovery of diverse microbes in source waters. Water 7, 1202–1216. 10.3390/w7031202. [DOI] [PMC free article] [PubMed] [Google Scholar]
  11. Khan AS, Morse S, Lillibridge S, 2000. Public-health preparedness for biological terrorism in the USA. Lancet 2000 (356), 1179–1182. 10.1016/S0140-6736(00)02769-0. [DOI] [PubMed] [Google Scholar]
  12. Khan AS, Swerdlow DL, Juranek DD, 2001. Precautions against biological and chemical terrorism directed at food and water supplies. Public Health Rep 116, 3–14. 10.1016/S0033-3549(04)50017-1. [DOI] [PMC free article] [PubMed] [Google Scholar]
  13. Leslie T, Whitehouse CA, Yingst S, Baldwin C, Kakar F, Mofleh J, Hami AS, Mustafa L, Omar F, Ayazi E, Rossi C, Noormal B, Ziar N, Kakar R, 2011. Outbreak of gastroenteritis caused by Yersinia pestis in Afghanistan. Epidemiol. Infect 139, 728–735. 10.1017/S0950268810001792. [DOI] [PubMed] [Google Scholar]
  14. Létant S, Murphy G, Alfaro T, Avila J, Kane S, Raber E, Bunt T, Shah S, 2011. Rapid viability polymerase chain reaction for detection of virulent Bacillus anthracis from environmental samples. Appl. Environ. Microbiol 77, 6570–6578. 10.1128/AEM.00623-11. [DOI] [PMC free article] [PubMed] [Google Scholar]
  15. Markman DW, Antolin MF, Bowen RA, Wheat WH, Woods M, Gonzalez-Juarrero M, Jackson M, 2018. Yersinia pestis survival and replication in potential ameba reservoir. Emerg. Infect. Dis 24 (2), 294–302. 10.3201/eid2402.171065. [DOI] [PMC free article] [PubMed] [Google Scholar]
  16. Miller RJ, Davey CB, 1965. Influence of the microphysical structure of a system on the growth of bacteria. Can. J. Microbiol 11, 761–763. 10.1139/m65-103. [DOI] [PubMed] [Google Scholar]
  17. Probes Molecular, 2004. LIVE/DEAD® BacLight™ Bacterial Viability Kits https://tools.thermofisher.com/content/sfs/manuals/mp07007.pdfpp.31.
  18. National Research Council (NRC), 1979. Iron A Report of the Subcommittee on Iron, Committee on Medical and Biologic Effects of Environmental Pollutants, Division of Medical Sciences, Assembly of Life Sciences University Park Press, Baltimore: (248 pp.). [Google Scholar]
  19. Pawlowski DR, Metzger DJ, Raslawsky A, Howlett A, Siebert G, Karalus RJ, Garrett S, Whitehouse CA, 2011. Entry of Yersinia pestis into the viable but nonculturable state in a low-temperature tap water microcosm. PLoS One 6, 1–9. 10.1371/journal.pone.0017585. [DOI] [PMC free article] [PubMed] [Google Scholar]
  20. Perry RD, Fetherston JD, 1997. Yersinia pestis—etiologic agent of plague. Clin. Microbiol. Rev 10, 35–66. https://www.ncbi.nlm.nih.gov/pmc/articles/PMC172914/pdf/100035.pdf. [DOI] [PMC free article] [PubMed] [Google Scholar]
  21. Riedel S, 2005. Plague: from natural disease to bioterrorism. Proc. (Baylor Univ. MedCent.) 18, 116–124. https://www.ncbi.nlm.nih.gov/pmc/articles/PMC1200711/. [DOI] [PMC free article] [PubMed] [Google Scholar]
  22. Rose LJ, Hodges L, O’Connell H, Noble-Wang J, 2011. National validation study of a cellulose sponge wipe-processing method for use after sampling Bacillus anthracis spores from surfaces. Appl. Environ. Microbiol 77, 8355–8359. 10.1128/AEM.05377-11. [DOI] [PMC free article] [PubMed] [Google Scholar]
  23. Slack P, 1989. The black death past and present. 2. Some historical problems. Trans. R. Soc. Trop. Med. Hyg 83, 461–463. 10.1016/0035-9203(89)90247-2. [DOI] [PubMed] [Google Scholar]
  24. Stenseth NC, Atshabar BB, Begon M, Belmain SR, Bertherat E, Carniel E, Gage KL, Leirs H, Rahalison L, 2008. Plague: past, present, and future. PLoS Med 5 (1), e3 10.1371/journal.pmed.0050003. [DOI] [PMC free article] [PubMed] [Google Scholar]
  25. Suchkov IuG, Khudiakov IV, Emel’ianenko EN, Levi MI, Pushkareva VI, Suchko IIu Litvin VIu, Kulesh EV, Zotova RS, Gintsburg AL, 1997. The possibility of preserving the causative agent of plague in resting (nonculturable) form. Zh Mikroiol. Epidemiol. Immunobiol 4, 42–46. [PubMed] [Google Scholar]
  26. Tang JW, Li Y, Chan PKS, Ridgway GL, 2006. Factors involved in the aerosol transmission of infection and control of ventilation in healthcare premises. J. Hosp. Inf 64, 100–114. https://www.ncbi.nlm.nih.gov/pubmed/16916564. [DOI] [PMC free article] [PubMed] [Google Scholar]
  27. Torosian SD, Regan PM, Taylor MA, Margolin A, 2009. Detection of Yersinia pestis over time in seeded bottled water samples by cultivation on heart infusion agar. Can.J. Microbiol 55, 1125–1129. 10.1139/W09-061. [DOI] [PubMed] [Google Scholar]
  28. US EPA, 2005. Method 415.3 Determination of Total Organic Carbon and Specific UVAbsorbance at 254 nm in Source Water. Revision 1.1. EPA/600/R-05/055 https://cfpub.epa.gov/si/si_public_file_download.cfm?p_download_id=525079. [Google Scholar]
  29. US EPA, 2017. Sampling Guidance for Unknown Contaminants in Drinking Water. EPA-817-R-08–003 https://www.epa.gov/waterlabnetwork/sampling-guidance-unknown-contaminants-drinking-water. [Google Scholar]
  30. Wai SN, Mizunoe Y, Takade A, Yoshida SI, 2000. A comparison of solid and liquid media for resuscitation of starvation- and low-temperature-induced nonculturable cells of Aeromonas hydrophila. Arch. Microbiol 173, 307–310. 10.1007/s002030000142. [DOI] [PubMed] [Google Scholar]
  31. Wilm, 1897. On the plague epidemic in Hong Kong in 1896 In: Frawkel C, Rubner M, Thierfelder H (Eds.), Hygienische Rundschau Vol. 7 A. Hirschwald Publishing, Berlin, Germany, pp. 285–300. [Google Scholar]
  32. World Health Organization, 2017. Plague Outbreak Madagascar, External Situation Report 14 http://www.afro.who.int/health-topics/plague/plague-outbreak-situation-reports.
  33. World Health Organization (WHO), 1996. Guidelines for drinking-water quality In: Health Criteria and Other Supporting Information, 2nd ed. vol. 2 World Health Organization, Geneva: (WHO/SDE/WSH/03.04/08). [Google Scholar]

RESOURCES