Upload
buiquynh
View
216
Download
2
Embed Size (px)
Citation preview
1
Evaluation of Sample Recovery Efficiency of Bacteriophage P22 on Fomites 1
2
Amanda B. Herzog 1, 4, Alok K. Pandey 1, David Reyes-Gastelum 5, Charles P. Gerba 4,6, Joan B. 3
Rose 3,4, and Syed A. Hashsham 1,2• 4
5
1 Department of Civil and Environmental Engineering, 2Center for Microbial Ecology, 6
3Department of Fisheries and Wildlife, 4Center for Advancing Microbial Risk Assessment, 7
5Center for Statistical Training and Consulting, Michigan State University, East Lansing MI; 8
6Department of Soil, Water and Environmental Science, University of Arizona, Tucson AZ 9
10
11
• Corresponding author: 12
Syed A. Hashsham 13
Department of Civil and Environmental Engineering 14
A126 Research Complex-Engineering 15
East Lansing, MI 48824 16
Phone: (517) 355 8241 17
Fax: (517) 355 0250 18
Email: [email protected] 19
20
21
22
23
Copyright © 2012, American Society for Microbiology. All Rights Reserved.Appl. Environ. Microbiol. doi:10.1128/AEM.01370-12 AEM Accepts, published online ahead of print on 31 August 2012
on June 20, 2018 by guesthttp://aem
.asm.org/
Dow
nloaded from
2
ABSTRACT 24
Fomites are known to play a role in the transmission of pathogens. Quantitative analysis of the 25
parameters that affect sample recovery efficiency (SRE) at the limit of detection of viruses on 26
fomites will aid in improving quantitative microbial risk assessment (QMRA) and infection 27
control. The variability in SRE as a function of fomite type, fomite surface area, sampling time, 28
application media, relative humidity (RH), and wetting agent was evaluated. To quantify the 29
SRE, bacteriophage P22 was applied onto the fomites at average surface densities of 0.4 ± 0.2 30
and 4 ± 2 PFU/cm2. Surface areas 100 and 1000 cm2 of nonporous fomites found in indoor 31
environments (acrylic, galvanized steel, and laminate) were evaluated with pre-moistened 32
antistatic wipes. The parameters with the most effect on the SRE were sampling time, fomite 33
surface area, wetting agent, and RH. At time zero (initial application of bacteriophage P22), the 34
SRE for 1000 cm2 was on average 40% lower than 100 cm2. For both fomite surface areas, 35
application media trypticase soy broth (TSB) and/or the laminate fomite predominantly resulted 36
in a higher SRE. After the applied samples dried on the fomites (20 min), the average SRE was 37
less than 3%. A TSB wetting agent applied on the fomite, improved the SRE for all samples at 38
20 min. In addition, RH greater than 28% generally resulted in a higher SRE than RH less than 39
28%. Parameters impacting SRE at the limit of detection have the potential to enhance sampling 40
strategies and data collection for QMRA models. 41
42
INTRODUCTION 43
Nonporous fomites (inanimate or nonliving objects) can be an important vehicle in the 44
transmission of viral disease especially for populated indoor environments such as schools, 45
daycare centers, nursing homes, hospitals, food preparation settings or any civil infrastructure (4-46
on June 20, 2018 by guesthttp://aem
.asm.org/
Dow
nloaded from
3
6, 26, 33). Human exposure can be through touching and transfer of pathogens present on the 47
fomite to the hands and then to the mouth, nasopharynx, and eyes (5, 25). Exposure can also be 48
from the inhalation of re-aerosolized organisms from contaminated fomites (5, 26). A 49
challenging task can be to control and remediate an indoor environment from an outbreak 50
resulting from an accidental or intentional release of viruses (1, 26). 51
52
After decontaminating an indoor environment, to declare a site as “clean”, quantification of the 53
loss due to sample recovery which is specific to the method(s) used is essential to verify the 54
efficacy of decontamination (17). Quantitative analyses of the parameters that affect SRE from 55
fomites are vital for implementing efficient sampling and detection methods (17). The use of 56
infection transmission models that include the environmental dynamics (environmental 57
conditions, human behavior, survival characteristics of the agent in the environment, etc.) can be 58
used to make decisions on interventions to prevent viral outbreaks (20). Without a quantitative 59
assessment of the abundance of such agents in the environment, generic intervention 60
recommendations could be ineffective (4, 36). 61
62
Survival and SRE studies with viruses have generally been conducted on fomites at surface 63
densities of 102 PFU/cm2 or higher by applying virus stocks in volumes ranging from 5 to 500-µl 64
on fomite areas ranging from 0.38 to 32 cm2 (Table 1). Use of higher initial titers is known to 65
extend the viral survival rate on fomites (5). Under these optimal conditions, results may 66
represent the upper limits of SRE. Surface densities may also be lower than what has been 67
studied so far and pose significant risk (Table 1). However, parameters affecting survival at very 68
low surface densities are less well studied. To our knowledge, only two survival studies (3, 7) 69
on June 20, 2018 by guesthttp://aem
.asm.org/
Dow
nloaded from
4
and two SRE studies (18, 37) have been conducted at surface densities ranging from 0.02 – 50 70
PFU or TCID50/cm2 (Table 1 and Table S1 in the supplemental material). These factors may 71
have a significant effect on quantifying the risk to human health after decontamination. 72
73
The objective of this study was to evaluate the parameters that affect SRE of bacteriophage P22, 74
a surrogate for DNA viruses (21, 31), at concentrations close to the limit of detection. 75
Bacteriophage P22 was chosen because it is a surrogate for DNA viruses such as Adenovirus 76
(13, 31), meets many of the desired characteristics of a surrogate (21, 31, 34), and has been 77
successfully used by our group in environmental release and recovery studies (21, 31). The 78
variability of SRE from the parameters such as fomite type, fomite surface area, sampling time, 79
application media, wetting agent, and RH was evaluated. Results presented here have 80
implications for sampling strategies and subsequent microbial risk assessment at low 81
concentrations. 82
83
MATERIALS AND METHODS 84
Bacteriophage P22: preparation, application and sample recovery 85
Bacteriophage P22 that infects the bacterial host Salmonella typhimurium LT2 (ATCC 19585), 86
was provided by Charles P. Gerba (University of Arizona). Bacteriophage P22 is a dsDNA 87
icosahedral-shaped virus with a short tail, 52 to 60 nm in size, and belongs to the family 88
podoviridae (31). To prepare bacteriophage P22, 1-ml of bacteriophage P22 stock was added to 89
25-ml of the bacterial host, S. typhimurium, at log phase in TSB (Difco, Becton Dickinson and 90
Company, Sparks, MD). After 24 hr incubation at 37 oC, 0.1-ml of lysozyme and 0.75-ml of 91
ethylenediamine tetraacetic acid (EDTA) were added to the solution and centrifuged at 2390 x g 92
on June 20, 2018 by guesthttp://aem
.asm.org/
Dow
nloaded from
5
for 10 minutes. The supernatant was filtered through a 0.45 µm filter (Millipore) to remove the 93
bacterial cells and debris (31). Bacteriophage P22 was then diluted in suspensions of phosphate 94
buffered saline tween-80 (Fisher Scientific, NJ) (PBST), TSB or sterile distilled water. 95
96
Fomites, simulating an indoor environment, included acrylic (Optix, Plaskolite Inc, Columbus, 97
OH), galvanized steel (Type 28 GA galvanized, MD Building Products, Oklahoma City, OK) 98
and laminate (Type 350 #60 mate finish, Wilsonart International Inc, Temple, Texas) with 99
surface areas of 100 and 1000 cm2. Fomites and testing area were disinfected with 70% ethanol, 100
rinsed with sterile distilled water and dried. Bacteriophage P22 was applied in PBST, TSB or 101
water on the fomite in a grid formation comprising of fifty 1-µl droplets. The average amount of 102
bacteriophage P22 applied to the fomite was 433.1 ± 194.5 PFU, approximately 8.66 103
PFU/droplet, with average surface densities of 4.3 ± 1.9 PFU/cm2 for 100 cm2 and 0.4 ± 0.2 104
PFU/cm2 for 1000 cm2. The recovery material, pre-moistened Fellowes Screen Cleaning Wipes 105
are generally used to remove dirt, dust and finger prints from office equipment (#99703, 106
Fellowes, Itasca, IL) and are antistatic, non-toxic and alcohol free. The pre-moistened wipes are 107
made of crepe fabric (crepe material is treated as a trade secret by Fellowes) and wetted by the 108
manufacture with water and detergent (propylene glycol ethers). Pre-moistened wipes were cut 109
into 48 cm2 pieces using sterilized scissors and stored in sterile whirl pack bags at room 110
temperature during the experiment lasting no more than 12 hr. Fresh pieces were cut and used 111
each day. Sampling was done by moving the pre-moistened wipes over the entire fomite twice 112
(in perpendicular directions to each other). Two samples were taken; one immediately after the 113
initial application (referred to as 0 min) and another after the samples were visibly dry (which 114
was 20 min). Control experiments conducted with bacteriophage P22 suspensions to determine 115
on June 20, 2018 by guesthttp://aem
.asm.org/
Dow
nloaded from
6
if the moistening agent had an effect on the viability of the virus indicated that on average 116
95% (range 80 – 125%) of bacteriophage P22 could be recovered with inoculation directly onto 117
the wipe and dissolution with PBST. Very high recoveries were also seen at time zero on 118
fomites with no drying. 119
120
After sampling, the recovery material was placed into a 50-ml tube containing 5-ml of PBST and 121
vortexed for 30 seconds. Bacteriophage P22 was assayed using a double agar layer method (14). 122
The sample containing bacteriophage P22 (1-ml) was added to 2.5-ml of melted 1% agar overlay 123
(1 g bacto agar/100 ml TSB) (Bacto agar: Difco, Becton Dickinson and Company, Sparks, MD) 124
with 0.3-ml of S. typhimurium in the log phase. The solution was rolled by hand for mixing and 125
immediately dispensed evenly onto 1.5% trypticase soy agar (TSA) (Difco, Becton Dickinson 126
and Company, Sparks, MD) plates. After the overlay agar solidified, the plates were incubated 127
at 37 oC for 24 hr; the number of plaque forming units was then counted. The SRE experiments 128
conducted used a total of 324 plates. These consisted of 3 fomite types, 2 sampling times, 3 129
application media, and 2 fomite surface areas. Each SRE measurement was made in triplicate 130
and repeated on three different days. Because the PFUs recovered were already very low, 131
dilution of samples was not necessary. For each sample recovery experiment, positive controls 132
were conducted in triplicate. Fifty 1-µl droplets of bacteriophage P22 inoculated in 950-μl of 133
PBST (same as extraction solution) were dispensed into a 1.5-ml microcentrifuge tube. The 1-ml 134
bacteriophage P22 control was dispensed as described above. 135
136
Single agar layer method to separate bacteriophage P22 survival from sample recovery 137
When evaluating the survival of bacteriophage P22 on fomites, fifty 5-μl droplets containing an 138
on June 20, 2018 by guesthttp://aem
.asm.org/
Dow
nloaded from
7
estimated 3.96 PFU/droplet suspended either in TSB or water were applied on polystyrene Petri 139
dish surface (100 x 15 mm) in a grid formation. An average of 198 ± 65 PFU were applied to 140
each plate with an average surface density of 2.5 ± 0.9 PFU/cm2. For this experiment, the time 141
of first sampling (other than the initial at time zero) was changed to 1 hr instead of 20 min 142
because the 5-μl droplets took longer to visibly dry on the Petri dish. Samples were evaluated at 143
0, 1, 2, 4, 8, 12, and 24 hr by implementing a single agar layer method. This method allowed us 144
to evaluate the PFU remaining but eliminated the need to recover them from a surface because 145
bacteriophage P22 was directly applied on the Petri dish surface. The assay consisted of 146
dispensing 3-ml of melted 1% agar overlay (1 g bacto agar/100 ml TSB) with 0.5-ml of S. 147
typhimurium in the log phase and 2-ml of TSB onto the Petri dish surface where bacteriophage 148
P22 were applied. After the overlay agar solidified, the plates were incubated at 37 oC for 24 hr 149
at which point the number of plaque forming units was then counted. The experiment was 150
conducted twice using six replicates per time point spanning 7 sampling time points and two 151
application media (thus using a total of 168 plates). For each survival experiment a positive 152
control was also included in triplicate which consisted of fifty 5-μl droplets of bacteriophage P22 153
inoculated in 750-μl of PBST in 1.5-ml microcentrifuge tubes. One ml of this positive control 154
was plated as described above. 155
156
Relative humidity and TSB wetting agent 157
The RH and temperature in the laboratory were measured before each experiment with a digital 158
RH and temperature meter (VWR Scientific Products). The average temperature of the 159
laboratory during these experiments was 20.8 ± 0.23 (mean ± standard deviation). The RH 160
ranges of 9 – 23% and 28 – 32% were the natural RHs of the laboratory during the winter and 161
on June 20, 2018 by guesthttp://aem
.asm.org/
Dow
nloaded from
8
summer months respectively (Figure 3). For RH range of 55 – 58%, a small laboratory space (14 162
ft x 7 ft x 9 ft) was equipped with a humidifier (Bionaire, Milford, MA). 163
164
Previous studies support that use of a wetting agent applied to the recovery material (wipe or 165
swab) to enhance the SRE (9, 18, 19, 29). In a preliminary experiment, PBST and TSB were 166
compared as wetting agents applied on the fomite surface to evaluate its effect on SRE 167
enhancement at 20 min. There was no statistical differences between the SREs when PBST or 168
TSB as wetting agent was applied on the fomite (p=0.232, n=27, student t-test, data not shown). 169
Hence in further SRE experiments, a TSB wetting agent was used (this step is referred to as TSB 170
wetting). Using a disposable spreader, 200 and 400 µl of TSB was applied and uniformly 171
distributed over 100 and 1000 cm2 fomite surface area, respectively. The recovery material 172
sampled both the disposable spreader and the fomite. The recovery materials were processed as 173
described above. This experiment used a total of 162 plates consisting of 2 wetting conditions, 1 174
fomite surface area, 1 sampling time, 3 fomites, 3 application mediums, and 3 RH. Each 175
measurement was made in triplicate. A positive control was also included in triplicate as 176
described previously for the SRE experiments. 177
178
Percent surface recovery efficiency computations and statistical analysis 179
Percent sample recovery efficiency was calculated as: 180
181
( )100% ×=
control
assay
N
DNSRE (1) 182
183
on June 20, 2018 by guesthttp://aem
.asm.org/
Dow
nloaded from
9
where %SRE was the sample recovery efficiency from the fomite, Nassay was the number of 184
plaque forming units counted on the agar plate from sampling the fomite, D was the dilution 185
factor (the total extraction volume divided by the volume of sample assayed), and Ncontrol was the 186
number of units on the agar plate from the control experiment. 187
188
The data (%SRE) had considerable differences in variance, especially between 0 min and 20 189
min. Due to this, the data were transformed by adding 1 (to account for the zero values) and 190
converted to a log scale. After analyzing the residuals, it was determined that the normality 191
assumption of the residual did not fit the equation, therefore the residuals were fitted under the 192
assumption of a gamma distribution. Two equations for the transformed outcome were used to 193
study the relationships between fomite type, application media, RH, and wetting condition. 194
195
Log (%SRE +1) = βo + β1X1 + β2X2 + β3X1X2 + e (2) 196
197
Where for equation 2, Log (%SRE +1) is the log transformed SRE, X1 is an independent variable 198
that denotes the fomite type so X1 is a nominal variable with no numerical value (acrylic, 199
laminate, and galvanized steel as categories) for which laminate was taken as the reference 200
category in the analyses, X2 is an independent variable that denotes the application media so X2 201
is a nominal variable (PBST, TSB, and water as categories) for which water was taken as the 202
reference category in the analyses, X1X2 is the interaction between fomite type and application 203
media, and e is the error term. The intercept βo represents the average value of the reference 204
group, in this case, the average value of the log of the reference categories laminate fomite and 205
on June 20, 2018 by guesthttp://aem
.asm.org/
Dow
nloaded from
10
water media. The terms β1, β2, and β3 are the regression coefficients known as the effect for the 206
corresponding independent variable X1, X2, and X1X2, respectively. 207
208
Log (%SRE +1) = βo + β1X1 + β2X2 + β3X3 + β4X4 + β5X1X2 + β6X2X3 + β7X1X3 + e (3) 209
210
In equation 3, Log(%SRE +1) is the log transformed SRE, X1 and X2 are defined as in equation 211
(2), X3 is an independent variable that denotes RH range so X3 is a nominal variable (9 – 23%, 212
28 – 32%, and 55 – 58% as categories) for which 55 – 58% was taken as the reference category 213
in the analyses, X4 is an independent variable that denotes the use of TSB wetting for the sample 214
collection so X4 is a nominal variable (no wetting and TSB wetting as categories) for which TSB 215
wetting was taken as the reference category in the analyses, and e is the error term. As before, 216
the intercept βo represents the average value of the log of the reference categories laminate 217
fomite, water media, 55 – 58% RH and TSB wetting. The interaction terms are X1X2 (fomite 218
type and application media), X2X3 (application media and RH), and X1X3 (fomite type and RH). 219
The regression coefficients (β1-7) are known as the effect for the corresponding independent 220
variable X1-4, X1X2, X2X3, and X1X3, respectively. 221
222
Because the independent variables used were nominal, dummy variables were used to compare 223
the different categories to the corresponding reference categories. The dummy variable described 224
the set of experimental conditions consisting of fomite type, application media, RH, and wetting 225
condition as single entity and evaluate the SRE for each set to the next by treating two such sets 226
as reference (laminate and water). A regression was run using SAS 9.2® with the GLIMMIX 227
procedure to evaluate the equations. Data was analyzed to evaluate the type III test of fixed 228
on June 20, 2018 by guesthttp://aem
.asm.org/
Dow
nloaded from
11
effects (emanating from the factors being investigated) to determine the significance of each of 229
the parameters specified in the model statement (27). Analysis of the model was performed in 230
groups wetting condition, fomite surface area and sampling time. Patterns in the experimental 231
data indicated differences to explore certain effects. This limited error rates and avoided 232
canceling of significant effects. 233
234
RESULTS 235
SRE of bacteriophage P22 from various fomites 236
For 100 cm2 fomite surface area and the three application media (PBST, TSB, or water), the 237
average SRE for the experimental data at 0 min was 46 ± 6.9% (SRE ± standard deviation) for 238
acrylic, 70 ± 7.7% for galvanized steel, and 92 ± 6.4% for laminate (Figure 1A). The type III 239
test of fixed effects (equation (2)) for 100 cm2 at 0 min was significant for fomite type 240
(p<0.0001), application media (p<0.0001) and the interaction between fomite type and 241
application media (p=0.0128) (Table S2). Based on equation (2), laminate yielded the highest 242
SRE while acrylic gave the lowest SRE regardless of which application media was used. 243
However, use of TSB did result in a higher SRE than the other media. PBST and TSB 244
performed similarly on acrylic while PBST and water performed similarly on laminate (Table 245
S3). At 20 min, the average SRE for acrylic, galvanized steel and laminate were all less than 1 ± 246
0.9% for all application media (Figure 1A). The type III test of fixed effects for 100 cm2 at 20 247
min was significant for fomite type (p=0.0047) and application media (p<0.0001) but not 248
significant (p=0.3589) for the interaction between fomite type and application media (Table S4). 249
For these conditions, the application media significantly affected the SRE and a higher SRE was 250
observed from application media TSB. Similar to 0 min, laminate resulted with a higher SRE 251
on June 20, 2018 by guesthttp://aem
.asm.org/
Dow
nloaded from
12
while acrylic resulted with a lower SRE. Similar results were observed on acrylic and 252
galvanized steel when applied in PBST media (Table S3). 253
254
Considering all application media for 1000 cm2 fomite surface area, the average SRE for the 255
experimental data at 0 min was 21 ± 6.9% for acrylic, 26 ± 3.1% for galvanized steel, and 42 ± 256
19.2% for laminate (Figure 1B). The type III test of fixed effects for 1000 cm2 at 0 min was 257
significant for fomite type (p<0.0001), application media (p=0.0037) and the interaction between 258
fomite type and application media (p=0.0998) (Table S5). The laminate fomite yielded the 259
highest SRE while acrylic fomite gave the lowest SRE irrespective of the application media. The 260
use of TSB resulted in higher SRE while PBST and water had statistically equivalent SREs 261
(Table S6). At 20 min, the average SRE for the 1000 cm2 fomite surface area was 2 ± 1.4% or 262
less for all surfaces and application media (Figure 1B). The type III test of fixed effects for 1000 263
cm2 at 20 min was significant for fomite type (p<0.0001) and application media (p=0.0053) but 264
not significant for the interaction between fomite type and application media (p=0.3720) (Table 265
S7). The laminate fomite had the highest SRE while acrylic and galvanized steel had lower and 266
comparable SREs. The use of TSB and water as application media resulted in a higher SRE than 267
PBST (Table S6). 268
269
SRE versus decay for bacteriophage P22 270
The method employed to determine SRE includes the loss due to decay. To separate this loss 271
from the SRE, bacteriophage P22 was directly applied onto a Petri dish (using TSB and water) 272
and decay was quantified as described in the Methods section using the single agar layer method. 273
The decay rates for bacteriophage P22 were 7.97x10-2 hr-1 when applied in TSB and 6.81x10-2 274
on June 20, 2018 by guesthttp://aem
.asm.org/
Dow
nloaded from
13
hr-1when applied in water. After 1 hr, when the 5-μl droplets were visibly dry on the Petri dish; 275
majority of the applied bacteriophage P22 were still infective (89.4 ± 6.7% in TSB and 87 ± 276
7.9% in water). This was substantially higher than the SREs at 20 min employing the double 277
agar layer method which was 0% in water and 0.62 ± 1.3% in TSB for the 100 cm2 acrylic fomite 278
and 0.76 ± 1.6% in water and 0.69 ± 1.5% in TSB for the 1000 cm2 acrylic fomite. Even at 24 hr 279
2 – 5% of bacteriophage P22 was detectable using the single agar method (Figure 2). These 280
results indicate that low or zero SRE may not always indicate absence of the target because SREs 281
also include loss due to sample recovery. 282
283
Impact of wetting agent at varying relative humidity 284
From the above experiment, it was clear that significant portion of the bacteriophage P22 was 285
still active on the fomite at 20 min and the recovery material was unable to recover the dried 286
sample. To enhance recovery, TSB was applied to the fomite as a wetting agent for SREs at 20 287
min (Figure 3). Each point on the distribution represents the experimental data for each fomite 288
type, application media, RH and TSB wetting combination. The type III test of fixed effects 289
using equation (3) was significant for application media (p<0.0001), RH (p<0.0001), the 290
interaction between fomite type and application media (p=0.0001), the interaction between 291
fomite type and RH (p=0.0048) and the interaction between application media and RH 292
(p<0.0001). It was not significant for fomite type (p=0.7634). Use of TSB wetting step was 293
significantly different from not using it (p<0.0001) (Table S8). The TSB wetting step improved 294
the mean SRE for all cases. For both TSB wetting and no TSB wetting, bacteriophage P22 295
applied in TSB media resulted in a higher SRE than the PBST and water. Exception to this was 296
the acrylic and galvanized steel where water gave higher SRE at 55 – 58% RH range (Table S9). 297
on June 20, 2018 by guesthttp://aem
.asm.org/
Dow
nloaded from
14
298
Overall, regardless of wetting agent, a higher average SRE result was primarily observed at RH 299
of 28 – 32% and 55 – 58%. The SRE values for both these humidity ranges were not statistically 300
different from each other. Mean predicted SRE was predominantly lower for RH range of 9 –301
23% compared to the other two ranges. When water was used as the application media, the 302
highest average SRE was always obtained for 55 – 58% RH and the lowest for 9 – 23% RH 303
range (Table S9). The effects of RH on SRE for the other application media (TSB and PBST) 304
were less obvious than water. 305
306
DISCUSSION 307
Once decontamination has been conducted on an indoor site due to a viral outbreak or 308
bioterrorism event, environmental monitoring and quantitative microbial risk assessment 309
modeling will help determine the risk to human health and if the indoor site can be declared 310
“clean” (16, 17). When monitoring fomites for viruses near the detection limit, the results from 311
the linear regression equation suggest that the sampling priority should be for 100 cm2 laminate 312
fomite. At both sampling times (0 min and 20 min) and both fomite surface areas evaluated (100 313
cm2 and 1000 cm2), laminate resulted in a higher SRE compared to the other fomites under the 314
same conditions. An increase in the fomite surface area from 100 to 1000 cm2 decreased the 315
average SRE at 0 min by approximately 25% for acrylic, 40% for galvanized steel, and 50% for 316
laminate (Figure 1). A lower SRE for the larger surface area was expected because the surface 317
density was also lower. Previous studies suggest that one method may not fit all scenarios and 318
sampling for larger fomite surface areas, use of alternative recovery material may be more 319
appropriate (12). Wipe methods are generally used for fomite surface areas of 10 – 25 cm2 but it 320
on June 20, 2018 by guesthttp://aem
.asm.org/
Dow
nloaded from
15
is unknown what influence fomite surface area may have on the SRE (12). Low surface 321
densities will require sampling of larger surface areas. Given that the SRE at 1000 cm2 was 322
lower than the SRE at 100 cm2 and SRE includes decay, sampling at low surface densities must 323
be carried out with caution. 324
325
In general, the application medium TSB produced higher SREs than PBST and water. TSB is an 326
organic medium used for the growth of bacteria and may have properties that were more 327
stabilizing for the bacteriophage P22 on the fomite than other media. It has been suggested that 328
suspension in more complex media may affect resistance to desiccation (30). Most of the SRE 329
studies reported earlier used organic media to suspend viral particles before applying to the 330
fomites (Table 1). The higher SRE in TSB application media suggest that application media may 331
also influence the SRE, especially at low surface densities. 332
333
The most dramatic reduction in the average SRE of bacteriophage P22 from the fomite was with 334
time (0 min vs. 20 min). Initially, inactivation of bacteriophage P22 could be the main reason for 335
this loss in SRE when the sample was dry on the fomite (20 min). Most of the rapid inactivation 336
occurs during the period of desiccation when bacteriophage P22 becomes less stable on the 337
fomites compared to a liquid medium (21). In addition, the concentration that was applied to the 338
fomite was rather low, close to the limit of detection of the plaque assay. Viral survival 339
increases with increase in concentration which can stabilize the virus against environmental 340
stressors (5). On average, less than 3% of bacteriophage P22 was recoverable after 20 min on 341
the fomite. The SREs reported at 20 min varied widely from 3 – 98.4% (Table 1). Each of the 342
studies had a different experimental approach for determining the SRE from fomites which may 343
on June 20, 2018 by guesthttp://aem
.asm.org/
Dow
nloaded from
16
account for the broad range of SREs reported. Keswick et al. (19), using cotton swabs recovered 344
rotavirus, poliovirus and bacteriophage f2 immediately after applying the samples to the fomite. 345
Similarly, cotton swabs were used to recover norovirus and rotavirus dried for 15 min on the 346
fomite by Scherer et al. (29). Taku et al. (37) evaluated three methods, moistened cotton swabs 347
or nylon filter, fomite contact with elution buffer and aspiration, and scraping with aspiration to 348
recover feline calicivirus dried for 15 min on the fomite. Recovery materials antistatic cloth, 349
cotton swab and polyester swab were evaluated by sampling bacteriophage MS2 dried for 45 min 350
on the fomite(18). Sattar et al. (28), analyzed the SRE of human rhinovirus 14 dried on 1 cm 351
diameter disks for 1 hr, then eluted the virus by submerging the disk in 1-ml of tryptose 352
phosphate broth and sonication. It is evident that the number of parameters influencing the SREs 353
is rather large posing a challenge for simple comparison. 354
355
A positive sample result indicates surface contamination and potential risk of exposure. 356
However, a negative result does not entirely assure the absence of infectious agents and the 357
absence of the potential risk of exposure (29). Following the same protocol, Masago et al. (21) 358
found bacteriophage P22 to survive for 36 hr on 10 cm2 fomites (aluminum, ceramic, glass, 359
plastic, stainless steel, and laminate) when applied at a surface density of approximately 107 360
PFU/cm2 (Table 1). The decay rate of bacteriophage P22, reported in Masago et al. (21), for the 361
plastic fomite was 5.2x10-3 hr-1. When eliminating the recovery method by applying the 362
bacteriophage P22 (surface density 2.5 ± 0.9 PFU/cm2) directly onto the Petri dish (plastic 363
fomite), 2 – 5% of bacteriophage P22 could be detected at 24 hr. The decay rate for 364
bacteriophage P22 on Petri dish was estimated to be 7.97x10-2 hr-1 when applied in TSB and 365
6.81x10-2 hr-1 when applied in water. The differences between the decay rates in Masago et al. 366
on June 20, 2018 by guesthttp://aem
.asm.org/
Dow
nloaded from
17
(21) and this study were most likely due to the sample concentrations, higher initial titers have 367
shown to extend survival on fomites (5). As seen in Figure 2, at 1 hr (5-μl droplets were visibly 368
dry) an average of 88.2 ± 7.3% of the applied bacteriophage P22 was still active. Majority of the 369
loss (40 – 60%) occurred between hours 1 and 2. Compared to this, the average SRE from the 370
100 cm2 and 1000 cm2 acrylic surface at 20 min (1-μl droplets visibly dry) was less than 1% 371
(Figure 2). Survival of organisms on fomites is known to be agent-specific and ranges from 0.75 372
hr for rotavirus to 90 days for astrovirus (Table 1). Temperature, RH, fomite surface area, and 373
sample concentration are all known to effect survival (5, 33, 40). Knowledge of the organisms’ 374
response to environmental stress on the fomite is important in determining the appropriate 375
detection methods and employing clean up strategies. 376
377
The results of the experiment designed to separate the decay from sample recovery (Figure 2) 378
revealed that for surface densities of 0.4 – 4 PFU/cm2, SRE was low due to poor efficiency of 379
recovery method rather than decay. The TSB wetting step improved the SRE for all cases at 20 380
min (Table S9). SRE results doubled in the majority of the cases, especially when the 381
application media was TSB. However this TSB wetting step, the combination of the scraping 382
from the disposable spreader and the TSB wetting solution applied onto the fomite, may have 383
physically dislodged the viral particles resulting with a higher SRE than without the TSB wetting 384
step (37). It can also be speculated that the bacteriophage P22 may adhere strongly to the fomite 385
surface after drying or attach to an imperfection on the fomite and the sampling material cannot 386
desorb the virus off the fomite. Surface roughness has been shown to influence adhesion and 387
cell retention to fomites, which can affect recovery (32, 39). In this study, surface roughness was 388
on June 20, 2018 by guesthttp://aem
.asm.org/
Dow
nloaded from
18
not measured. The addition of the TSB wetting step demonstrates the potential to further desorb 389
viruses from the fomite and improve SRE. 390
391
The RH and temperature are crucial parameters to viral survival on fomites (Table 1) (4, 5, 33). 392
A higher SRE was observed for bacteriophage P22 at RH ranges of 28 – 32% and 55 – 58% 393
regardless of the use of a wetting agent (Figure 3). At RH range of 9 – 23%, lowest SREs were 394
obtained (Table S9). The combination of application media and RH may also play a significant 395
role in SRE. Bacteriophage P22 applied in water consistently had the highest SRE at 55 – 58% 396
and the lowest SRE at 9 – 23%. However, for the RH ranges evaluated, its effect was not as 397
obvious for the other application media. The interaction between RH and application media may 398
be a useful parameter in implementing sampling strategies. 399
400
CONCLUSIONS 401
Efficient sample recovery and detection methods are essential in determining the exposure of 402
humans to viruses and the resulting risk in a contaminated indoor environment. The SRE of 403
bacteriophage P22 from fomites at concentrations near the limit of detection was most influenced 404
by time of sampling, fomite surface area, use of a wetting agent and RH. The observations made 405
here using bacteriophage P22 as a surrogate highlight some of the factors that must be 406
considered when sampling for very low surface densities of threat agents. Understanding the 407
contributions of decay and recovery in the overall measured SREs under various conditions and 408
the parameters affecting them will assist in implementing appropriate sampling methods and 409
decontamination strategies. 410
on June 20, 2018 by guesthttp://aem
.asm.org/
Dow
nloaded from
19
411
ACKNOWLEDGMENTS 412
This study was supported by the Center for Advancing Microbial Risk Assessment funded by the 413
U.S. Environmental Protection Agency and Department of Homeland Security grant number 414
R83236201. We thank Rebecca Ives, Yoshifumi Masago, and Tomoyuki Shibata at Michigan 415
State University for their technical support. 416
417
LIST OF TABLES 418
Table 1 - Parameters from survival and SRE studies evaluating viruses applied to fomites. 419
420
LIST OF FIGURES 421
Figure 1- Experimental SRE of bacteriophage P22 from fomites acrylic, galvanized steel, and 422
laminate. Bacteriophage P22 was applied in media PBST, TSB and water to fomites surface 423
areas of (a) 100 cm2 and (b) 1000 cm2. The pre-moistened wipe recovered bacteriophage P22 at 424
the initial application time (0 min) and after drying (20 min). Each bar represents the average of 425
nine plates and the error bars represent the standard deviation of those nine plates. 426
427
Figure 2– Loss of bacteriophage P22 due to decay versus the loss due to sample recovery and 428
decay. Survival of bacteriophage P22 is signified by circles, ● applied in TSB and ○ applied in 429
water, and each point represents the mean and standard deviation of twelve plates. SRE from 100 430
cm2 is signified with triangles,▼applied in TSB and ∆ applied in water. SRE from 1000 cm2 is 431
signified by squares, ■ applied in TSB and □ applied in water. Each point of the SRE data 432
represents the mean and standard deviation of nine plates. 433
on June 20, 2018 by guesthttp://aem
.asm.org/
Dow
nloaded from
20
434
Figure 3 - The experimental impact of RH and TSB wetting agent on the SRE of bacteriophage 435
P22 after drying (20min) on 100 cm2 fomite surface area. Bacteriophage P22 was sampled with 436
pre-moistened wipes at RH ranges of (a) 9 – 23%, (b) 28 – 32%, and (c) 55 – 58%. Each dot on 437
the distribution represents the SRE from a single fomite (acrylic, galvanized steel, and laminate) 438
and application media (PBST, TSB, and water) combination. Those with the highest SRE were 439
labeled. The solid line of the box plot represents the median SRE and the dashed line represents 440
the mean SRE. The box plot whiskers above and below the box indicates the 90th and 10th 441
percentiles, respectively. 442
443
on June 20, 2018 by guesthttp://aem
.asm.org/
Dow
nloaded from
21
Table 1 4
Organism Sample
Concentration
Fomite Area
(cm2)
Application
Volume
(μl)
Surface
Density
Application
Medium
Relative
Humidity
( % )
Temperature
(oC)
Survival or
SRE
Ref.
Survival studies
Alphaviruses
Ebola virus
Lassa virus
1.5x107–4.5x1010 PFU/ml 0.25 NRa 6.4x106 PFU/cm2 (A), 7.6x107 PFU/cm2 (E), 5.6x107 PFU/cm2
NR 30–40 20, 25 6–14 days (26)
Astrovirus 1x105–5x105 PFU 1,3 20, 50 3.3x104-1.6x105PFU/cm2 Phosphate buffered saline (PBS) or 20% fecal suspension (FS)
90±5 4, 20 10–90 days (2)
Avian metapneumovirus
Avian influenza virus
3.1x106–6.3x106 TCID50/mlb
1 10 3.4x104–6.3x104 TCID50/cm2 NR NR NR 1–6 days (39)
Bacteriophage P22 1x109 PFU/ml 10 10 107 PFU/cm2 NR 50 25 859 hours (21)
Calicivirus 107 TCID50/ml 1 20 2.0x105 TCID50/cm2 NR NR NR 4–72 hours (10)
Coronavirus 104–105 MPNc 1 10 104–105 MPN/cm2 Cell culture medium 20±3, 50 ±3, 80 ±3 4, 20, 40 0.25–28 days (8)
Coronavirus 107 PFU/ml 0.79 10 1.3x105 PFU/cm2 PBS 55, 70 21 3–6 hours (35)
Feline calicivirus
Norwalk virus
109 PFU/ml (FC),
106 RT-PCRU/ml (N)
25 NR 4x107PFU/cm2 (FC), 4x102 RT-PCRU/cm2
20% FS in PBS 75–88 22±2 7 days (11)
Hepatitis A virus
Polio virus
10–fold dilution 0.79 10 N/A 10% FS in saline 25±5, 55±5, 80±5, 95±5
5, 20, 35 4–96 hours HAV,
4–12 hours PV
(22)
Hepatitis A virus
Human rotavirus
Enteric adenovirus
Poliovirus
NR 1 20, 50, 100 N/A PBS or FS 50±5, 85±5, 90±5 4, 20 30–60 days HAV,
30–60 days HRV,
5–60 days ADV,
5–60 days PV
(1)
Rotavirus
Poliovirus
Bacteriophage f2
103–105 PFU/ml 250 Misted N/A Distilled water, distilled water with 10% FS
NR NR 0.75–1.5 hours (19)
Influenza A virus 2x108 PFU/ml NR 20 N/A NR 50–60 22±2 6–24 hours (23)
Influenza A virus 106 TCID50/ml 1 10 104 TCID50/cm2 Eagle minimal essential medium with 25mM HEPES and Earle’s salts
30–50 21–28 2 hours–17 days (38)
Influenza A virus 1.5x108 TCID50/ml 2 10 7.5x105 TCID50/cm2 1% BSA 23–24 17–21 4–9 hours (15)
Influenza A and B virus 103–104 TCID50/0.1 ml 7.07–19.63 100 50–1.4x103 TCID50/cm2 NR 35–40(A),
55–56(B)
27.8–28.3(A), 26.7–28.9(B)
24–48 hours (3)
Parainfluenza 1.5x100, 1.5x103,
1.5x104 TCID50/ml
32 500 0.02–2.0x102 TCID50/cm2 Minimun essential medium with Earle’s salts
NR 22 6–10 hours (7)
on June 20, 2018 by guesthttp://aem
.asm.org/
Dow
nloaded from
22
a. NR-not reported 5 b. TCID50-median tissue culture infective dose 6 c. MPN-most probable number 7 d. N/A – not applicable, not able to calculate surface density from information reported 8 9
Rhinovirus 107 PFU/ml 0.79 10 1.3x105 PFU/cm2 Tryptose phosphate broth, bovine mucin, human nasal discharge
20±5, 50±5,
80±5
20±1 2–25 hours (28)
Zaire ebolavirus
Lake Victoria marbugvirus
1x106 TCID50/ml 0.38 20 5.2x104 TCID50/cm2 Guinea pig sera, tissue culture media
55±5 4, 22 14–50 days (24)
SRE studies
Bacteriophage MS2 1x106 PFU/ml 25 5 3.7 PFU/cm2 50% solution of TSB and dilution buffer (5 mM NaH2PO4 and 10 mM NaCl)
45–60 20–22 7–40% (18)
Feline calicivirus 7.0x105–1.3x106 TCID50/100µl
25.8, 929, 5,290
20 26–104 TCID50/cm2 10% FS in PBS NR NR 3–71% (37)
Rotavirus
Poliovirus
Bacteriophage f2
103–105 PFU/ml 250 Misted N/A Distilled water, distilled water with 10% FS
NR NR 16.8±6% (R), 42.3±1.9% (P), 10.6±5.7%(B)
(19)
Norovirus
Rotavirus
2.0x107 RT-PCRU/ml(N)
2.0x105 RT-PCRU/ml(R)
10 100 2.0x103 , 2.0x104 RT-PCRU/cm2 (N)
2.0x101 , 2.0x102 RT-PCRU/cm2 (R)
10% PBS NR NR 10.3±13.0–51.9±38.5% (N)
5.4±1.5–57.7±25.9% (R)
(29)
Rhinovirus 107 PFU/ml 0.79 10 1.3x105 PFU/cm2 Tryptose phosphate broth, bovine mucin, human nasal discharge
50±5 22 40.3–98.4% (28)
on June 20, 2018 by guesthttp://aem
.asm.org/
Dow
nloaded from
23
Figure 1 450 451 452 453 a. 454 455 456 457 458 459 460 461 462 463 464 465 466 467 468 469 470 471 472 473 474 b. 475 476 477 478 479 480
481 482 483 484 485 486 487 488 489 490 491 492 493 494 495
0
20
40
60
80
100
120
Acr
ylic
Gal
vani
zed
Ste
el
Lam
inat
e
PBST PBSTTSB TSBWater Water
SR
E (
%)
100 cm2
0 min 20 min
Application Medium
0 min
0
20
40
60
80
100
120
PBST TSB Water
20 min
PBST TSBWater
Acr
ylic
G
alva
nize
d S
teel La
min
ate
SR
E (
%)
1000 cm2
Application Medium
on June 20, 2018 by guesthttp://aem
.asm.org/
Dow
nloaded from
24
496 497 498 499 Figure 2 500 501 502 503
Log Time (hr)
0.1 1 10 100
Su
rviv
al o
r S
RE
(%
)
0
20
40
60
80
100No Loss
Loss due to decay (single agar layer method)
Loss due to sample recovery and decay (double agar layer method)
on June 20, 2018 by guesthttp://aem
.asm.org/
Dow
nloaded from
25
Figure 3 504 505
(a) 9-23% rH
SR
E (
%)
0
10
20
30
40
No Wetting
TSB Wetting
(b) 28-32% RH
0
10
20
30
40
No Wetting
TSB Wetting
(c) 55-58% RH
0
10
20
30
40
No Wetting
TSB Wetting
Acrylic + TSB
Laminate + TSB Laminate + TSB
Acrylic + TSB
Laminate + TSB
Acrylic + TSB
on June 20, 2018 by guesthttp://aem
.asm.org/
Dow
nloaded from
26
REFERENCES 506 507 1. Abad, F. X., R. M. Pinto, and A. Bosch. 1994. Survival of enteric viruses on 508
environmental fomites Appl. Environ. Microbiol. 60:3704-3710. 509 510 2. Abad, F. X., C. Villena, S. Guix, S. Caballero, R. M. Pinto, and A. Bosch. 2001. 511
Potential role of fomites in the vehicular transmission of human astroviruses. Appl. 512 Environ. Microbiol. 67:3904-3907. 513
514 3. Bean, B., B. M. Moore, B. Sterner, L. R. Peterson, D. N. Gerding, and H. H. Balfour. 515
1982. Survival of influenza viruses on environmental surfaces J. Infect. Dis. 146:47-51. 516 517 4. Boone, S. A., and C. P. Gerba. 2010. The prevalence of human parainfluenza virus 1 on 518
indoor office fomites. Food Environ. Virol. 2:41-46. 519 520 5. Boone, S. A., and C. P. Gerba. 2007. Significance of fomites in the spread of respiratory 521
and enteric viral disease. Appl. Environ. Microbiol. 73:1687-1696. 522 523 6. Boxman, I. L. A., R. Dijkman, N. Loeke, G. Hagele, J. Tilburg, H. Vennema, and M. 524
Koopmans. 2009. Environmental swabs as a tool in norovirus outbreak investigation, 525 including outbreaks on cruise ships. J. Food Prot. 72:111-119. 526
527 7. Brady, M. T., J. Evans, and J. Cuartas. 1990. Survival and disinfection of 528
parainfluenza viruses on environmental surfaces. Am. J. Infect.Control 18:18-23. 529 530 8. Casanova, L. M., S. Jeon, W. A. Rutala, D. J. Weber, and M. D. Sobsey. 2010. 531
Effects of Air Temperature and Relative Humidity on Coronavirus Survival on Surfaces. 532 Appl. Environ. Microbiol. 76:2712-2717. 533
534 9. Centers for Disease Control and Prevention (CDC). 2002. Emergency preparedness 535
and response: comprehensive procedures for collecting environmental samples for 536 culturing Bacillus anthracis http://www.bt.cdc.gov/Agent/Anthrax/environmental-537 sampling-apr2002.asp. 538
539 10. Clay, S., S. Maherchandani, Y. S. Malik, and S. M. Goyal. 2006. Survival on 540
uncommon fomites of feline calicivirus, a surrogate of noroviruses. Am. J. Infect. Control 541 34:41-43. 542
543 11. D'Souza, D. H., A. Sair, K. Williams, E. Papafragkou, J. Jean, C. Moore, and L. 544
Jaykus. 2006. Persistence of caliciviruses on enviromnental surfaces and their transfer to 545 food. Int. J. Food Microbiol. 108:84-91. 546
547 12. Edmonds, J. M. 2009. Efficient methods for large-area surface sampling of sites 548
contaminated with pathogenic microorganisms and other hazardous agents: current state, 549 needs, and perspectives. Appl. Microbiol. Biotechnol. 84:811-816. 550
551
on June 20, 2018 by guesthttp://aem
.asm.org/
Dow
nloaded from
27
13. Enriquez, C., A. Alum, E. M. Suarez-Rey, C. Y. Choi, G. Oron, and C. P. Gerba. 552 2003. Bacteriophages MS2 and PRD1 in turfgrass by subsurface drip irrigation. J. 553 Environ. Engin. ASCE 129:852-857. 554
555 14. EPA, U. S. 2001. Method 1602: male-specific (F+) and somatic coliphage in water by 556
single agar layer (SAL) procedure. , EPA 821-R-01-029, U.S. Environmental Protection 557 Agency, Washington, DC. 558
559 15. Greatorex, J. S., P. Digard, M. D. Curran, R. Moynihan, H. Wensley, T. Wreghitt, 560
H. Varsani, F. Garcia, J. Enstone, and J. S. Nguyen-Van-Tam. 2011. Survival of 561 influenza A(H1N1) on materials found in households: implications for infection control. 562 PLoS One 6:1-6. 563
564 16. Haas, C. N., J. B. Rose, and C. P. Gerba. 1999. Quantitative microbial risk assessment. 565
John Wiley, New York. 566 567 17. Herzog, A. B., S. D. McLennan, A. K. Pandey, C. P. Gerba, C. N. Haas, J. B. Rose, 568
and S. A. Hashsham. 2009. Implications of limits of detection of various methods for 569 Bacillus anthracis in computing risks to human health. Appl. Environ. Microbiol. 570 75:6331-6339. 571
572 18. Julian, T. R., F. J. Tamayo, J. O. Leckie, and A. B. Boehm. 2011. Comparison of 573
surface sampling methods for virus recovery from fomites. Appl. Environ. Microbiol. 574 77:6918-6925. 575
576 19. Keswick, B. H., L. K. Pickering, H. L. Dupont, and W. E. Woodward. 1983. Survival 577
and detection of rotaviruses on environmental surfaces in day care centers. Appl. 578 Environ. Microbiol. 46:813-816. 579
580 20. Li, S., J. N. S. Eisenberg, I. H. Spicknall, and J. S. Koopman. 2009. Dynamics and 581
control of infections transmitted from person to person through the environment. Am. J. 582 Epidemiol. 170:257-265. 583
584 21. Masago, Y., T. Shibata, and J. B. Rose. 2008. Bacteriophage P22 and Staphylococcus 585
aureus attenuation on nonporous fomites as determined by plate assay and quantitative 586 PCR. Appl. Environ. Microbiol. 74:5838-5840. 587
588 22. Mbithi, J. N., V. S. Springthorpe, and S. A. Sattar. 1991. Effect of relative humidity 589
and air temperature on survival of hepatitis A virus on environmental surfaces. Appl. 590 Environ. Microbiol. 57:1394-1399. 591
592 23. Noyce, J. O., H. Michels, and C. W. Keevil. 2007. Inactivation of influenza A virus on 593
copper versus stainless steel surfaces. Appl. Environ. Microbiol. 73:2748-2750. 594 595
on June 20, 2018 by guesthttp://aem
.asm.org/
Dow
nloaded from
28
24. Piercy, T. J., S. J. Smither, J. A. Steward, L. Eastaugh, and M. S. Lever. 2010. The 596 survival of filoviruses in liquids, on solid substrates and in a dynamic aerosol. J. Appl. 597 Microbiol. 109:1531-1539. 598
599 25. Rusin, P., S. Maxwell, and C. Gerba. 2002. Comparative surface-to-hand and fingertip-600
to-mouth transfer efficiency of gram-positive bacteria, gram-negative bacteria, and 601 phage. J. Appl. Microbiol. 93:585-592. 602
603 26. Sagripanti, J. L., A. M. Rom, and L. E. Holland. 2010. Persistence in darkness of 604
virulent alphaviruses, Ebola virus, and Lassa virus deposited on solid surfaces. Arch. 605 Virol. 155:2035-2039. 606
607 27. SAS/STAT(R) 9.22 User's Guide. 2012. Type III Test of Fixed Effects. SAS Institute 608
Inc.,http://support.sas.com/documentation/cdl/en/statug/63347/HTML/default/viewer.htm609 #statug_glimmix_a0000001479.htm. 610
611 28. Sattar, S. A., Y. G. Karim, V. S. Springthorpe, and C. M. Johnsonlussenburg. 1987. 612
Survival of human rhinovirus type 14 dried onto nonporous inanimate surfaces: effect of 613 relative humidity and suspending medium. Can. J. Microbiol. 33:802-806. 614
615 29. Scherer, K., D. Made, L. Ellerbroek, J. Schulenburg, R. Johne, and G. Klein. 2009. 616
Application of a swab sampling method for the detection of norovirus and rotavirus on 617 artificially contaminated food and environmental surfaces. Food Environ. Virol. 1:42-49. 618
619 30. Shams, A. M., L. J. Rose, L. Hodges, and M. J. Arduino. 2007. Survival of 620
Burkholderia pseudomallei on environmental surfaces. Appl. Environ. Microbiol. 621 73:8001-8004. 622
623 31. Shen, C. P., M. S. Phanikumar, T. T. Fong, I. Aslam, S. P. McElmurry, S. L. Molloy, 624
and J. B. Rose. 2008. Evaluating bacteriophage P22 as a tracer in a complex surface 625 water system: The Grand River, Michigan. Environ. Sci. Technol. 42:2426-2431. 626
627 32. Silva, S., P. Teixeira, R. Oliveira, and J. Azeredo. 2008. Adhesion to and viability of 628
Listeria monocytogenes on food contact surfaces. J. Food Prot. 71:1379-1385. 629 630 33. Sinclair, R., S. A. Boone, D. Greenberg, P. Keim, and C. P. Gerba. 2008. Persistence 631
of category A select agents in the environment. Appl. Environ. Microbiol. 74:555-563. 632 633 34. Sinclair, R. G., J. B. Rose, S. A. Hashsham, C. P. Gerba, and C. N. Haas. 2012. 634
Criteria for selection of surrogates used to study the fate and control of pathogens in the 635 environment. Appl. Environ. Microbiol. 78:1969-1977. 636
637 35. Sizun, J., M. W. N. Yu, and P. J. Talbot. 2000. Survival of human coronaviruses 229E 638
and OC43 in suspension and after drying on surfaces: a possible source of hospital-639 acquired infections. J. Hosp. Infect. 46:55-60. 640
641
on June 20, 2018 by guesthttp://aem
.asm.org/
Dow
nloaded from
29
36. Spicknall, I. H., J. S. Koopman, M. Nicas, J. M. Pujol, S. Li, and J. N. S. Eisenberg. 642 2010. Informing optimal environmental influenza interventions: how the host, agent, and 643 environment alter dominant routes of transmission. PLoS Comput. Biol. 6:11. 644
645 37. Taku, A., B. R. Gulati, P. B. Allwood, K. Palazzi, C. W. Hedberg, and S. M. Goyal. 646
2002. Concentration and detection of caliciviruses from food contact surfaces. J. Food 647 Prot. 65:999-1004. 648
649 38. Thomas, Y., G. Vogel, W. Wunderli, P. Suter, M. Witschi, D. Koch, C. Tapparel, 650
and L. Kaiser. 2008. Survival of influenza virus on banknotes. Appl. Environ. 651 Microbiol. 74:3002-3007. 652
653 39. Tiwari, A., D. P. Patnayak, Y. Chander, M. Parsad, and S. M. Goyal. 2006. Survival 654
of two avian respiratory viruses on porous and nonporous surfaces. Avian Dis. 50:284-655 287. 656
657 40. Vasickova, P., I. Pavlik, M. Verani, and A. Carducci. 2010. Issues concerning survival 658
of viruses on surfaces. Food Environ. Virol. 2:24-34. 659 660 661
on June 20, 2018 by guesthttp://aem
.asm.org/
Dow
nloaded from