Journal of Microbiological Methods 96 (2014) 35–41

Contents lists available at ScienceDirect

Journal of Microbiological Methods

journal homepage: www.elsevier.com/locate/jmicmeth

Evaluation of swabs and transport media for the recovery of

Yersinia pestis
Sarah E. Gilbert a, Laura J. Rose b,⁎, Michele Howard b, Meranda D. Bradley b, Sanjiv Shah c, Erin Silvestri c,
Frank W. Schaefer III c, Judith Noble-Wang b
a
North Adams Regional Hospital, North Adams, MA, USA
b
Division of Healthcare Quality Promotion, Centers for Disease Control and Prevention, Atlanta, GA, USA
c
Office of Research and Development, National Homeland Security Research Center, U.S. Environmental Protection Agency, Cincinnati, OH, USA

a r t i c l e i n f o a b s t r a c t

Article history: The Government Accountability Office report investigating the surface sampling methods used during the 2001
Received 20 September 2013 mail contamination with Bacillus anthracis brought to light certain knowledge gaps that existed regarding envi-
Received in revised form 22 October 2013 ronmental sampling with biothreat agents. Should a contamination event occur that involves non-spore forming
Accepted 23 October 2013
biological select agents, such as Yersinia pestis, surface sample collection and processing protocols specific for
Available online 1 November 2013
these organisms will be needed. Two Y. pestis strains (virulent and avirulent), four swab types (polyester,
Keywords:
macrofoam, rayon, and cotton), two pre-moistening solutions, six transport media, three temperatures, two
Yersinia pestis levels of organic load, and four processing methods (vortexing, sonicating, combined sonicating and vortexing,
Transport media no agitation) were evaluated to determine the conditions that would yield the highest percent of cultivable
Swab Y. pestis cells after storage. The optimum pre-moistening agent/transport media combination varied with the
Recovery Y. pestis strain and swab type. Directly inoculated macrofoam swabs released the highest percent of cells into so-
lution (93.9% recovered by culture) and rayon swabs were considered the second best swab option (77.0% recov-
ered by culture). Storage at 4 °C was found to be optimum for all storage times and transport media. In a worst
case scenario, where the Y. pestis strain is not known and sample processing and analyses could not occur until
72 h after sampling, macrofoam swabs pre-moistened with PBS supplemented with 0.05% Triton X-100
(PBSTX), stored at 4 °C in neutralizing buffer (NB) as a transport medium (PBSTX/NB) or pre-moistened with
NB and stored in PBSTX as a transport medium (NB/PBSTX), then vortexed 3 min in the transport medium,
performed significantly better than all other conditions for macrofoam swabs, regardless of strain tested
(mean 12 – 72 h recovery of 85.9–105.1%, p b 0.001). In the same scenario, two combinations of pre-
moistening medium/transport medium were found to be optimal for rayon swabs stored at 4 °C (p b 0.001),
then sonicated 3 min in the transport medium; PBSTX/PBSTX and NB/PBSTX (mean 12–72 h recovery of
83.7–110.1%).
Published by Elsevier B.V.

1. Introduction
After the Bacillus anthracis attacks in the fall of 2001, gaps in the US
bioterrorism preparedness and response were highlighted (US GAO,
2005). A lack of standardized sampling and processing techniques for en-
vironmental samples complicated the assessment of contamination and
clean-up (Canter, 2005; Teshale et al., 2002). The methods were not
well characterized or standardized; therefore, post-decontamination
sampling data offered limited confidence that the buildings were safe to
re-occupy. Several ways that the country could better prepare itself for
a bioterrorist attack were identified. One critical need is for validated sam-
pling methods that could be used by all laboratories in the event of a
bioterrorist incident.
⁎ Corresponding author at: 1600 Clifton Rd NE, MS-C16, Atlanta, GA, 30329, USA.
E-mail address: lrose@cdc.gov (L.J. Rose).
Much effort and resources have been allocated to the development
of detection assays; however, the initial sample collection and process-
ing lags behind in development. Sample processing involves recovery of
the biological agent from the sampling device, and remains the limiting
step in the detection methods, whether the methods use non-culture
assays (e.g., polymerase chain reaction [PCR]), or culture-based assays.
The efficiency and limit of detection (LOD) of an analytical method de-
pend on the use of optimized sampling materials, conditions that can
best maintain the integrity of the sample, and optimized protocols for
extraction of the target organism from the sampling tools. Pathogen re-
covery is highly dependent on the collection device used for sampling
the environment and the surface material that is being sampled
(Buttner et al., 2007; Edmonds et al., 2009).
In this study, previously developed Centers for Disease Control and
Prevention (CDC) methods for recovery of Bacillus anthracis spores
from non-porous surfaces (Hodges et al., 2006; Rose et al., 2004) were
investigated for their application to Yersinia pestis, the causative agent

0167-7012/$ – see front matter. Published by Elsevier B.V.

http://dx.doi.org/10.1016/j.mimet.2013.10.017
36 S.E. Gilbert et al. / Journal of Microbiological Methods 96 (2014) 35–41
for plague. Though one study demonstrated that Y. pestis only remained
viable on surfaces for 5 days or less (Rose et al., 2003), other studies
show that the organism can persist in soil for 40 weeks or more
(Ayyadurai et al., 2008), in municipal water beyond 30 days (Gilbert
and Rose, 2012), and in bottle water for 138 days or more (Torosian et
al., 2009). Swabs may be used for collection and transport of Y. pestis
from a variety of environmental surfaces during an outbreak investiga-
tion. The current work did not investigate recovery of cells from envi-
ronmental surfaces, but focused on post-collection variables (swab
types, transport media and extraction steps). The primary goal of this
project was to determine the best combination of sampling swab, pre-
moistening agent, transport medium, and extraction method for a
high efficiency recovery of Y. pestis vegetative cells from swabs. Due to
the delays encountered with courier services, laboratory capacity and
laboratory scheduling of analysis, a secondary goal was to determine if
the percent recovery of Y. pestis would be affected by the transport
medium holding time (e.g., 24, 48, or 72 h).
2. Materials and methods
2.1. Culture preparations
Virulent Y. pestis CO92 and low-virulent Y. pestis A1122 strains were
obtained from CDC's Division of Vector Borne Diseases, Ft. Collins, CO.
For all swab inoculations, one mL of frozen stock (48 h growth in
Brain Heart Infusion Broth (BHIB) and frozen with 10% glycerol), was
added to a 100 mL flask containing 30 mL of BHIB and incubated in a
shaker-incubator at 25 °C for 26–30 h to reach late log phase growth.
Cells were harvested by centrifugation at 3000 ×g, 4 °C, for 10 min.
The supernatant was decanted and the pellet resuspended in 25 mL of
sterile phosphate buffered saline (PBS). This wash step was repeated
two additional times. After the third wash, the pellet was resuspended
in PBS to have a turbidity equivalent to a 0.5 McFarland standard
(108 CFU/mL). The suspension density was checked using a PBS dilution
series plated on Trypticase Soy Agar containing 5% sheep's blood (TSAII,
Becton Dickinson, Sparks, MD). For inoculation of liquid transport medi-
um, suspensions were created in Butterfield buffer (Becton Dickinson,
Sparks, MD) from 48 h plate growth. One mL of the 105 CFU/mL suspen-
sion was used to inoculate 9 mL of transport medium (resulting in
104 CFU/mL). One hundred μL of the 105 CFU/mL suspension was used
to inoculate the swabs (resulting in 104 CFU/swab).
2.2. Survival in liquid transport media
Six liquid transport media were evaluated for the ability to maintain
viability of Y. pestis A1122: (1) Stuart, Toshach, and Patsula medium,
(2) Cary and Blair without CaCl2 (C&Bmod) medium, (3) Amies medium
without charcoal, (4) Amies medium with charcoal, (5) neutralizing
buffer (NB) and (6) phosphate buffered saline with 0.05% Triton X-
100 (PBSTX) surfactant, (Atlas, 2010). Transport media 1 – 4 are tradi-
tional clinical transport media that are meant to be prepared with
0.2% agar to form a semi-solid mass in a transport tube, neutralizing
buffer (5) is a liquid buffer which is helpful in neutralizing most types
of disinfectants that may be present on the surfaces being sampled,
and PBSTX (6) is a liquid buffer solution typically used to elute organ-
isms from the swab during the laboratory processing. Unlike environ-
mental samples, recovery from clinical swabs does not require
quantitation, but simply preservation of the cells for identification. The
current practice for clinical swabs is to submerge the swab in semi-
solid media to preserve the cells on the swab during transport. For the
purposes of this study and following a bioterror incident, however,
quantitation of the cells recovered from (environmental) swabs was
required. When swabs were removed from semi-solid medium and
processed in a separate tube, significant numbers of Y. pestis cells
remained in the semi-solid medium. Therefore, the agar was removed
from all transport medium formulations. In order to determine the
best transport medium, 1 mL of a 105 CFU/mL suspension was added
to 9 mL of each liquid transport medium (fifteen replicates each). The
tubes were held at 4 °C, 25 °C, or 35 °C over a 96-h holding period,
with 1 mL samples taken at 0, 12, 24, 48, 72 and 96 h. Tubes were
vortexed immediately before samples were taken, and the samples
were diluted in series and spread-plated in triplicate onto TSAII. All
plates were incubated at 25 °C and colonies counted after 48 h. The
total CFU recovered for the sample was determined, and the results
were reported as the mean log10 CFU recovered at each temperature,
time point, transport medium and organism evaluated. To identify the
condition(s) that allowed for the least amount of change in cell number
(growth or death), the percent of samples with either ≤0.5 log10 or
≤0.3 log10 CFU change in recovered cells relative to the T0 (time zero)
CFU recovery was calculated for each transport medium, and tempera-
ture. Based on these results, two optimal transport media were selected
for the remainder of the study.
2.3. Evaluation of swabs and extraction methods
Four types of swab materials were evaluated: (1) cotton (Baxter
Healthcare Corp., Deerfield, IL cat # A5002-5), (2) polyester (Falcon™
# 220690, BD, Franklin Lakes, NJ), (3) macrofoam (Puritan Medical,
Guilford, ME, # 25-1607 1PF SC), and (4) rayon (Puritan, # 25-806 1
WR). Two pre-moistening solutions were evaluated; (1) PBSTX and
(2) NB. The swabs were pre-moistened by submersion in a tube with
one of the pre-moistening liquids for 10 min. The swabs were then
pressed against the inside wall of the tube, as they were removed, to ex-
press any excess pre-moistening liquid. The pre-moistened swabs were
directly inoculated with 100 μL of 105 CFU/mL suspension of the test or-
ganism, resulting in 104 CFU/swab, and then were placed in a 15 mL
conical tube and held for 1 h at room temperature. After the one hour
hold time, the swabs were placed into tubes containing 5 mL of phos-
phate buffered saline supplemented with 0.02% Tween® 80 (PBST)
(Sigma–Aldrich, St. Louis, MO). Preliminary work demonstrated that
no significant number of Y. pestis cells was lost during the transfer of
the swabs from one tube to another after the 1 h hold time. The swabs
in the 5 mL PBST were then processed by one of four methods:
(1) vortexing for 3 min (VX-2500 multi-tube vortexer set on the
highest speed, VWR, Sewanee, GA), (2) water bath sonicating for
3 min (FS 20, 40-KHz sonic cleaner; Fisher Scientific, Pittsburg, PA),
(3) vortexing and sonicating for 30 s each — repeated three times for
a total of 3 min, and (4) no extraction method (submersion in PBST
only). After the extraction, the swabs were removed and the excess liq-
uid was expressed from the swab heads by pushing the swabs against
the inside wall of the tube. The PBST containing the extracted cells
was diluted in series, and each dilution plated in triplicate on TSAII. Pos-
itive controls consisted of inoculating 5 mL of PBST (no swab) with
100 μL of the 105 CFU/mL Y. pestis suspension and processing alongside
the swabs. The positive control dilutions were plated onto ten replicate
plates. All plates were incubated for 48 h at 25 °C. Colonies were count-
ed and the percent of viable cells recovered was determined, relative to
the positive controls for each test parameter. Each swab material and
pre-moistening solution combination was tested with ten replicate
swabs and one positive control. The two swab material and processing
method combinations (highest percent recovery) were selected
(macrofoam and rayon) for the remainder of the study.
2.4. Evaluation of sample storage parameters: clean swabs
To evaluate the storage parameters of clean inoculated swabs
(no dust added), the swabs were pre-moistened with either PBSTX or
NB, inoculated with a known concentration of low-virulent Y. pestis
A1122 or virulent Y. pestis CO92 as described previously, then placed
into one of the two best transport media chosen in the survival evalua-
tions. The swabs were held at 4 °C, with 10 swab samples being
removed and processed at each of the following time points: 0, 12, 18,
 S.E. Gilbert et al. / Journal of Microbiological Methods 96 (2014) 35–41 37

24, 48 and 72 h. Swabs were processed by the optimum method for Table 1
each, as determined in processing evaluations. Percent of transport media samples that yielded Y. pestis recoveries of ≤0.5 log10 and ≤0.3
log10 change relative to T0 for storage times of 12, 24, 48 and 72 h combined (15 tubes
During this phase of the study, omission of the agar from the C&B sampled at each time point, n = 60).
transport medium resulted in the precipitation of CaCl2 from the trans-
port media solution. Although no data are available indicating CaCl2 is Storage temp Transport media

an important component in transport medium, CaCl2 is an important Amies with Amies no Cary & Neutralizing Stuart, PBSTX
nutrient for virulent Y. pestis growth (Bearden et al., 2009; Higuchi charcoal charcoal Blair (mod) buffer Toshach
& Patsula
and Smith, 1961). Because of the uncertainty as to the consequences
of omitting the CaCl2, the use of C&Bmod transport medium was Percent of tubes with ≤0.5 log10 change
discontinued when conducting the experiments using “dirty” swabs 4 °C 77 100 100 100 100 100
25 °C 55 100 100 68 0 100
(dust added). None of the traditional clinical transport media are com-
mercially available without agar, so NB was chosen to replace the Percent of tubes with ≤0.3 log10 change
C&Bmod, because it performed similarly to C&B in the transport medium 4 °C 67 71 97 77 97 92
25 °C 0 85 77 0 0 100
evaluations (based on the 4 °C data) and because it is commercially
available.
PBSTX demonstrated that 100% of samples yielded recoveries ≤0.5
2.5. Evaluation of sample storage parameters: “dirty” swabs log10 of the T0 recovery at both temperatures (4 °C and 25 °C). When
the selection criterion was narrowed to ≤0.3 log10 change relative to
To simulate swabs collected from surfaces with dust or grime pres- the T0 recovery, C&Bmod and PBSTX maintained Y. pestis viability better
ent, a slurry of characterized Arizona test dust (ATD, A-3 Medium, Pow- than Amies without Charcoal, and the study proceeded with these two
der Technology Inc., Burnsville, MN) was prepared to pre-moisten the transport media. Y. pestis was more stable in most transport media at
swabs for experiments which called for “dirty” swabs. The ATD inher- 4 °C, so subsequent evaluations were conducted at 4 °C. The survival
ently contains a consortia of environmental organisms including Bacillus of Y. pestis A1122 in the three best performing transport media at 4 °C
spp, actinomycetes, fungal spores, yeasts, Micrococcus spp., etc. as de- and 25 °C over a 96 h storage period is shown in Fig. 1.
scribed in Rose et al. (2011). One gram of ATD was added to 10 mL of
the pre-moistening solution to achieve a concentration of 100 mg 3.2. Evaluation of swab extraction methods
ATD/mL. The slurry was diluted once more, by adding 10 mL to 90 mL
of the pre-moistening solution (PBSTX or NB), creating a final concen- Table 2 shows the mean percent recovery of cells from each swab
tration of 10 mg ATD/mL in the pre-moistening solution. This diluted when pre-moistened with the designated buffer and processed by the
slurry was stored at 4 °C for up to one week before use. The slurry designated method. The highest percent recovery was from macrofoam
was re-suspended by vortexing 1 min, and then swabs were dipped swabs pre-moistened with NB and processed by vortexing alone, or by
into the slurry before direct inoculation with Y. pestis and were sonication and vortexing together (93.9% or 93.5%, respectively)
processed as described previously in Section 2.4. (Table 2, shaded cells). Since no significant difference was seen between
these two processing methods (p = 0.94), the method with the fewer
2.6. Statistical analysis steps and simpler equipment (vortexing only) was chosen for subse-
quent evaluations of macrofoam swabs.
When evaluating survival in transport media, percent recovery was Rayon swabs pre-moistened with NB, inoculated with Y. pestis A1122
calculated relative to the known inoculum CFU. When comparing swab and processed using sonication alone yielded the highest percent recov-
types, processing methods, storage temperatures, pre-moistening solu- ery (77.0%) for this swab type (Table 2, shaded). Polyester swabs yielded
tions and transport media over several time periods, the percent recover- the lowest percent recoveries of all the swab types and hence, were
ies were calculated relative to the T0 total CFU recovery. The statistical discontinued for future testing. Although the optimal condition for cot-
analysis was completed using SPSS software, version 18 (IBM, Armonk, ton swabs (pre-moistened with PBSTX and vortexed and sonicated)
NY). Each data set was tested for normality. If a normal distribution
was found, then ANOVA was performed between various combinations
5.5
of the two selected swabs, pre-moistening solutions and transport
media when grouped 12 to 24 h, 12 to 48 h, and 12 to 72 h
5.0
(α ≤ 0.05). Tukey's Highest Significant Difference test was performed
as a Post Hoc Test. If the data were not normally distributed, then non-
4.5
Log10 CFU/mL

parametric tests (Kruskal–Wallis and Mann–Whitney) were performed.

4.0
3. Results
3.5
3.1. Survival of Y. pestis in various liquid transport media
3.0
Recovery of Y. pestis in transport media held at 35 °C either declined
significantly or increased significantly (p b 0.01) over time as compared 2.5
to 4 °C or 25 °C (Fig. 1a–f of the supplemental material). The differences
between 4 °C and 25 °C were not as readily apparent, so the data for 2.0
multiple time points was pooled for statistical analysis. The percent of 0 12 24 48 72 96
transport medium samples held between 12 and 72 h that yielded Time, hr
changes in Y. pestis recoveries of ≤0.5 log10 or ≤0.3 log10 relative to o o
PBSTX, 4 C PBSTX, 25 C
the recovery at T0 at 4 °C and 25 °C is presented in Table 1. Though o
NB, 4 C o
NB, 25 C
there was some variability in inocula between transport media tests, o
C&B (mod), 4 C C&B (mod), 25 C
o

this variability is unlikely to have influenced the results, since the

log10 changes were determined relative to the recovery at T0 for each Fig. 1. Survival of Y. pestis A1122 over time in the 3 best performing liquid transport media
medium. Three transport media; Amies without charcoal, C&Bmod, and (n = 15 at each time point).
38 S.E. Gilbert et al. / Journal of Microbiological Methods 96 (2014) 35–41

Table 2 percent recoveries than all other conditions p b 0.05) whether proc-
Mean percent recovery (SD) of Y. pestis for each swab material, pre-moistening liquid, and
essed within 24, 48 or 72 h (data not shown).
extraction method (n = 10).
In contrast to macrofoam swabs, clean rayon swabs inoculated with
a
Extraction method Pre–moistening agent Swab material Y. pestis A1122 and Y. pestis CO92 showed narrower ranges of recovery
Cotton Macrofoam Polyester Rayon
(73.6–118.1% and 74.7–126.5%, respectively, Fig. 3 of supplemental ma-
b
NB 60.9 (8.6) 93.9 (13.1) 44.6 (7.1) 55.2 (12.2) terial), when all storage time points, pre-moistening solutions and
Vortex only PBSTX 69.0 (2.03) 86.5 (2.59) 55.5 (1.86) 68.7 (5.20)
p–value 0.01 0.10 <0.01 0.01
transport media were considered. When all time points within the 12
to 72 h storage periods were averaged, the conditions (pre-moistening
NB 53.7 (8.8) 89.0 (12.7) 44.6 (11.7) 77.0 (14.4)
Sonicate only PBSTX 69.4 (2.31) 84.8 (3.89) 50.6 (2.73) 64.6 (5.0)
solution/transport medium stored at 4 °C) with the highest mean per-
p–value <0.01 0.33 0.13 0.02 cent recoveries were PBSTX/PBSTX for Y. pestis A1122 strain (103.7%,
NB 53.6 (11.0) 93.5 (9.4) 53.7 (11.0) 73.0 (15.7)
SD = 17.1, n = 50) and NB/PBSTX for Y. pestis CO92 strain (110.1%,
Vortex & sonicate PBSTX 79.4 (3.08) 86.3 (3.24) 53.7 (3.59) 67.8 (4.6) SD = 21.6, n = 40) (Table 3). When we consider the data at 24 and
p–value <0.01 0.04 0.99 0.33
48 h separately, the optimum condition for Y. pestis A1122 on rayon
NB 40.4 (5.0) 68.7 (9.2) 25.1 (5.0) 28.6 (7.8)
None PBSTX 48.4 (2.50) 62.5 (3.86) 46.4 (1.43) 29.8 (3.1)
swabs (PBSTX/PBSTX) yielded significantly higher percent recoveries
p–value <0.01 0.07 <0.01 0.64 than the PBSTX/C&Bmod and NB/C&Bmod conditions when processed
a
Extractions were performed one hour after swab inoculation. All extractions were
within 12 to 24 h (p b 0.03) and 12 to 48 h (p b 0.004). The optimum
performed in PBS + 0.02% Tween® 80. condition for Y. pestis CO92 on rayon swabs (NB/PBSTX) only yielded
b
Shading indicates the best extraction methods for the two best swab materials. significantly higher recovery than the PBSTX/PBSTX and NB/C&Bmod
conditions (p = 0.01) when processed after a storage time of 48 h
(data not shown).

yielded a statistically equivalent percent recovery to the optimum con-
dition for rayon swabs (79% and 77%, respectively), cotton swabs were
excluded from subsequent evaluations because of anecdotal concerns
for potential PCR inhibition. Subsequent phases of the project focused
on the use of macrofoam swabs processed by vortexing (3 min) and
rayon swabs processed by sonication (3 min).
3.3. Evaluation of sample storage parameters: clean swabs
When all storage time points, pre-moistening solutions and trans-
port media were considered, the recovery ranges from clean macrofoam
swabs inoculated with Y. pestis showed high variation, especially for the
virulent strain (46.9–106.2% for Y. pestis A1122 and 2.3–114.3% for
Y. pestis CO92, Fig. 2 of supplemental material). The conditions (pre-
moistening solution/transport medium stored at 4 °C) with the highest
mean percent recovery for clean macrofoam swabs when data from
all storage time points (12 to 72 h) were averaged occurred when NB/
C&Bmod (99.6%, SD = 10.9, n = 49) and PBSTX/C&Bmod (101.8%,
SD = 18.0, n = 49) were used for Y. pestis A1122 and Y. pestis CO92
strains, respectively (Table 3). When we consider the data at 24 h and
48 h separately, the optimum conditions for Y. pestis A1122 on
macrofoam swabs (NB/C&B mod) yielded significantly higher recoveries
than one other condition at 24 h (NB/PBSTX) and significantly higher
recovery than two others at 48 h (PBSTX/C&Bmod and NB/PBSTX,
p b 0.04, data not shown). The optimum condition for clean macrofoam
swabs inoculated with Y. pestis CO92 (PBSTX/C&Bmod) yielded higher
Table 3
Optimum conditions for recovery of Y. pestis A1122 (low-virulence) and Y. pestis CO92
(virulent) from each swab type.
Swab type Strain Clean or Pre-moistening Transport
dirty agent medium
3.4. Evaluation of sample storage parameters: “dirty swabs”
When all storage time points, pre-moistening agents and transport
media were considered, dirty macrofoam swabs yielded similar ranges
of recoveries for both Y. pestis strains (65.8–111.6% and 58.0–100% for
Y. pestis A1122 and Y. pestis CO92, respectively; Fig. 2). When all time
points within the 12 to 72 h storage periods were averaged, the condi-
tions with the highest mean percent recoveries were PBSTX/NB for dirty
swabs inoculated with Y. pestis A1122 (105.1%, SD = 11.7, n = 40) and
NB/PBSTX for dirty swabs inoculated with Y. pestis CO92 (85.9%,
SD = 10.6, n = 40) (Table 3). The optimum condition of PBSTX/NB
for dirty swabs inoculated with Y. pestis A1122 yielded significantly
higher percent recoveries than the conditions PBSTX/PBSTX and NB/
NB (p b 0.001), but only when stored for up to 24 h. In comparison,
the optimum condition of NB/PBSTX for dirty swabs inoculated with
Y. pestis CO92 yielded significantly higher percent recovery than the
condition NB/NB, whether processed within 24, 48, or 72 h (p b 0.022).
Dirty rayon swabs inoculated with Y. pestis A1122 and ATD yielded a
narrower range of recovery (79.1–112.8%; Fig. 3A) than the dirty rayon
swabs inoculated with Y. pestis CO92 and ATD (34.3–100%; Fig. 3B).
When all time points within the 12 to 72 h storage periods were aver-
aged, the highest mean percent recovery for dirty rayon swabs inoculat-
ed with Y. pestis A1122 was 107.2% (SD = 14.3, n = 40) for the
optimum condition (pre-moistening solution/transport medium stored
at 4 °C) PBSTX/PBSTX. The highest mean percent recovery for dirty
rayon swabs inoculated with Y. pestis CO92 and ATD was 83.7%
(SD = 14.3, n = 40) for the optimum condition NB/PBSTX. The opti-
mum condition for the Y. pestis A1122 strain (PBSTX/PBSTX) yielded sig-
nificantly higher percent recoveries than all other conditions for storage
times of 24 h (p b 0.02) or 48 h (p b 0.001).The percent recovery for
the optimum condition for Y. pestis CO92 stain (NB/PBSTX) was only sig-
nificantly higher than one condition (PBSTX/NB, p b 0.001) for both 24
Mean percent and 48 h storage times.
recovery

24 ha 48 hb 72 hc 3.5. Statistical analysis

Macrofoam A1122 Clean NB C&Bmod 101.9 99.2 99.6
Dirty PBSTX NB 106.7 106.1 105.1 When recovery of Y. pestis was compared from all conditions tested
CO92 Clean PBSTX C&Bmod 110.6 106.1 101.8 (clean or dirty, pre-moistening solution and transport medium), no sin-
Dirty NB PBSTX 84.2 85.3 85.9 gle condition was identified as the best for recovery from both
Rayon A1122 Clean PBSTX PBSTX 101.0 100.1 103.7
macrofoam and rayon swabs or for both strains (Table 3), indicating
Dirty PBSTX PBSTX 105.3 107.8 107.2
CO92 Clean NB PBSTX 120.3 115.9 110.1 that the Y. pestis strain should be considered, if possible, when
Dirty NB PBSTX 82.5 84.3 83.7 conducting sampling during a response or conducting research of this
a
24 h represents the mean percent recovery for 12 and 24 h storage combined.
type. Another approach to analyzing the data was considered by exam-
b
48 h represents the mean percent recovery for 12, 24, and 48 h storage combined. ining the optimum conditions for the worst case scenario of an investi-
c
72 h represents the mean percent recovery for 12, 24, 48, and 72 h storage combined. gation (when processing and analysis do not occur until 72 h after
 S.E. Gilbert et al. / Journal of Microbiological Methods 96 (2014) 35–41
Fig. 2. Mean percent recovery of two Y. pestis strains from dirty macrofoam swabs stored at
4 °C; A) Y. pestis A1122 and B) Y, pestis CO92 (104 CFU/swab). Bars represent the 95%
confidence interval of the mean recovery (n = 10).
sampling, and the strain is not known). The data for each swab type
were combined, including the data for both Y. pestis strains, both clean
and dirty swabs, and all time points up to 72 h. For these analyses, the
data for the substituted transport media (NB) was pooled with the orig-
inally tested C&Bmod data set (PBSTX/C&Bmod data pooled with PBSTX/
NB and NB/C&B pooled with NB/NB). The combined data for each
swab type were not normally distributed, and therefore were analyzed
using the Kruskal–Wallis and Mann–Whitney tests. The results of these
analyses for macrofoam swabs showed one condition that resulted in
statistically higher recoveries than the others; the PBSTX/C&Bmod
pooled with PBSTX/NB (p b 0.001), data not shown). Since the standard
C&B formulation was found to be unstable in liquid form and was not
commercially available, PBSTX/NB was chosen as the optimum pre-
moistening solution/transport medium for macrofoam swabs (in the
event that the Y. pestis strain is not known). When the data was grouped
similarly for rayon swabs (both Y. pestis strains, dirty and clean swabs,
and all time points), the test results showed that two pre-moistening
solution/transport medium combinations stood out with statistically
39
Fig. 3. Mean percent recovery of two Y. pestis strains from dirty rayon swabs stored at 4 °C
A) Y. pestis A1122 and B) Y. pestis CO92 (104 CFU/swab). Bars represent the 95% confidence
interval of the mean recovery (n = 10).
higher recoveries than the others; PBSTX/PBSTX (p b 0.001, data not
shown) and NB/PBSTX (p b 0.001, data not shown).
4. Discussion
Of the four swabs evaluated, rayon and macrofoam were chosen as
the two best materials, because they released Y. pestis cells significantly
better than polyester, and they did not have the PCR reagent inhibition
concerns associated with cotton. Though no conclusive study has been
conducted to confirm that cotton inhibits PCR, cotton fibers are known
to absorb heavy metals and humic acids in the environment, both of
which are known PCR inhibitors (Deng and Bai, 2003). Recovery from
rayon swabs was higher when processed by sonication and recovery
from macrofoam swabs was higher when processed by vortexing. Dur-
ing the course of the study, recoveries of N100% were sometimes ob-
served, though this is likely due to occasional underestimation of the
40 S.E. Gilbert et al. / Journal of Microbiological Methods 96 (2014) 35–41
inoculating suspension that included clusters cells that were subse-
quently dispersed during processing. This phenomenon has been ob-
served by others conducting similar research (Downey et al., 2012;
Rose et al., 2004; Moore and Griffith, 2007).
Six liquid transport media were evaluated for their ability to support
Y. pestis A1122 cells at 4, 25, and 35 °C over several holding times up to
72 h. The most consistent recovery of Y. pestis in all transport media was
found when the storage temperature was maintained at 4 °C. These
findings are similar to the findings of Hubbard et al. (2011) in which
Y. pestis was found to be best preserved on rayon swabs at 4 °C when
held in liquid Stuart medium and processed by vortexing 2 min
(Hubbard et al., 2011). The transport system that Hubbard et al.
(Hubbard et al., 2011) used consisted of a swab in contact with a sponge
saturated with the transport medium, rather than the swab submerged
in the transport medium, as evaluated in the current work. This differ-
ence in transport systems may explain the contrasting results, as
Hubbard et al. found that PBST performed worse than the other trans-
port media.
For the current experiments, quantitation of Y. pestis recovered from
swabs was required, so liquid transport media were used rather than
semi-solid transport media. This was done because a significant number
of Y. pestis may remain in the semi-solid medium when swabs are re-
moved and processed in a separate tube of extraction fluid. Although
C&Bmod and PBSTX were found to be the two best transport media of
the six evaluated, C&Bmod may not be the ideal transport medium
choice. Modification of the traditionally semi-solid C&B medium into a
liquid medium required the omission of calcium chloride from the for-
mulation, an ingredient that influences the virulence factors of Y. pestis
in growth medium (Higuchi and Smith, 1961; Bearden et al., 2009).
Additionally, the modified C&B is not commercially available and must
be formulated in a laboratory. Following a contamination event involv-
ing Y. pestis, quick access to transport media and other sampling mate-
rials would be needed. The next best commercially available transport
medium (NB) was substituted for all subsequent work using the
“dirty” swabs.
Results of this study indicated that no single pre-moistening agent/
transport medium combination stood out as the best for all swab condi-
tions tested (swab types, clean and dirty, Y. pestis strains). The strain
variability noted in the present work emphasizes the need for evaluat-
ing multiple strains rather than generalizing results based solely on
one strain. For this study, when rayon swabs were analyzed, the condi-
tion PBSTX/PBSTX gave the highest percent recovery for the Y. pestis
A1122 strain. The condition that gave the highest percent recovery for
Y. pestis CO92 was, NB/PBSTX regardless of whether the swabs were
clean or dirty. When “dirty” macrofoam swabs were analyzed, PBSTX/
NB and NB/PBSTX were the conditions that gave the highest percent re-
coveries for Y. pestis A1122 strain and for Y. pestis CO92, respectively.
Because the strain may not be known at the time of sampling, the choice
of pre-moistening solutions may be decided upon by availability or may
be based on the desire to reduce the residual disinfectant carryover.
The data for both strains, for both dirty and clean swabs, and for all
time points were combined for each swab type. The results of the anal-
ysis of macrofoam swab data indicated that macrofoam swabs pre-
moistened with PBSTX and stored in NB (PBSTX/NB) provided signifi-
cantly higher percent recovery of Y. pestis than any other combinations
of pre-moistening agent and transport medium tested (after C&Bmod
formulation was omitted from consideration for reasons noted above).
Analysis of the results for the combined rayon swabs data indicated
that two combinations of pre-moistening agent/transport medium,
PBSTX/PBSTX and NB/PBSTX, stood out as significantly better for
recovery of Y. pestis than the other two.
It should be noted that in evaluating sample storage parameters, the
swabs were stored and processed in the given transport medium, so the
differences in preferred transport medium for each swab type may
reflect the differences in the surface chemistry of each material
(macrofoam and rayon) and the need for the surfactant in PBSTX to
enhance release of the Y. pestis during processing. Each swab material
has unique electrochemical properties that interact with the cells and
influence adherence (Da Silva et al., 2011).
5. Conclusions
The findings of this work provide first responders and the research
community with valuable input on the optimum methods for recovery
of viable Y. pestis from swabs after storage and/or shipping. These data
suggests that 4 °C is the optimum storage temperature, and that if the
optimum combinations of swab, pre-moistening agent and transport
medium are used, swabs may be held for 72 h without significant loss
of cell viability. The optimum combinations of pre-moistening agent
and transport media differ for macrofoam and rayon swabs, and for
each of the two strains tested, but holding samples at 4 °C for up to
72 h with any combination of NB and PBSTX may allow for a mean of
≥83% recovery of cells from swabs. Since one study showed that Y.
pestis remained viable on surfaces for 5 days or less (Rose et al., 2003),
detection may only be possible using molecular assays. Additional stud-
ies should be conducted to confirm that these pre-moistening solutions
and transport media are compatible with detection assays such as PCR
and immunoassays.
Disclaimer and Acknowledgements
The U.S. Environmental Protection Agency (USEPA), through its
Office of Research and Development, collaborated with the Centers
for Disease Control and Prevention in the research described herein
under EPA IA# DW75-92259701. We thank Eugene Rice (EPA/
NHSRC) for his valuable input and assistance in planning this
study. This content has been peer and administratively reviewed
and has been approved for publication as a joint USEPA and CDC doc-
ument. Note that approval does not signify that the contents neces-
sarily reflect the views of the USEPA, the CDC, the Public Health
Service, or the U.S. Department of Health and Human Services. Refer-
ence herein to any specific commercial product, process, or service
by trade name, trademark, manufacturer, or otherwise does not nec-
essarily constitute or imply its endorsement, recommendation, or fa-
voring by the United States government. The views and opinions
expressed herein do not necessarily state or reflect those of the
United States government and shall not be used for advertising or
product endorsement purposes.
Appendix A. Supplementary data
Supplementary data to this article can be found online at http://dx.
doi.org/10.1016/j.mimet.2013.10.017.
References
Atlas, R.M., 2010. Handbook of Microbiological Media. CRC Press, Taylor and Francis
Group, Boca Raton, FL.
Ayyadurai, S., Houmadi, L., Lepidi, H., Nappez, C., Raoult, D., Drancourt, M., 2008. Long-
term persistence of virulent Yersinia pestis in soil. Microbiology 154, 2865–2871.
Bearden, S.W., Sexton, C., Pare, J., Fowler, J.M., Arvidson, C.G., Yerman, L., Viola, R.E.,
Brubaker, R.R., 2009. Attenuated enzootic (pestoides) isolates of Yersinia pestis
express active aspartase. Microbiology 155, 198–209.
Buttner, M., Cruz, P.P., Stetzenbach, L.D., Cronin, T., 2007. Evaluation of two surface sam-
pling methods for detection of Erwinia herbicola on a variety of materials by culture
and quantitative PCR. Appl. Environ. Microbiol. 73, 3505–3510.
Canter, D.A., 2005. Addressing residual risk issues at anthrax cleanups: how clean is safe?
J. Toxicol. Environ. Health A 68, 1017–1032.
Da Silva, S.M., Filliben, J.J., Morrow, J.B., 2011. Parameters affecting spore recovery from
wipes used in biological surface sampling. Appl. Environ. Microbiol. 77, 2374–2380.
Deng, S., Bai, R.B., 2003. Aminated polyacrylonitrile fibers for humic acid adsorption:
behaviors and mechanisms. Environ. Sci. Technol. 37, 5799–5805.
Downey, A.S., DaSilva, S.M., Olson, N.D., Filliben, J.J., Morrow, J.B., 2012. Impact of process-
ing method on recovery of bacteria from wipes used in biological surface sampling.
Appl. Environ. Microbiol. 7, 5872–5880.
 S.E. Gilbert et al. / Journal of Microbiological Methods 96 (2014) 35–41
Edmonds, J.M., Collett, P.J., Valdes, E.R., Skowronski, E.W., Pellar, G.J., Emanuel, P.A., 2009.
Surface sampling of spores in dry-deposition aerosols. Appl. Environ. Microbiol. 75,
39–44.
Gilbert, S.E., Rose, L.J., 2012. Survival and persistence of nonspore-forming biothreat
agents in water. Lett. Appl. Microbiol. 55, 189–194.
Higuchi, K., Smith, J.L., 1961. Studies on the nutrition and physiology of Pasteurella pestis.
VI. A differential plating medium for the estimation of the mutation rate to
avirulence. J. Bacteriol. 81, 605–608.
Hodges, L.R., Rose, L.J., Peterson, A., Noble-Wang, J., Arduino, M.J., 2006. Evaluation of a
macrofoam swab protocol for the recovery of Bacillus anthracis spores from a steel
surface. Appl. Environ. Microbiol. 72, 4429–4430.
Hubbard, K., Pellar, G., Emanuel, P., 2011. Suitability of commercial transport media for
biological pathogens under nonideal conditions. Int. J. Microbiol. 2011, 9 (Article ID
463096).
Moore, G., Griffith, C., 2007. Problems associated with traditional hygiene swabbing: the
need for in-house standardization. J. Appl. Microbiol. 103, 1090–1103.
41
Rose, L.J., Donlan, R., Banerjee, S.N., Arduino, M.J., 2003. Survival of Yersinia pestis on
environmental surfaces. Appl. Environ. Microbiol. 69, 2166–2171.
Rose, L., Jensen, B., Peterson, A., Banerjee, S.H., Arduino, M.J., 2004. Swab materials and Bacil-
lus anthracis spore recovery from nonporous surfaces. Emerg. Infect. Dis. 10, 1023–1029.
Rose, L.J., 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.
Teshale, E.H., Painter, J., Burr, G.A., Mead, P., Wright, S.V., Cseh, L.F., Zabrocki, R., Collins, R.,
Kelley, K.A., Hadler, J.L., Swerdlow, D.L., 2002. Environmental sampling for spores of
Bacillus anthracis. Emerg. Infect. Dis. 8, 1083–1087.
Torosian, S.D., Regan, P.M., Taylor, M.A., 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.
Government Accountability Office, U.S., 2005. Anthrax detection: agencies need to vali-
date sampling activities in order to increase confidence in negative results. GAO-05-
251, Washington, DC.