COMPARISON OF THE RECOVERY OF BACILLUS ANTHRACIS AND BACILLUS
THURINGIENSIS SPORES FROM POROUS MEDIA: CONSIDERING TIME AND
MOISTURE CONTENT
By
Bharathi Murali

A THESIS
Submitted to
Michigan State University
in partial fulfillment of the requirements
for the degree of
Biosystems Engineering – Master of Science
2014

ABSTRACT
COMPARISON OF THE RECOVERY OF BACILLUS ANTHRACIS AND BACILLUS
THURINGIENSIS SPORES FROM POROUS MEDIA: CONSIDERING TIME AND
MOISTURE CONTENT
By
Bharathi Murali
Bacillus anthracis spores are agents used for biological warfare owing to high virulence,
and extreme resistance to environmental stresses with negligible degradation in viability. There
is an interest in understanding the persistence and subsequent health risks due to B. anthracis
spores in the indoor environment during bioterror attacks, but neither the time-dependent
behavior of B. anthracis spores on porous media nor the suitability between B. anthracis and its
recommended experimental surrogate, B. thuringiensis spores, are well known. This work
specifically focused on recovery of B. anthracis and B. thuringiensis spores from HVAC filters,
a porous fomite, considering the effects of sampling time and moisture content using culturebased quantification. Effects of time significantly affected the recovery of Bacillus spores.
Although there was an overall reduction in viability of Bacillus spores over time, the reduction in
recovery of B. anthracis and B. thuringiensis varied remarkably. Clumping in B. thuringiensis
spores and accelerated dehydration rates at higher moisture levels might have resulted in
increased observed recovery of B. thuringiensis. Hence it is concluded that, B. thuringiensis
spores is not an ideal experimental surrogate for B. anthracis in recovery studies. Inclusion of
time and moisture emphasizes the importance of variation in recoveries and inferences between
the surrogate and B. anthracis spores. The data generated in this study will be useful for
persistence studies that measure reduction or decay but are dependent on recovery. Furthermore,
this information will inform quantitative microbial risk assessment (QMRA) models that
estimate the risk of exposure to B. anthracis spores in indoor environments.

I dedicate this thesis to Harshavardhan Nirmal, a.k.a. Jutty

iii

ACKNOWLEDGMENTS

I would like to express my sincere gratitude to my advisor Dr. Jade Mitchell for all her
support, patience, and motivation during my Master’s program. I am thankful for the opportunity
provided to me by the Department of Biosystems and Agricultural Engineering to obtain my
degree. I also express my appreciation to Dr. Thomas Voice and the Department of Civil and
Environmental Engineering for their support in transferring my major.
I would like to extend my sincere thanks to my thesis committee members Dr. Joan Rose
and Dr. Evangelyn Alocilja for their constant help, motivation, and insightful inputs. I have
learned so much from Drs. Mitchell, Rose and Alocilja, and I am very lucky to have had their
guidance. I would also like to thank Dr. Kyle S. Enger, Andrew Bruce, and Carol Flegler for
their assistance in technical and experimental aspects, during the development of my research.
I sincerely thank Dr. Daniel Guyer and Dr. Ajit Srivatsava for providing me with a great
deal of encouragement and optimism from an academic perspective. Furthermore, I would like to
thank the Department of Biosystems and Agricultural Engineering and the Merle and Catherine
Esmay Scholarship for providing financial support. I also would like to thank my friends and
colleagues in the department for providing a cheerful and productive work environment.
I wholeheartedly express my love and appreciation to all my friends back home, who
morally supported me during tough times.
Last but not least, I express my greatest adoration and gratitude to my parents D. Murali
and Saraswathi Murali for being my biggest source of love and support during every moment of
my education.
This study would not have been possible without all of your help. Thank you!

iv

TABLE OF CONTENTS

LIST OF TABLES ........................................................................................................................ vii
LIST OF FIGURES ....................................................................................................................... ix
CHAPTER 1: INTRODUCTION ................................................................................................... 1
1.1. Research Objectives ..................................................................................................... 4
1.1.1. Specific Objective I...................................................................................... 6
1.1.2. Specific Objective II .................................................................................... 6
1.1.3. Specific Objective III ................................................................................... 7
CHAPTER 2: LITERATURE REVIEW ........................................................................................ 8
2.1. Bacillus anthracis Spores ............................................................................................ 8
2.2. Bacillus anthracis spores – a Bioweapon .................................................................... 9
2.3. Inhalation Anthrax ..................................................................................................... 10
2.4. Indoor Air and HVAC Filters .................................................................................... 11
2.5. HVAC Filters – Reliable Sampling Sources for Indoor Microbial Contamination ... 12
2.6. Recovery Studies ....................................................................................................... 14
2.7. Surrogate Selection .................................................................................................... 19
2.8. Risk Assessment Studies............................................................................................ 22
2.9. Knowledge Gaps ........................................................................................................ 23
CHAPTER 3. EVALUATING TIME-DEPENDENT VARIABILITY IN RECOVERY OF
BACILLUS SPORES FROM HVAC FILTERS .......................................................................... 26
3.1. Abstract ...................................................................................................................... 26
3.2. Introduction ................................................................................................................ 27
3.3. Materials and Methods ............................................................................................... 30
3.3.1. Species Used ............................................................................................... 30
3.3.2. Inoculum Preparation .................................................................................. 30
3.3.3. Preparation and Inoculation of HVAC Filters ............................................ 31
3.3.4. Elution of Spores from Filters..................................................................... 32
3.3.5. Culture-based Quantification ...................................................................... 33
3.3.6. Tests for Clumping and Drying Rates ........................................................ 33
3.3.7. Scanning Electron Microscopy (SEM) ....................................................... 34
3.3.8. Data Treatment............................................................................................ 35
3.3.9. Statistical Tests ........................................................................................... 36
3.4. Results ........................................................................................................................ 37
3.4.1. Concentrations of applied and 0 hour recovered spores ............................. 37
3.4.2. Comparison of recovery rates and normalized recovery rates .................... 37
3.4.3. Comparison of B. anthracis and B. thuringiensis spores ............................ 40
3.4.4. Effect of surfactants on clumping of B. thuringiensis spores ..................... 43
3.4.5. Comparison of log reductions between different moisture levels .............. 44
3.4.6. Comparison of Drying Trends of Two Water Contents: ............................ 48
3.4.7. Microscopy ................................................................................................. 48
v

3.4.8. Statistical Analyses: .................................................................................... 51
3.5. Discussion .................................................................................................................. 52
3.5.1. Comparison of Normalization Factors ........................................................ 52
3.5.2. Recovery of Bacillus spores over Time ...................................................... 53
3.5.2.1. Clumping in B. thuringiensis spores ............................................ 54
3.5.2.2. Comparison of drying rates of different moisture contents ......... 56
3.5.3. Surrogate Suitability ................................................................................... 58
3.5.4. Summary ..................................................................................................... 59
3.6. Concluding Remarks .................................................................................................. 61
3.7. Acknowledgements .................................................................................................... 62
CHAPTER 4: DISCUSSION AND CONCLUSIONS ................................................................. 63
4.1. Hypothesis I ............................................................................................................... 63
4.2. Hypothesis II .............................................................................................................. 64
4.3. Hypothesis III............................................................................................................. 67
4.4. Conclusions ................................................................................................................ 69
4.5. Recommendation for Future Work ............................................................................ 70
APPENDICES .............................................................................................................................. 71
APPENDIX A: SUMMARY OF EXPERIMENTAL PLANNING AND DESIGN ........ 72
APPENDIX B: APPLIED AND RECOVERY DATA COLLECTED FROM BACILLUS
ANTHRACIS AND BACILLUS THURINGIENSIS SPORES AT DIFFERENT
MOISTURE LEVELS OVER TIME................................................................................ 77
APPENDIX C: RECOVERY DATA COLLECTED FROM SUPPLEMENTARY
EXPERIMENTS TO TEST FOR CLUMPING AND DRYING RATES........................ 83
APPENDIX D: SCANNING ELECTRON MICROSCOPY ........................................... 88
APPENDIX E: STATISTICAL ANALYSES ................................................................ 100
APPENDIX F: SUMMARY OF REAL-TIME PCR FOR DIFFERENT MOISTURE
LEVELS TO QUANTIFY BACILLUS ANTHRACIS ................................................. 107
REFERENCES ........................................................................................................................... 113

vi

LIST OF TABLES

Table 1. Summary of applied and average 0 hourour recovered concentrations of Bacillus
spores.. .......................................................................................................................................... 39
Table 2. Summary of recovery rates and normalized recovery rates over time (N=9) ................. 39
Table A. 1. Generalized timeline for recovery experiments: planning module for future use ..... 74
Table B. 1. Mean concentrations of master stocks of Bacillus spores .......................................... 78
Table B. 2. Concentrations of stock solutions of Bacillus spores needed for 0.4 mL inoculation
...................................................................................................................................................... .78
Table B. 3. Concentrations of stock solutions of Bacillus spores needed for 0.2 mL inoculation
...................................................................................................................................................... .78
Table B. 4. Summary of plate counts and recovered concentrations of B. anthracis (0.4 mL) .... 79
Table B. 5. Summary of plate counts and recovered concentrations of B. thuringiensis (0.4 mL)
....................................................................................................................................................... 80
Table B. 6. Summary of plate counts and recovered concentrations of B. anthracis (0.2 mL) .... 81
Table B. 7. Summary of plate counts and recovered concentrations of B. thuringiensis (0.2 mL)
....................................................................................................................................................... 82
Table C. 1. Summary of plate counts and recovered concentrations of B. thuringiensis without
Tween-20 (0.4 mL) ....................................................................................................................... 84
Table C. 2. Summary of plate counts and recovered concentrations of B. thuringiensis with
Tween-20 (0.4 mL) ....................................................................................................................... 85
Table C. 3. Summary of weights of filters loaded with 0.4 mL and 0.2 mL water content ......... 86
Table E. 1. Descriptive statistics of the dependent variable of the complete data set ................ 102
Table E. 2. Tests of Between-Subjects effects (Time, Species, Treatment) ............................... 102
Table E. 3. Parameter estimates for complete data set ............................................................... 103
Table E. 4. Descriptive statistics of the dependent variable of surfactant data set ..................... 105
Table E. 5. Tests of Between-Subjects effects (Time, Tween-20) ............................................. 105
Table E. 6. Parameter estimates for surfactant data set .............................................................. 105
Table F. 1. Forward and Reverse Primer Sequences for B. anthracis Sterne spores*................ 108
vii

Table F. 2. Composition of Reaction Mixture ............................................................................ 108
Table F. 3. Summary of Lightcycler® Operation Cycles ........................................................... 108
Table F. 4. Summary of applied and recovered concentrations of B. anthracis Sterne spores
enumerated using qPCR .............................................................................................................. 109

viii

LIST OF FIGURES

Figure 1. Log-reduction from Eq. (1) as a function of time for 0.4 mL volume of inoculation for
both B. anthracis and B. thuringiensis. Error bars indicate standard deviation from the mean. .. 41
Figure 2. Log-reduction from Eq. (1) as a function of time for 0.2 mL volume of inoculation for
both B. anthracis and B. thuringiensis. Error bars indicate standard deviation from the mean. .. 41
Figure 3. Normalized log-reduction from Eq. (2) as a function of time for 0.4 mL volume of
inoculation for both B. anthracis and B. thuringiensis. Error bars indicate standard deviation
from the mean. .............................................................................................................................. 42
Figure 4. Normalized log-reduction from Eq. (2) as a function of time for 0.2 mL volume of
inoculation for both B. anthracis and B. thuringiensis. Error bars indicate standard deviation
from the mean. .............................................................................................................................. 42
Figure 5. Log-reduction from Eq. (1) as a function of time for 0.2 mL volume of inoculation of
B. thuringiensis spores with and without addition of Tween-20. Error bars indicate standard
deviation from the mean. .............................................................................................................. 43
Figure 6. Log-reduction from Eq. (1) as a function of time for B. anthracis spores for both
volumes of inoculation. Error bars indicate standard deviation from the mean. .......................... 46
Figure 7. Log-reduction from Eq. (1) as a function of time for B. thuringiensis spores for both
volumes of inoculation. Error bars indicate standard deviation from the mean. .......................... 46
Figure 8. Normalized log-reduction from Eq. (2) as a function of time for B. anthracis spores for
both volumes of inoculation. Error bars indicate standard deviation from the mean. .................. 47
Figure 9. Normalized log-reduction from Eq. (2) as a function of time for B. thuringiensis spores
for both volumes of inoculation. Error bars indicate standard deviation from the mean. ............ 47
Figure 10. SEM images of B. thuringiensis spores (A, C, E) with addition of Tween-20 and (B,
D, F) without the addition of Tween-20 (More images using SEM are presented in APPENDIX
C)................................................................................................................................................... 50
Figure A. 1. (L) Overall schematics representing sequential processes, inoculation, elution, and
quantification involved in the recovery experiments for a given sampling time. (R) Series of
mechanical processes, sonication, vortexing, and shaking involved in the elution stage. ............ 76
Figure B. 1. Weight of filters (mg) as a function of time for filters loaded with 0.4 mL and 0.2
mL volumes of water .................................................................................................................... 87
Figure C. 1. Weight of filters (mg) as a function of time for filters loaded with 0.4 mL and 0.2
mL volumes of water .................................................................................................................... 87
Figure D. 1. Scattered clusters of B. thuringiensis spores without Tween-20 (200x) .................. 91
ix

Figure D. 2. Spores of B. thuringiensis with Tween-20 (200x) showing lesser clustering as
compared to the absence of Tween-20.......................................................................................... 91
Figure D. 3. Joint clumps of B. thuringiensis spores without Tween-20 (1000x) ........................ 92
Figure D. 4. Clumps of B. thuringiensis spores with Tween-20 (1000x) ..................................... 92
Figure D. 5. Images of B. thuringiensis spores without Tween-20 (1900x)................................. 93
Figure D. 6. Images of B. thuringiensis spores with Tween-20 (1900x) showing an overall lesser
degree of attachment ..................................................................................................................... 93
Figure D. 7. Spores of B. thuringiensis without Tween-20 (2500x)............................................. 94
Figure D. 8. Spores of B. thuringiensis with Tween-20 (2500x).................................................. 94
Figure D. 9. Dense clump of B. thuringiensis spores without Tween-20 (3000x) ....................... 95
Figure D. 10. Spores of B. thuringiensis with Tween-20 (3000x)................................................ 95
Figure D. 11. Spores of B. thuringiensis without Tween-20 (8000x) with strong clustering
observed at higher magnifications ................................................................................................ 96
Figure D. 12. Images of B. thuringiensis spores with Tween-20 (8000x) .................................... 96
Figure D. 13. Surface of clumped B. thuringiensis spores without Tween-20 (18000x) ............ .97
Figure D. 14. Separated spores of B. thuringiensis with Tween-20 (18000x).............................. 97
Figure D. 15. Coiled spores of B. thuringiensis without Tween-20 (20000x).............................. 98
Figure D. 16. Coiled spores of B. thuringiensis with Tween-20 (20000x) similar to Fig. C. 16 .98
Figure D. 17. Connection between two B. thuringiensis spores without Tween-20 (55000x)
observed at very high magnification ............................................................................................. 99
Figure D. 18. Spores of B. thuringiensis with Tween-20 (55000x).............................................. 99
Figure E. 1. Normal distribution of the response variable, log-reduction (LR) for a combined data
set of both B. anthracis and B. thuringiensis using both moisture levels at all sampling times
................................................................................................................................................. …101
Figure E. 2. Scatter plot matrix of standard residuals, observed, and predicted values of the
dependent variable of the complete data set ............................................................................... 104
Figure E. 3. Scatter plot matrix of standard residuals, observed, and predicted values of the
dependent variable of the surfactant data set .............................................................................. 106

x

Figure F. 1. Log-concentration of B. anthracis Sterne spores as a function of time for 0.4 mL
volume of inoculation quantified using culture-based and molecular-based techniques. Error bars
indicate standard deviation from the mean. ................................................................................ 110
Figure F. 2. Log-concentration of B. anthracis Sterne spores as a function of time for 0.2 mL
volume of inoculation quantified using culture-based and molecular-based techniques. Error bars
indicate standard deviation from the mean. ................................................................................ 111

xi

CHAPTER 1: INTRODUCTION
Bacillus endospores are formed in response to acute unfavorable environmental
conditions such as starvation, chemical and thermal stresses, desiccation, pH extremes,
mechanical force, and antibiotics (Setlow, 2003; Nicholson et al., 2000; Sternbach, 2003). When
exposed to favorable conditions these endospores are capable of germinating. Spores of Bacillus
anthracis maintain their viability over decades, with negligible degradation (Dragon & Rennie,
1995). The virulence and durability of B. anthracis spores, along with the ease and affordability
of their manufacture, have made them agents of biological warfare (Inglesby et al., 1999;
Inglesby et al., 2002). The detrimental effect of B. anthracis in biological warfare is well
documented (Cole, 2010; Riedel, 2005; Riedel, 2004; Christopher et al., 1997; Block, 2001).
Although the use of B. anthracis as a bioterror agent has been documented as early as World
War I, it was not until the 2001 anthrax attacks in the United States that B. anthracis and
bioterrorism gained global attention (Inglesby et al., 2002; Riedel, 2005, Sternbach et al., 2003;
Cole, 2010). The aftermath of the 2001 anthrax attacks included economic losses, public chaos,
and political turbulence of significant magnitude, because the spores were delivered in letters
targeting senate and media persons and contaminating government buildings (Day, 2003).
According to a PubMed search result using the keyword “anthrax”— the number of studies
immediately following the 2001 anthrax attacks increased by approximately 414% during the
years 2001 and 2002 compared to the year 2000.
The persistence of B. anthracis endospores in different environmental matrices such as
soil, water, and air makes the possible exposure scenarios more minatory in cases of bioterror
events (Manchee et al., 1981; Dragon et al., 2001; Minett & Dhanda, 1941; Brachman et al.,
1966). Understanding the behavior of aerosolized biothreat agents over time is a major concern
1

of the US Environmental Protection Agency and the Department of Homeland Security.
Considering the persistence of Bacillus endospores in air, it is also vital to quantitatively and
qualitatively monitor the indoor environment over time in scenarios of intentional bioweapon
releases. Because humans spend 94% of the day indoors at work, in enclosed vehicles, and
during indoor recreation, etc., the quality of the air inhaled over time correlates to the quality of
their health (Jenkins et al., 1992; Lai & Nazaroff, 2000). HVAC filters are part of this indoor
environment, which are used in more than 70% of residential and commercial establishments in
the United States, and have been used as sampling sources in cases of an intentional indoor
release or an outdoor infiltration of weaponised endospores (Noris et al., 2011; Kim et al., 2008;
Krauter & Biermann, 2007; Van Cuyk et al., 2012; Stanley et al., 2008; Calfee et al., 2014).
Sampling these filters from different intakes/ outlets of an air handling unit within a given
building could help predict the pattern of spread of pathogens used during the attack (Calfee et
al., 2014).
Sampling and decontamination studies have been successful in documenting the efficacy
of the techniques and materials used, but the recovery over time has been rarely considered. To
our knowledge, the following recovery studies using bacterial spores have been the only studies
carried out on filter media to date: Wang et al. (2001), Moritz et al. (2001), Farnsworth et al.
(2006), Kim et al. (2008), Solon et al. (2012), Stanley et al., (2008); Van Cuyk et al., (2012);
Burton et al. (2005), Maus et al. (2000), Mittal et al. (2011), and Calfee et al. (2014). However,
only four studies to date have considered time-dependent (24 hours or greater) recovery of
bacterial spores: Moritz et al. (2001) – 450 days, Mittal et al. (2011) – 210 days, Kim et al.
(2008) – 28 days, and Maus et al. (2000) – 5 days.

2

It is interesting to find that these filter-based studies have produced some widely differing
results such as: very high (90% or greater) extraction or recovery efficiencies (Wang et al., 2001;
Burton et al., 2005; Farnsworth et al., 2006); high extraction efficiency but poor culturability
(Wang et al., 2001); high extraction efficiency with no change or minimal loss in culturability
(Maus et al., 2000; Burton et al., 2005; Farnsworth et al., 2006; Kim et al., 2008; Mittal et al.,
2011); higher recoveries at high ambience moisture content (Wang et al., 2001; Moritz et al.,
2001); reported high survival or proliferation of microbial spores during higher moisture content
of the filters (Moritz et al., 2001; Solon et al., 2012); and no change in viability or evidence of
proliferation observed in the recovery, under both high and low relative humidity (Maus et al.,
2000).
All of these studies were invariably carried out using the spores of species other than B.
anthracis: B. subtilis, B. atrophaeus, B. golbigii, and B. thuringiensis. Moritz et al. (2001) did
not specify the type of bacterial species used. Although it is an accepted practice to use surrogate
species in experimental research, it is important to test the similarities between the surrogate and
the species of interest experimentally. Most of the commonly used surrogate spores such as B.
subtilis, B. atrophaeus, and B. golbigii are approximately half the diameter and 25% the volume
of B. anthracis spores. Furthermore, these surrogate species lack essential glycoproteins that are
responsible for the formation of the outermost layer of B. anthracis spores, called exosporium
(Hachisuka et al., 1984; Carrera et al., 2007; Greenberg et al., 2010). Therefore, use of such
morphologically and phylogenetically different surrogates in some experiments would yield
different results especially on porous media where the chances of infiltration are directly
correlated to the size and aerodynamic properties of the spores. Bacillus thuringiensis spores are
morphologically and genetically close to B. anthracis which make them ideal surrogate

3

candidates for experimental purposes (Carrera et al., 2007; Greenberg et al., 2010; Tufts et al.,
2013). Although physical parameters of B. thuringiensis and B. anthracis spores are comparable,
time-dependent recovery of either from filters has not been experimentally studied to date.
Therefore, prior to using the suggested best-suited surrogate in place of B. anthracis spores, it is
strongly warranted to test both the species using identical experimental conditions.
Quantitative studies of B. anthracis spores are needed in order to improve understanding
for persistence modeling and risk assessment. The efficacy of decontamination and disinfection
processes following bioterror attacks are evaluated by wipes and swabs and sampling of
recovered spores. This information is integrated with fate-and-transport, and dose-response
models within QMRA frameworks. Therefore, a robust set of recovery data of B. anthracis over
time would prove useful for persistence and risk assessment research in the future. Furthermore,
an evaluation of whether surrogate data, specifically B. thuringiensis spores is suitable for
predicting the behavior of B. anthracis spores on porous media would prove useful for future.
1.1. Research Objectives
The objective of this study was to compare the recovery of B. anthracis Sterne (nonpathogenic surrogate) to B. thuringiensis HD 2-61 spores from HVAC filters over a short-term
period of 168 hours. This study is the first one of its kind to carry out recovery of B. anthracis
and its surrogate B. thuringiensis experimentally in the same study with identical sporulation and
experimental procedures. Hence, the results from this study could help in understanding the
short-term, time-dependent recovery of B. anthracis and B. thuringiensis spores as well as help
in future selection of surrogates and quantification techniques for experimental purposes.

4

Therefore, goals of this study are to:


Quantify and understand the time-dependent recovery of spores of B. anthracis
alongside its surrogate B. thuringiensis from HVAC filters



Investigate if B. thuringiensis and B. anthracis are experimentally similar



Investigate the effects of varied moisture content loaded on the filters on recovery for
both the species
Hypotheses specific to each goal in this study, are presented in the following sub-

sections.

5

1.1.1. Specific Objective I
To quantify the time-dependent recovery of spores of B. anthracis alongside B.
thuringiensis
It is hypothesized that Bacillus spores will have a time-dependent reduction in recovery,
from HVAC filters. Filter based recovery studies of Bacillus spores over time are scarce. Based
on culture-based results from Kim et al. (2008) in which B. golbigii spores were inoculated and
sampled from HVAC filters over a period of about 30 days, it is expected that there will be a
reduction in the recovery of spores from HVAC filters. To achieve this objective, sample
coupons of HVAC filters will be artificially loaded with B. anthracis and B. thuringiensis spores
of known similar concentrations, and recovered at the same time intervals using a series of
mechanical extraction processes. Amounts of recovered spores at each time point were
quantified and will be represented in spores/ filter.
1.1.2. Specific Objective II
To investigate if B. thuringiensis and B. anthracis spores behave similarly in recovery
studies from porous media
It is hypothesized that B. anthracis and B. thuringiensis spores will have identical timedependent recovery from HVAC filters. Most of the published studies in the past have used the
species that belonged to the B. subtilis family. Surrogates such as B. atrophaeus, B. subtilis, and
B. golbigii are approximately half the diameter and 25% the volume of pathogenic B. anthracis.
Therefore, use of such surrogates with morphological and phylogenetic differences in
experiments would yield different results especially on porous media where there are chances of
infiltration which is correlated to the size and aerodynamic properties of the spores. In 2010,
Bacillus thuringiensis, was suggested to be an ideal suitable surrogate for pathogenic B.
6

anthracis sharing morphological and genetic similarities (Carrera et al., 2005; Greenberg et al.,
2010; Tufts et al., 2013). Although Solon et al. (2012) studied the recovery of B. thuringiensis
spores from HVAC filters, there is no experimental evidence to date that B. anthracis and B.
thuringiensis behave identically. The results generated in Objective I will be used to achieve this
objective.
1.1.3. Specific Objective III
To study the effect of different moisture content loaded on the filters on recovery
It is hypothesized that Bacillus spores will have a lower recovery when less water is
applied to the HVAC filter. Recoveries obtained at high moisture contents on filter or fomites
have resulted in greater culturability and recovery, suggesting that there could be greater
detachment or poorer attachment when the sampling matrix is moist (Wang et al., 2001; Moritz
et al., 2001; Nanasaki et al., 2010; Lopez et al., 2013). Therefore the sensitivity of the wet
application method used in this study will be tested. Considering that HVAC systems encounter
frequent conditions of fluctuating moisture levels, and higher moisture content have found to
favor greater culturability of spores, this study will also inform the effect of a saturated and a
lower water content applied to the filter over time. To achieve this objective, sample coupons of
HVAC filters will be artificially loaded with B. anthracis and B. thuringiensis spores of known
similar concentrations with two different moisture contents, and subsequently recovered and
extracted over time as described in Objective I.

Summary of experimental planning and design are provided in detail in APPENDIX-A.
Chapter 2 provides a comprehensive literature review of related research and motivation for this
study. Results and discussion are presented in Chapters 3 and 4, respectively.
7

CHAPTER 2: LITERATURE REVIEW
2.1. Bacillus anthracis Spores
Bacillus is the genus of rod-shaped, endospore forming, Gram-positive bacteria that are
ubiquitous in the environment which includes both pathogenic and non-pathogenic species.
Bacillus species have a wide range of applications such as antibiotic production as well as in
biological warfare (Katz & Demain, 1977; Spencer, 2003).
Intentional mailing of B. anthracis spore-laden letters sent to the senate and media offices
in 2001 highlighted the need for data on B. anthracis to protect public health. The average
number of studies on anthrax in 2001 and 2002, following the 2001 anthrax attacks increased by
414% when compared to publications in 2000, according to a PubMed search. The increased
concern for B. anthracis spores was due to their high pathogenicity, durability, and ease of
production and dissemination in the environment.
Spores are formed from the live vegetative cells as a result of starvation and adverse
environmental conditions which are not favorable for germination and proliferation. They then
can later germinate under favorable conditions. The robustness of the spores and their ease to
spread have increased the need to understand their persistence over time. Unlike vegetative cells,
spores remain viable, resist heat, cold-treatment, light, chemicals, and antibiotics (Dragon &
Rennie, 1995; Dixon et al., 1999; Inglesby et al., 1999). Spores are reported to be viable after
decades of isolation (Graham-Smith, 1930; Graham-Smith, 1941). For example, Redmond et al.
(1998) found that samples of sugar used by the German saboteurs in 1917 during World War I
contained viable spores of B. anthracis even after nearly 80 years of archiving.

8

2.2. Bacillus anthracis spores – a Bioweapon
Bacillus anthracis has been used in bioterrorism throughout time. In American Scientist,
2001, Steven M. Block wrote that “Biological weapons have been called “the poor man’s atom
bomb.””. Several attacks and outbreak have occurred in the past in different countries leading to
social chaos and political disruption.


1917-1918: Bacillus anthracis spores laden sugar samples were used by the
German army to infect the livestock of Allied Nations, which resulted in death of
200 mules in Argentina during the World War I (Redmond et al., 1998;
Christopher et al., 1997).



1932- 1945: Bioweapon research and experiments were conducted by Japan in
Japanese-occupied Manchuria. Bacillus anthracis and other lethal pathogens
were intentionally used to infect the prisoners, resulting in at least 10,000 deaths.
During this period, over 11 Chinese cities were impacted by B. anthracis by
contaminating the municipal water supplies, food, and direct spraying into the
homes (Christopher et al., 1997).



1942: In an experiment with bioweapons during the World War II carried out by
Great Britain, a large number of bombs filled with B. anthracis spores were
detonated on Gruinard Island in Scotland. The spread of the spores were tested on
80 sheep resulting in death of all sheep due to anthrax infection. The persistent
nature of the spores made the island inhabitable for decades (Manchee et al.,
1981; Manchee et al., 1994).

9



1979: An anthrax outbreak in Sverdlovsk military microbiology facility resulted
in death of about 86% of the people exposed to B. anthracis spores in less than a
week’s time (Meselson et al., 1994).



2001: Bioterror attacks by mail containing B. anthracis spores in the United
States, killed five people and hospitalized at least 17 people. Over 10,000 people
were potentially exposed to B. anthracis in different states in the United States.
Post terror decontamination measures were highly labor intensive and expensive
(Jernigan et al., 2001).

It was estimated that an aerosol release of 100 kg of B. anthracis spores upwind of a city
size of Washington D.C. could result in nearly 3 million deaths, which is an equivalent of
hydrogen bomb explosion (Inglesby et al., 1999).
2.3. Inhalation Anthrax
The use of B. anthracis spores as a bioterrorism agent poses a threat to human health.
Exposure and infection may occur via three routes: dermal (cutaneous anthrax), ingestion
(gastrointestinal anthrax), and inhalation (inhalation anthrax). The mortality rates of inhalational
anthrax in humans are as high as 100% in the absence of any medical intervention (Bush et al.,
2001; Jernigan et al., 2001; Borio et al., 2001; Barakat et al., 2002; Chensue, 2003). Within three
months of the 2001 anthrax attacks, 22 individuals were infected with the bacteria of which
eleven resulted in inhalation anthrax. A total of five patients died due to inhalation anthrax (Day,
2003). Credible and validated exposure assessment for inhalation anthrax could aid in better
management of bioterror events in the future. Due to the presence of numerous uncertainties,
exposure assessments for B. anthracis spores have been previously conducted with large
numbers of assumptions in which decay of spores over time was not considered (Canter, 2005;
10

Hong et al., 2010). It is a major concern of the US Environmental Protection Agency and the
Department of Homeland Security to understand the behavior of such aerosolized biothreat
agents with time because consequences of an attack may be directly correlated to both the
virulence and the persistence of these spores under ambient environmental conditions.
2.4. Indoor Air and HVAC Filters
A single exposure to a large concentration of B. anthracis spores in scenarios of bioterror
attacks or accumulated exposures to lesser concentration of B. anthracis spores over time could
lead to inhalation anthrax risks. The major underlying concern with B. anthracis spores is, in
case of an intentional release, inhalational risk will be the major component of overall risk from
all possible exposure routes, when a contaminated area is reoccupied (Hong & Gurian, 2012). In
case of an indoor release, spores have a high capability to be reaerosolized and infect building
inhabitants. It was predicted using mechanistic modeling that more than 90% of aerosolized
anthrax remain aerosolized inside the building for the first 48 hours after a hypothetical indoor
release (Sextro et al., 2002). Even if decontamination procedures follow immediately after an
attack, there are high possibilities of anthropogenic activities and chaos following the attack
which could cause reaerosolization of spores. Greater than 80% of reaerosolized B. anthracis
spores under simulated office conditions were observed to be in human respirable range (Weis et
al., 2002). Once inhaled, there is a high probability of B. anthracis spores to germinate into
vegetative cells and multiply in the respiratory system (Cote et al., 2009; Day et al., 2011; Dixon
et al., 1999).
The 1979 anthrax outbreak at Sverdlovsk was found to be the result of a missing air filter
that failed to capture B. anthracis spores coming from the production line in the drying facility
(Wilkening, 2006). This indicates that high-grade air filters are proficient units to entrap spores
11

from spreading into outer environment. Nearly 18% of particles in the size range of B. anthracis
spores could be captured by the air filters in Minimum Efficiency Reporting Value (MERV) 6
category (Carrera et al., 2005; Waring &Siegel, 2008).
More than 70% of residential air handling units (AHU) are comprised of HVAC systems
in the United States (Noris et al., 2011). In cases of indoor bioterror attacks or infiltration of
biological weapons from outdoor events, HVAC filters have found to act as distribution conduits
in the building and also as potential sampling agents to determine the spread of microorganisms
in indoor environment (Van Cuyk et al., 2012; Krauter & Biermann, 2007; Noris et al., 2011;
Stanley et al., 2008; Kim et al., 2008; Calfee et al., 2014). Therefore, HVAC filters may act as
concentrating systems to classify the quality and quantity of the biological agent used in the
attack if the extraction, recovery, quantification and survival of the bioterror agents are clearly
understood over time.
2.5. HVAC Filters – Reliable Sampling Sources for Indoor Microbial Contamination
In scenarios of intentional bioterror attacks, HVAC filters could act as inexpensive
sampling sources to classify the quality and quantity of the biological agent used in the attack
over a wide area, because of their abundance.
A large-scale study of the aerosol composition from two buildings located in Minneapolis
and Seattle, respectively carried out by Stanley et al. (2008), examined loaded filter samples
from the air handing units over a period of five and three months, respectively to look for
airborne microbial contamination on filters. Comparison between the recoveries of fresh and
shipped samples of air filters found that, shipping time played a significant role in affecting
(lowering) the culturability of B. subtilis spores. Survivability of the spores was also suggested to
be dependent on shipping conditions and time. This study did not aim to recover any specific
12

microorganism, but the goal was to identify the background culturable bacterial aerosols from
large public buildings. Multiple filters of different classes composed of cotton or polyester blend,
synthetic fibers, or fiberglass media were used in each of the four AHUs in each building. The
results showed that a total of 39 different species were detected from the filters samples from
both the buildings. It was demonstrated that the most commonly found species in both the
buildings despite their demographic and seasonal diversities belonged to 11 different spore
forming species of Bacillus genus with evident indoor sources and growth promoters such as
potted plants (including the soil and insecticides), food (cooked, spoilt, improperly refrigerated).
Of the hardy spore-forming entrapped Bacillus spores, three of them (B. cereus, B. subtilis, B.
thuringiensis) belonged to the commonly used surrogate species for B. anthracis. This highlights
the significance that indoor ambience and human activities could promote the thriving of the
hardy spore forming, Bacillus population once they are captured in the AHU from an indoor or
an outdoor release (Stanley et al., 2008).
A study carried out by Moritz et al. (2001) aimed to identify the capability of HVAC
filters to capture, proliferate, and also release the bacterial species, even though the specifics of
the microorganisms used in this study were not provided in detail. The increased release of
unspecified bacterial spores was observed during cold and damp season of higher relative
humidity (>80%) in his work. This 15-month long experiment carried out on two HVAC
handling systems aimed at studying the capabilities of the air filters in HVAC units in retaining
microorganisms, and to examine if they act as a source of airborne microbial contamination.
Microbial content captured by filters were measured regularly at 14-day intervals. Bacterial
counts from one of the filter systems increased, which was attributed to the plausible
proliferation on the filters and subsequent dissemination through the filter air under increased

13

relative humidity. The released bacterial population was found to fall under the respirable size
range of the humans. It was recommended that by the use of a pre-heater in front of the HVAC
filters, the microbial growth could be annulled by controlling the relative humidity. This study
indicated that the fluctuating moisture content present in the filters over time could possibly
influence the adhesive behavior of vegetative bacteria depending on the degree of wetness;
however this needs further investigation (Moritz et al., 2001).
In general, a better understanding of the fate and transport of spores, time-dependent
recovery, and subsequent persistence pertaining to HVAC filters could serve as a constructive
tool for future risk assessment studies and clean-up strategies in indoor environments. Based on
the results of previous work described herein, it was shown that, HVAC filters are capable of
capturing, releasing, and possibly providing a substrate for proliferation of different bacterial
species. Extending the same findings towards B. anthracis spores warrants further investigation
in light of their resiliency and potential use in bioterrorism. The following sections discuss the
previously published research studies on recovery and quantification techniques of bacterial
spores from filters and porous media.
2.6. Recovery Studies
Microbial recovery studies with various types of filters, that have been performed through
seeding and cultivation experiments in order to define extraction and recovery efficiencies of
Bacillus spores on media are presented in this section.
Wang et al. (2001) observed that inoculating polycarbonate filters with B. subtilis
endospores and subsequent extraction, under varied relative humidity (rH) conditions over time
(up to 8 hours), resulted in the extraction efficiencies of the spores as high as 98 ± 1%. The
extraction efficiency was enumerated from the total microbial count in the eluted fluid via
14

epifluorescence microscopic quantification. It was found that efficient extraction techniques of
vortexing followed by ultrasonic treatment could yield high recoveries of spores from the
inoculated samples. Contrast to high recovery and extraction efficiencies measured using
epifluorescence microscopy, the relative culturability of these spores enumerated using plateculturing technique were as low as 5% and 17% for rH 30% and 85% respectively. Variables
such as sampling time, rH, and type of the sampler, significantly and independently affected the
culturability of the B. subtilis endospores. Additionally the study suggested that the desiccation
stresses in the spores could account for their decrease in culturability over time. It was suggested
that the higher culturability could be due to higher rH and higher amounts of water passing
through the unit filter area. It could be inferred from this study that variables like water content
on the filters, extraction methods, and filter types could affect the quantity and quality (viability
or culturability) of the spores released from the filters. Though the culturability was low, the
extraction was found to be as high as 100% by epifluorescence microscopy which indicates the
need for better understanding of the effect of humidity or water content on recovery and
culturability (Wang et al., 2001).
In a study by Burton et al. (2005), four types of filters of organic and inorganic origins
were loaded with B. subtilis var. niger spores whose extraction efficiencies were reported for two
extraction sequences. Extraction techniques of vortexting with ultrasonic agitation and vortexing
with mechanical shaking were compared in this study. Extraction sequence of vortexting
followed by shaker agitation was demonstrated to be the efficient technique of the two. The
samples were extracted at different time steps of 15min, 1h, and 4h. The recovery efficiencies
from extraction processes ranged between 66 -123% over time for different filters. High
extraction efficiencies that exceeded 100% were not explained in this study. Culturability was

15

measured using plate-culturing quantification, and total counts of spores extracted as well as
those present on the filters were measured using optical particle counter and epifluorescence
microscopy. There was no loss in culturability over time reported in this study in contrast to
Wang et al. (2001). The differences in the culturability observed in both the studies were
attributed to the types of filters used, and differences in experimental procedures adopted for
sporulation. It was observed that the overall culturability reduced in the case of polycarbonate
filters compared to other types of filters (Burton et al., 2005).
A recovery optimization study of B. golbigii from commonly used filters in HVAC
systems was carried out by Kim et al. (2008). This work aimed at an eluting technique in which
maximum possible culturability could be attained. The test filters were loaded with B. golbigii
spores by two methods namely, manual and aerosol loading with subsequent elution by handshaking agitation (500 shakes). The spores were eluted using deionized-filtered water and
Tween-80 solution separately which had mean recovery rates of 102% and 118%, respectively
for manual loading. A slightly greater recovery was observed when eluted with Tween-80. It was
hypothesized that the surfactant Tween-80, could reduce the clustering or clumping of the B.
golbigii spores. Spores were eluted and recovered with 100% culturability which was determined
using plate-culturing technique. Recoveries greater than 100% were observed for artificial as
well as aerosol loading when eluted with water, which were 102% and 106%, respectively.
While aerosolized tests for the vegetative bacterial cells were futile due to the fragility of the
cells. It is reasonable to assume that spores are more likely to maintain their culturability than
vegetative bacteria cells during filtration and elution. This contributes to the reliability of HVAC
filters as sampling media for spores after an attack (Kim et al., 2008).

16

A relatively long-term survival study was carried out by Mittal et al. (2011) on B.
atrophaeus spores, on HEPA filters, which are of higher efficacy for capturing smaller sizes of
bacteria compared to the HVAC filters. The test filters were loaded with the spores through
aerosolized dispersion with constant airflow through them over time, which aimed to simulate a
real-life AHU. It was observed that, all the other microorganisms survived no more than 6 days
on their filters, however the spores of B. atrophaeus lasted on the HEPA filters up to 210 days
(in absence of nutrients) with no significant loss in its viability.
Lack of validation of the extraction methods and recoveries greater than 100% led to the
recovery study of B. thuringiensis spores from HVAC filters by Solon et al. (2012), which aimed
at determining the most suited methodology for filter elution. Bacillus thuringiensis spores were
artificially loaded on to HVAC filters of MERV 7 (commonly used filter type in residential and
commercial buildings). A series of elution techniques such as sonication, shaking, and vortexing
were tested in varied combinations to understand the best order for maximum recovery. It was
found that the elution processes in the sequence of sonicating – vortexing – mechanical shaking
yielded the maximum recovery compared to other combinations of techniques. The results of this
study suggested the possibility of growth on the filters, suspecting that the filter media could act
as a substrate for the growth of the deposited spores. Growth analysis was carried out by eluting
the spores with and without the presence of filters during the extraction processes. When there
was a filter sample present for 32 minutes during the agitation sequence, it resulted in
approximately twice the recovery of bacterial spores. The authors hypothesized that it was due to
growth that an increase in spore counts was observed, however factors such as wetness of the
filter, disintegration of spores from the filters over longer periods of time when soaking,
chemistry between the spores and the filters’ surfaces were not considered (Solon et al., 2012).

17

Higher recovery percentages observed in the previous works (Wang et al., 2001; Burton et al.,
2005; Farnsworth et al., 2006) as well as in the above mentioned study create the need to
investigate growth as a factor over time with a better and consistent quantification technique.
A study carried out by Calfee et al. (2014) aimed to compare the efficiencies in recovery
between extraction-based and two vacuum-based sampling methods for B. atrophaeus spores
from HVAC filters of MERV 8. It was found that extraction-based yield were approximately
33% higher than the vacuum-based results. It was demonstrated that in the presence of dust,
there was a higher culturability of spores, though there was no statistical difference made by the
presence of dust. The recoveries were normalized to that of the reference steel coupon whereby
biases during recovery were eliminated (Calfee et al., 2014). In case of normalizing with the
initial applied concentration of aerosolized spores, the recovery was found to be as low as 4.6%.
This indicates that, in a real-life scenario of aerosolized or reaerosolized spores, it is more likely
that a major fraction of the spores remain on the surface of filter media. Interestingly, the results
of this study did not observe a recovery that exceeded the initial inoculation volume. In other
words recoveries close to or greater than 100% were not observed. Further research is needed in
this area because of contrasting recoveries observed in different studies mentioned above in
comparison to this work.
Since all of the recovery studies mentioned herein invariably carried out research using
agents other than B. anthracis spores, it is of interest to study B. anthracis directly because
agents such as B. atrophaeus, B. subtilis, and B. golbigii spores are smaller in size compared to
B. anthracis spores, therefore their transport properties in air will vary from that of B. anthracis
(Carrera et al., 2007). Surrogate agent used by Solon et al. (2012) was although morphologically
and genetically similar to B. anthracis, it lacks the recovery over time, and the benchmark to

18

compare with B. anthracis spores. There it is important to carry out recovery studies as a
function of time using B. anthracis spores.
2.7. Surrogate Selection
Due to the extensive pathogenicity and the ease of dissemination, the use of B. anthracis
spores in most research facilities is prohibited due to the requirements of minimum bio-safety
levels. The virulence of the spores mandates that experiments need to be conducted under
Biosafety Level (BSL) – 3 conditions. Experimental release or recovery studies of such spores
outside the BSL- 3 facility are forbidden. Therefore, it is required to select non-pathogenic
surrogate strains for B. anthracis, which could successfully mimic its behavior, on recovery or
on release for experimental purposes. Studies that have identified such reliable surrogates,
compared to B. anthracis in the past were considered for this work.
Bacillus species such B. anthracis, B. thuringiensis, and B. cereus belong to B. cereus
family, and are morphologically, dimensionally and genetically comparable with variations in
their pathogenicity (Greenberg et al., 2010). On the other hand, commonly used surrogates such
as B. atrophaeus, B. subtilis belong to the B. subtilis family. Due to numerous taxonomical
methods, there were reclassifications of several strains of B. subtilis and B. golbigii as B.
atrophaeus (Fritze & Pukall, 2001). Frequent references of B. atrophaeus as previously known
as B. golbigii or as B. subtilis var. niger could be encountered in some studies (Carrera et al.,
2007; Sagripanti et al., 2007; Kim et al., 2008).
Multiple closely related spores of different species of Bacillus were closely observed
using microscopy along with the virulent strain of B. anthracis in order to determine the ones
with maximum closeness to the species of interest, B. anthracis based on the size distribution by
Carrera et al. (2007). A total of 14 types of spores were assessed for morphological similarities
19

under transmission electron microscope, of which seven were different strains of B. anthracis,
and the other seven belonged to the same genus, Bacillus. It was reported that despite a similarity
in genome sequences among the seven different strains of B. anthracis, their lengths varied
significantly. The authors suggested that the best alternatives for B. anthracis are B. thuringiensis
and B. cereus based on their similarities in dimensional and aerodynamic properties. The most
commonly used surrogates B. atrophaeus or B. subtilis for B. anthracis, used in recovery studies
are different from B. anthracis spores as they are thinner and smaller (nearly half the size of B.
anthracis). Besides the commonality in the structure of the Bacillus spores, B. anthracis, B.
cereus, and B. thuringiensis possess an outermost layer called exosporium, which are lacking in
B. subtilis, B. golbigii, or B. atrophaeus (Carrera et al., 2007).
Since most of the published studies reported in the previous sections performed
inoculation through aerosolization, the aerosolized spores of B. atrophaeus or B. subtilis would
likely differ significantly from that of B. anthracis when released.
In an experimental-surrogate identification study by Greenberg et al. (2010), it was
reported that B. atrophaeus was most often used for research purposes followed by B. cereus and
B. subtilis in place of B. anthracis. The earliest reported usage of B. atrophaeus was by the US
government for biological warfare test studies in 1943. However, there was no basis or a
concrete criterion for selecting these species as surrogates. Though this study was a review and
not experimental, it was the first one of its kind to comprehensively provide a selection protocol
and identify the best suited surrogate for experimental purposes (Greenberg et al., 2010). Species
with genetic similarity to B. anthracis, morphology, pathogenicity, and response to chemical and
environmental stresses, ease of culture, cost, and history of use were several benchmarks used
for surrogate selection. It was suggested that the potential surrogates for B. anthracis could be B.

20

cereus, B. thuringiensis, B. atrophaeus, and B. megaterium. However, based on the genetic and
morphological parallels, B. thuringiensis and B. cereus were recommended to the best suited
surrogates for B. anthracis.
A comparative review by Tufts et al. (2013) on the behavior of B. thuringiensis and B.
anthracis spores when aerosolized, reported that, aerosolized B. thuringiensis and B. anthracis
are similar in behavior, however all the B. thuringiensis spores in a sample could not be
considered as a representative of B. anthracis spores. The most influencing factors for the
aerodynamic behavior of spores such as size, exosporium, morphology, hydrophobicity, and
density were compared. It was suggested that, when prepared in similar conditions, such as
sporulation media, rH, temperature, pH, sporulation time, the aerodynamic properties of B.
thuringiensis and B. anthracis appear similar. The restoration of size to aspect ratio and presence
of exosporium are important because they play a significant role in aerosolization. The
underlying statement of this review was that, though all B. thuringiensis spores cannot be
considered as B. anthracis spores owing to heterogeneity in preparation methods, if the spores
are prepared in a similar manner their heterogeneity in preparation would be minimized and
considered analogous.
Spores of B. thuringiensis were used as surrogate for B. anthracis in the recovery study
by Solon et al. (2012) from HVAC filters. Minimal differences in the genetics and morphology
between B. anthracis and B. thuringiensis, non-pathogenicity, ease to accessibility, compatibility
to laboratory usage, non-clumping property of B. thuringiensis HD1011 strain were highlighted
in this work for the selection of B. thuringiensis.
Therefore, from the above referred studies, based on the laboratory suitability,
availability, genetic and size similarities, the spores of B. thuringiensis appears to be the best

21

candidate for the surrogate usage in the research studies. However, there is no study that
compares the behavior of B. thuringiensis along the side of B. anthracis to evaluate recovery
experimentally on porous media. The study by Solon et al. (2012) could provide useful
information about the behavior of B. thuringiensis on filter media; however extrapolating it to B.
anthracis spores requires further investigation with similar experimental conditions.
2.8. Risk Assessment Studies
Since the spores of B. anthracis are highly persistent in the environment, and possess
very low ability for environmental attenuation and decay, there arises a need for active
decontamination of the site of interest in case of any deliberate release. According to the U.S.
Environmental Protection Agency (EPA), cleaning of the site is considered safe enough for
reoccupancy, when there is no longer any growth of B. anthracis spores from all the postremediation samples (Canter, 2005). This highly conservative practice is due to the absence of
confidence in the limit of detection (LOD) for the spores. LOD represents the instrumental or
environmental limit that is considered as an output during analytical methods to detect B.
anthracis. These values serve as an input variable in risk assessment studies. Estimation of the
LOD depends on recovery efficiencies with the consideration of time and experimental stages as
recovery efficacy directly affects the LOD. Improving the recovery efficiency could amplify the
sensitivity of the LOD (Herzog et al., 2009). Though recovery efficiency and LOD do not
definitively indicate the risk, they are needed to estimate exposure concentrations and thereby
interrelated and required to develop a credible and substantial risk assessment. Risk assessments
so far are based on a large number of assumptions and uncertainties as well as the absence of
time-varying recoveries of spores. Spores accumulated in air filters such as HVAC are
considered as samples that could help in inferring the exposure and risk resulting from the
22

release itself (Hong et al., 2010; Hong et al., 2012). The remarkable degree of persistence of the
spores and high mortality rates of inhalation anthrax has highlighted the importance of risk
estimation for the long-term fate and transport of the spores. Long-term modeling strongly
depends on validated short-term data that contains acceptable uncertainties and assumptions.
Therefore more studies on time-dependent recovery are needed to credibly assess the LOD, fate
and transport of the spores, exposure scenario and hence the risk.
2.9. Knowledge Gaps
Previous research leaves some important knowledge gaps that need to be addressed in
order to understand the time-dependent recovery of spores. It is vital to carry out studies with
respect to time to discretely distinguish between variability in recovery over time and the decay
of spores, particularly when subsequent extrapolations to long-term scenarios will be made.
High extraction efficiencies as observed in the past studies (Wang et al., 2001; Burton et
al., 2005; Farnsworth et al., 2006) provide information about the success of their respective
recovery strategies. However, all of these studies have used spores belonging to the B. subtilis
family, which are morphologically and dynamically different from B. anthracis (Greenberg et
al., 2010). Additionally, the spores of B. atrophaeus or B. subtilis are nearly half the size of that
of B. anthracis (Carrera et al., 2007). Due to smaller size, wet spores of B. subtilis or B.
atrophaeus become denser than B. anthracis, thereby affecting transport properties in air or
water (Greenberg et al., 2010). Use of dissimilar surrogates for recovery studies could be
considerably misleading, especially on porous media where size of the spores could play an
influencing factor on recovery because of entrapment within the media. Hence, it is critical to
employ a surrogate that is similar in relevant properties as that of pathogenic B. anthracis.

23

Though the theoretical compatibility of B. anthracis and B. thuringiensis has been
reviewed and published previously (Tufts et al., 2013; Carrera et al., 2007; Greenberg et al.,
2010; Solon et al., 2012) there are no studies to date that have experimentally demonstrated the
recovery of B. anthracis and B. thuringiensis from HVAC filters in parallel. Therefore, studies
are needed to experimentally test the behavior of B. anthracis and B. thuringiensis concurrently
to verify their theoretical compatibility.
Recovery studies that were carried out under different relative humidity (rH) conditions
and moisture levels saw an increased spore recovery and culturability (Moritz et al., 2001; Wang
et al., 2001). Greater culturability of B. subtilis spores was correlated to greater amount of water
loaded per unit filter (Wang et al., 2001). Fomite to finger transfer efficiency of B. thuringiensis
spores for porous media such as cotton and polyester also increased with increasing rH (Lopez et
al., 2013). Higher transfer efficiency was attributed to the wetness of the spores facilitating them
to transfer. The authors suggested that, a low rH could result in stronger adhesion between the B.
thuringiensis spores and the non-porous fomites resulting in lesser transfer efficiency. A similar
finding was reported in which, adhesion of B. subtilis spores increased upon drying on a nonporous surface (Nanasaki et al., 2010). Although the underlying role of moisture content over
time, upon recovery is clearly not understood, further experimental work with different moisture
levels could help address the gap pertaining to dryness and adhesion.
To summarize, the identified gaps from the previous studies are:


Lack of time-dependent recovery studies of B. anthracis spores



Lack of conclusive evidence that B. anthracis and B. thuringiensis spores have identical
or similar behavior when identical sporulation and experimental procedures are used to
measure recovery from porous media

24



Lack of quantitative studies on the effect of moisture levels on recovery from porous
media over time

25

CHAPTER 3. EVALUATING TIME-DEPENDENT VARIABILITY IN RECOVERY OF
BACILLUS SPORES FROM HVAC FILTERS

3.1. Abstract
Time-dependent recovery of B. anthraics spores is not well understood to date. The
objective of this study was to quantify and compare the recoveries of Bacillus anthracis Sterne
and Bacillus thuringiensis HD 2-61 spores from HVAC filters, a porous fomite, over a period of
168 hours under different moisture conditions. The results were analyzed using a general linear
model (GLM) with factors of sampling time, moisture content, and species. The recovered
concentrations of B. anthracis spores statistically decreased over time while B. thuringiensis
spores demonstrated an increase followed by a decrease in all trials. In the GLM, main effects as
well as the interaction effects of sampling time, water content, and the species, significantly
affected the log-reductions. Results from supplemental tests, indicated that, the increase in the
recovery of B. thuringiensis might have been due to individual or combined effects of clumping,
moisture content, and rate of dehydration. Overall, the resulting trends in recovery of B.
anthracis and B. thuringiensis spores from this study were dissimilar and indicated that B.
thuringiensis spores could not be an ideal experimental surrogate for B. anthracis spores in
recovery studies. The data generated in this study may be useful for persistence modeling and
subsequent studies in quantitative microbial risk assessment (QMRA) to estimate the risk of
exposure to B. anthracis spores in indoor air.

Keywords: B. anthracis, B. thuringiensis, HVAC, Anthrax, Bioterrorism

26

3.2. Introduction
Weaponized Bacillus anthracis spores gained importance as a public health issue due to
fear of bioterrorism related exposures and subsequent infection. Mortality rates as high as 100%
were observed in cases of inhalational anthrax, without medical intervention in humans,
compared to other forms of anthrax such as cutaneous and gastrointestinal anthrax (Bush et al.,
2001; Jernigan et al., 2001; Borio et al., 2001; Barakat et al., 2002; Chensue, 2003). Due to the
presence of numerous uncertainties about inhalational exposure to B. anthracis, risk assessments
have been based on a large number of assumptions where the fate of spores over time have not
been considered (Canter, 2005; Hong et al., 2010). Time-dependent predictions of recovery and
persistence modeling of the spread of spores in the indoor environments are important for more
credible risk assessments and decontamination strategies.
Heating, ventilation and air conditioning (HVAC) systems are widely used in more than
70% of the residential air handling units in the United States (Noris et al., 2011). In cases of
indoor bioterror attacks or infiltration of biological weapons from outdoor events, HVAC filters
have been found to act as distribution conduits in buildings and potential sampling media to
determine the spread of microorganisms (Van Cuyk et al., 2012; Krauter & Biermann, 2007;
Norris et al., 2011; Sextro et al., 2002; Stanley et al., 2008; Kim et al., 2008; Calfee et al., 2014).
Therefore, HVAC filters may act as pointers to classify the quality and quantity of the biological
agent used in the attack if the extraction, recovery, quantification and survival of the agents are
clearly understood over time.
Several studies have carried out extraction and recovery of spores of B. anthracis
surrogates from various filter media (Wang et al., 2001; Burton et al., 2005; Farnsworth et al.,
2006; Kim et al., 2008; Solon et al., 2012; Calfee et al., 2014; Moritz et al., 2001; Mittal et al.,
27

2011; Maus et al., 2000; Van Cuyk et al., 2012; Stanley et al., 2008); however none of them
have quantitatively reported the recovery efficiency as a function of time for B. anthracis spores
from filters. Instead studies were conducted with spores of B. thuringiensis, B. subtilis, B.
atrophaeus and, B. golbigii.
Furthermore, recovery studies carried out under different relative humidity conditions
found greater recovery at higher relative humidity conditions (Moritz et al., 2001; Wang et al.,
2001). Humidity was also directly correlated to culturability and inversely correlated to adhesion
of spores on porous and non-porous surfaces (Lopez et al., 2013; Nanasaki et al., 2010). Effect of
moisture on filters should to be tested because the indoor air system is constantly subjected to
fluctuating relative humidity variations in order to maintain a stable relative humidity within the
building (Woloszyn et al., 2009).
Biosafety concerns mandate the need for experimental surrogates for B. anthracis spores.
Theoretical reviews and microscopic examinations have suggested that the best suited surrogate
for B. anthracis in aerosol and in water was B. thuringiensis (Greenberg et al., 2010; Tufts et al.,
2013; Solon et al., 2012; Carrera et al., 2007). However, similarities between B. anthracis and B.
thuringiensis spores are valid if and only if the spores are prepared and experiments are carried
out identically (Tufts et al., 2013). Thus, the absence of recovery studies using both the species
warrants a need to test them under identical physical and experimental conditions over time for
comparison.
The goals of this study are to quantify the recoveries of B. anthracis Sterne strain and B.
thuringiensis HD 2-61 spores from HVAC filters over a 168-hour period with two different
moisture contents. The quantified time-based recovery of spores, will be used to address, (1) the
time-dependent behavior of Bacillus spores on HVAC filters; (2) the suitability of the surrogate

28

(B. thuringiensis spores) with respect to B. anthracis spores; and (3) the effect of moisture
content on the recovery of Bacillus spores.

29

3.3. Materials and Methods
3.3.1. Species Used
The B. anthracis Sterne strain and B. thuringiensis HD 2-61 were used for this
experiment because of their physical similarities to the virulent strain of B. anthracis. Strains of
B. anthracis Sterne and B. thuringiensis were provided by Dr. Frances Downes (Program in
Public Health, Michigan State University) and Dr. Julia Bell (Department of Large Animal
o

Clinical Sciences, Michigan State University) respectively. The strains were stored at -80 C until
initiation of the experiments. The experiments were carried out at different times in the lab to
ensure that the species were evaluated separately and the potential for cross-contamination was
eliminated.
3.3.2. Inoculum Preparation
The experiments were carried out in a biosafety level (BSL)—2 facility. Spores of each
species were separately produced from vegetative cells by suspending a loopful of each strain in
o

5mL of Tripticase Soy Broth (TSB) medium at 37 C for 24 hours. A loopful of the incubated
TSB suspension was spread onto Tripticase Soy Agar (TSA) slants. The slants were incubated at
o

37 C for 10 days. After 10 days, over 90% of the cells had formed spores. On day 11, the spore
layer was removed by washing, and resuspended in 5 mL of Type 1 ultrapure water at room
temperature. This spore solution is hereafter referred to as “master stock”. The master stocks
were quantified by plating 0.1 mL of serial dilutions onto TSA plates on the same day. Spores of
B. anthracis and B. thuringiensis were differentiated based on their colonies. Bacillus
thuringiensis spores form colonies of bigger size on TSA plates compared to B. anthracis spores.
o

Master stocks were then stored at 4 C to prevent germination of spores. The master stocks were
30

rapidly vortexed to ensure homogeneity and diluted in Type 1 ultrapure water to yield 10 mL of
6

7

desired spore concentrations in the range of 10 to 10 spores/ mL. These diluted solutions, are
hereafter referred to as “stock solutions”, and were used to spike the HVAC filters during the
inoculation process. Calculation details for preparation of master stock and stock solutions are
presented in APPENDIX-B.
3.3.3. Preparation and Inoculation of HVAC Filters
A MERV 7, Hi-E40 HVAC filter (Purolator Products Air Filtration Company,
Jeffersonville, IN) was cut into 1 x 1 inch pieces. These pieces were wrapped in aluminum foil to
o

avoid contamination by dust particles, and sterilized by autoclaving at 121 C for 15 minutes.
Three replicate filters were each spiked with 20 droplets of 20 µL, 0.4 mL stock solution,
of B. anthracis spores. A volume of 0.4 mL was used in order to saturate the entire surface of the
filter with the inoculum (Solon et al., 2012). The same inoculation procedures were used for the
stock solution of B. thuringiensis spores. To measure the effects of lower water content on
recovery, a separate stock solution was prepared with half the water used in the original stock
solution. The same inoculation procedure described above for the first stock solution was
repeated, but only 20 droplets of 10 µL, 0.2 mL, were spiked on the filters. This amount of
solution covered the surface of the filters but did not saturate them. Therefore, a lower volume of
inoculum with a similar concentration of spores was achieved. The exact spore concentrations
applied to each filter for both the 0.4 mL or 0.2 mL inoculation volumes for both species are
summarized in Table 1. Inoculated filters were gently wrapped in aluminum foil and placed
inside a sterile desiccator to ensure a dust-free and moisture-free environment. Spores were
recovered and quantified at 0 hour (30 minutes after application when all droplets had been

31

absorbed by the filter), 24, 48, and 168 hours. Thus a total of 48 filters (three filter replicates x
two concentration of spores x four time points x two species), along with eight negative controls
(one blank filter per time point per species) were tested.
3.3.4. Elution of Spores from Filters
Spores were recovered from the filters using elution sequences following the method
described by Solon (2010). At each time point, filters were carefully unwrapped and suspended
in 20 mL Type 1 ultrapure water in sterilized 50 mL centrifuge tubes. Aseptic tongs were used to
transfer the filters from the aluminum foil to the centrifuge tubes. Tongs were washed with 70%
ethanol solution, rinsed with deionized water, and dried after each transfer to avoid mixing of
spores or contamination between the filters. Elution of spores from the filters was carried out in a
three-stage process. The stages included sonication, followed by vortexing, and mechanical
shaking. During sonication, filter samples suspended in 20 mL of distilled water were placed in a
water bath inside a Branson Mechanical Ultrasonic Cleaner, Model 2510R-DTH operating at
frequency of 42 kHz ± 6% for 15 minutes. Following sonication, solutions were vortexed in a
Fischer Scientific Digital Vortex Mixer at 1000 rotations per minute for 2 minutes. The final
stage of elution was mechanical shaking in a VWR Minishaker at 170 rotations per minute for 15
minutes. Quantification was done by plate-culturing serial dilutions of the eluted solutions in
triplicates on TSA plates.

32

3.3.5. Culture-based Quantification
For all samples, a volume of 0.1 mL of the eluted solution was spread on TSA plates. The
o

plates were incubated at 37 C for about 24 hours. Plates with colony-forming units (CFUs) in the
range of 25 to 250 were considered for estimating the number of CFUs per mL of the original
eluted sample. It was assumed that each colony originated from an individual spore. The data
were converted to the spore equivalent of spores per mL, and subsequently spores per filter based
on the volume of the eluting fluid used (20 mL). At each time point, nine recovery data were
obtained. Blank TSA plates and plates with Type 1 ultrapure water were also incubated as
negative control plates to ensure contaminant-free elution and plating during quantification.
3.3.6. Tests for Clumping and Drying Rates
A set of supplemental experiments was carried out in order to test the probable clustering
of B. thuringiensis spores, using surfactant Tween-20 to separate spores in the solution. The
recovered CFUs from the surfactant experiments were used to qualitatively identify the presence
of clumping. An additional analysis to investigate the severity of the “bump” was carried out by
comparing the areas under the curve enclosed by the coordinates of the recovery points.
Dehydration rates of the water content loaded on filters were also measured separately to
understand the effect of moisture content on culturability.
It was previously observed that addition of surfactants during the elution process (to the
eluting solution) yielded higher numbers of spores during the recovery (Kim et al., 2010). In this
study, surfactant was added directly to the stock solution prior to inoculation. It was
hypothesized that clumping, if any, would occur prior to inoculation, and not during drying, on
the media. A concentration of 0.1% of Tween-20 was achieved in the surfactant-spores solution

33

(He et al., 2003; Dang et al., 2001). Inoculation volume of 0.4 mL, with a concentration of
6

approximately 2.6 x 10 spores per mL, followed by elution and quantification procedures as
discussed above were repeated for B. thuringiensis samples with and without the addition of
Tween-20. The B. thuringiensis spores, with and without the addition of Tween-20 were also
observed with a scanning electron microscope (SEM).
To observe the drying rates of the filters over 168 hours for each water content, 0.4 mL
and 0.2 mL of Type 1 ultrapure water was spiked on to separate filter coupons in triplicates. The
filters were wrapped and stored in desiccator at identical physical conditions. Filters were
weighed over 168 hours with time intervals of 8 hours and 16 hours.
3.3.7. Scanning Electron Microscopy (SEM)
SEM was used to observe the tangible effects of autoclaving on HVAC filters, and to
observe the nature of B. thuringiensis spores, in the presence and absence of Tween-20. Pieces of
HVAC filter samples were examined in a Joel JSM-6610LV scanning electron microscope
(JOEL Ltd., Tokyo, Japan). Samples of B. thuringiensis spores with and without the presence of
Tween-20 were processed and fixed on aluminum stubs, according to SEM procedure described
in APPENDIX-D. Fixed samples were examined in a Joel JSM-7500F (cold field emission
electron emitter) scanning electron microscope (JOEL Ltd., Tokyo, Japan).

34

3.3.8. Data Treatment
The total plate counts from each triplicate plate were converted to spores recovered/ filter
by multiplying by the inverse of the volume of extract plated (0.1 mL), by the dilution factor of
the corresponding plate, and the volume of total extract used to elute the filter (20 mL). The
calculated concentrations of recovered spores were represented in terms of spores per filter for
easier interpretation. To normalize the recovery data, all the recoveries were divided by the
number of applied spores (Table 1). These data were subsequently log10 transformed and
averaged for each time point (N=9). This is hereafter referred to as “log-reduction”, which was
calculated using Eq. (1). In order to show the differences in the results produced by the choice of
denominator, the recoveries were divided by the average recovery obtained at the 0 hour, and
log10 transformed. By normalizing with 0 hour recovery, a 100% initial recovery rate was
assumed. This is hereafter referred to as “normalized log-reduction”, which was calculated using
Eq. (2). Recovery rates and normalized recovery rates were computed using Eq. (3) and Eq. (4)
respectively, to represent the recoveries in terms of percentages.
Nt

Log Reduction = log10 (Ni )

…. (1)
Nt

Normalized Log Reduction = log10 (N0)
Recovery Rate =

Nt
Ni

…. (2)

∗ 100%

Normalized Recovery Rate =

…. (3)
Nt
N0

∗ 100%

…. (4)

where Nt=Average number of spores recovered at time, t, Ni=Initial number of spores applied to
the filters, N0=Average number of spores recovered at time 0 hour.

35

3.3.9. Statistical Tests
Statistical analyses were performed using the general linear model (GLM) Univariate
procedure of SPSS version 20 (SPSS Inc., Chicago, IL) on data sets containing: (i) the dependent
variable log-reductions, at all four time steps (0, 24, 48, and 168 hours) across species (B.
anthracis and B. thuringiensis), and moisture contents (0.4 mL and 0.2 mL), (N=144); (ii) logreductions of B. thuringiensis spores with and without the addition of Tween-20 (N = 72) to
assess for significant differences due to the addition of Tween-20 over time. The analyses
included both main effects as well as all interaction effects on the response variable.
One-tailed t-tests were performed to assess for significant reduction on log-reductions
between two time steps for both the species separately. A significance level (α) of 0.05 was used
for all tests.

36

3.4. Results
Results obtained in this study are presented in the following order: (i) comparison of data
treatment techniques; (ii) evaluation of log-reductions and normalized log-reductions; (iii)
comparison of B. anthracis and B. thuringiensis spores; (iv) evaluation of the effect of
surfactants on clumping of B. thuringiensis spores; and (v) comparison of log-reduction between
different moisture levels.
3.4.1. Concentrations of applied and 0 hour recovered spores
Mean concentrations obtained from the 0 hour extractions and enumeration ranged from
6

5

2.27 ± 0.167 x 10 CFU to 7.00 ± 1.34 x 10 CFU per filter. Applied concentration and average 0
hour recovered total concentrations obtained for B. anthracis and B. thuringiensis spores at both
moisture levels are summarized in Table 1. The calculated recoveries obtained during 0 hour
were used to normalize the recoveries measured over time using Eq. (2) and (4). Applied
concentrations of spores (Table 1) were used in Eq. (1) and (3), to obtain log ratios. The 0 hour
recovery of B. anthracis spores was close to the applied concentration but B. thuringiensis spores
lowered by an order of magnitude or 1-log10.
3.4.2. Comparison of recovery rates and normalized recovery rates
Table 2 presents the summary of t recovery rates and normalized recovery rates of B.
anthracis and B. thuringiensis spores. Recovery rates of B. anthracis were statistically
decreasing (p = 0.008) over time. For B. thuringiensis, recovery rates statistically increased (p <
0.001) greater than 100% at 24 hours.
As expected, normalized recovery rates for B. anthracis and B. thuringiensis at the 0 hour
were the same, due to the assumption of 100% initial recovery. Because the B. anthracis

37

recovery at 0 hour was close to applied concentrations, the differences between the 0 hour
recovery rates and normalized recovery rates were insignificant. However, the differences for B.
thuringiensis were significantly different (p<0.001).
Due to this variation both the values were retained for comparing their respective logreduction values.
By 168 hours, recovery for B. anthracis declined to 10% for 0.4 mL and 34% for 0.2 mL
and recovery for B. thuringiensis declined to 38% for 0.4 mL and 56% for 0.2 mL. For both
water levels, B. thuringiensis increased at 24 hours sampling followed by a decrease. This is
discussed in the next section.

38

Table 1. Summary of applied and average 0 hour recovered concentrations of Bacillus spores
Species
Volume of inoculum/ filter Total Applied Concentration a Total 0 hour
(mL)
Recovered
(CFU)
a, b, c
Concentration
(CFU ± SD)
6
6
B. anthracis
0.4
1.15 x 10
1.13 ± 0.203 x 10
6
5
B. thuringiensis 0.4
1.09 x 10
7.00 ± 1.34 x 10
6

B. anthracis

0.2

2.29 x 10

B. thuringiensis

0.2

2.19 x 10

6

2.27 ± 0.167 x 10

6

5

8.47 ± 2.15 x 10

a, Total concentration of CFUs per filter
b, Average of three TSA plate counts for each of the three filter replicates. Plates were
o
incubated at 37 C for 24 hours
c, N=9

Table 2. Summary of recovery rates and normalized recovery rates over time (N=9)
Recovery Rate (%)
Normalized Recovery Rate (%)
Species
(volume/ filter, mL)
0h
24 h
48 h
168 h
0h
4h
8h
B. anthracis (0.4)
B. thuringiensis (0.4)
B. anthracis (0.2)
B. thuringiensis (0.2)

98 ± 18
64 ± 12
99 ± 7
39 ± 10

78 ± 18
116 ± 31
77 ± 9
55 ± 7

52 ± 13
79 ± 19
54 ± 9
59 ± 12

10 ± 3
38 ± 2
34 ± 5
56 ± 10

39

100 ± 18
100 ± 19
100 ± 7
100 ± 25

79 ± 18
180 ± 48
78 ± 9
143 ± 18

53 ± 13
123 ± 30
54 ± 9
153 ± 31

168 h
10 ± 3
58 ± 3
35 ± 5
146 ± 26

3.4.3. Comparison of B. anthracis and B. thuringiensis spores
Figures 1 and 2 illustrate the comparison of the log-reductions for B. anthracis and B.
thuringiensis over time, at the two moisture levels respectively. Bacillus anthracis and B.
thuringiensis behaved dissimilarly over time. Recoveries of B. anthracis were gradually
reducing, while recoveries of B. thuringiensis increased followed by a decrease over time. At the
higher water content (0.4 mL) there was a slight overlap between the error bars of B. anthracis
and B. thuringiensis reductions (Fig.1) up to the 48 hours period which was not observed at the
lower water content (0.2 mL) in Fig. 2.
Figures 3 and 4 illustrate the comparison of the normalized log-reductions for B.
anthracis and B. thuringiensis over time, for two moisture levels respectively. With the
assumption of 100% recovery, the data indicates that B. anthracis is reducing over time, while B.
thuringiensis increased at 24 hours over 100%, then decreased as their corresponding recoveries
lie on either sides of the X-axis. This was observed for both moisture levels.
Overall, as a function of time B. anthracis reduced with respect to applied concentration
of the inoculum while, B. thuringiensis consistently increased at 24 hours, which appears like a
“bump” in all figures.

40

Figure 1. Log-reduction from Eq. (1) as a function of time for 0.4 mL volume of inoculation for
both B. anthracis and B. thuringiensis. Error bars indicate standard deviation from the mean.

Figure 2. Log-reduction from Eq. (1) as a function of time for 0.2 mL volume of inoculation for
both B. anthracis and B. thuringiensis. Error bars indicate standard deviation from the mean.

41

Figure 3. Normalized log-reduction from Eq. (2) as a function of time for 0.4 mL volume of
inoculation for both B. anthracis and B. thuringiensis. Error bars indicate standard deviation
from the mean.

Figure 4. Normalized log-reduction from Eq. (2) as a function of time for 0.2 mL volume of
inoculation for both B. anthracis and B. thuringiensis. Error bars indicate standard deviation
from the mean.

42

3.4.4. Effect of surfactants on clumping of B. thuringiensis spores
The apparent bump in B. thuringiensis was thought to be due to clumping. Results from
additional experiments with and without the addition of surfactant Tween-20 to test for clumping
of B. thuringiensis spores revealed that there was an increase followed by a decrease in the logreduction trend even in the presence of surfactants. The “bump” was observed in the logreductions with and without Tween-20 (Fig. 5). A linear trend line was fit to evaluate the severity
of the bump in each case. On comparing their error sum of squares (SSE) of their average plate
count results, it was found that the recoveries in the presence of Tween-20 (SSE= 0.0002013)
deviated less from the linear trend than in the absence of Tween-20 (SSE=0.00109). In other
words, recoveries were more consistent in the presence of Tween-20.

Figure 5. Log-reduction from Eq. (1) as a function of time for 0.4 mL volume of inoculation of
B. thuringiensis spores with and without addition of Tween-20. Error bars indicate standard
deviation from the mean.
43

The area enclosed by the coordinates of recovery over time for samples with Tween-20
was 1.472 square units, and that of coordinates of recovery over time for samples without
Tween-20 was 2.610 square units. Larger area enclosed by the samples without Tween-20
indicates that the bump is more severe and pronounced compared to the samples with Tween-20.
Also the standard deviation of the entire sample (N=36) with Tween-20 was 0.011, and that for
the sample (N=36) without Tween-20 was 0.024. Greater standard error about the mean also
indicates the larger spread in recovery values which may be due to the presence of clumps
(Ilstrup, 1990).
3.4.5. Comparison of log reductions between different moisture levels
Figures 6 and 7 represent the log-reductions of the two species at two moisture levels as a
function of time. Log-reductions of B. anthracis at both moisture levels for the first 48 hours
practically coincided with each other (Fig. 6). The log-reductions for 0.4 mL samples were
statistically lower than those of the 0.2mL at the 168 hours. Interestingly in Figure 7, the
behavior of B. thuringiensis spores was different in comparison to that of B. anthracis spores.
There was no concurrence between the data points of two moisture levels for B. thuringiensis.
The log-reductions for 0.2 mL samples were statistically lower than that of 0.4 mL except for
168 hours, which is opposite of the observations for B. anthracis.
Figures 8 and 9, show the normalized log-reductions at the two moisture levels as a
function of time according to Eq. (2), for B. anthracis and B. thuringiensis respectively. The
results from the two moisture levels for the first 48 hours were similar for both the species. The
normalized log-reductions till the 48 hours were not significantly different from one another for
either species. Therefore, from the normalized log-reductions, the change in moisture levels does

44

not appear to cause a significant difference in recovery for about the first 50 hours of the
experiment.
Overall, less water content loaded per filter did not affect the measured recovery for 48
hours for B. anthracis spores (Fig. 6, 8); less water content loaded per filter resulted in lesser
culturability of B. thuringiensis spores (Fig. 7); less water content resulted in higher culturability
for wither species in the 168 hours measured recovery (Fig. 6, 7).

45

Figure 6. Log-reduction from Eq. (1) as a function of time for B. anthracis spores for both
volumes of inoculation. Error bars indicate standard deviation from the mean.

Figure 7. Log-reduction from Eq. (1) as a function of time for B. thuringiensis spores for both
volumes of inoculation. Error bars indicate standard deviation from the mean.

46

Figure 8. Normalized log-reduction from Eq. (2) as a function of time for B. anthracis spores for
both volumes of inoculation. Error bars indicate standard deviation from the mean.

Figure 9. Normalized log-reduction from Eq. (2) as a function of time for B. thuringiensis spores
for both volumes of inoculation. Error bars indicate standard deviation from the mean.

47

3.4.6. Comparison of Drying Trends of Two Water Contents
The rates of dehydration were tested using sterile water on HVAC filters with identical
experimental conditions. HVAC filter coupons loaded with 0.4 mL and 0.2 mL of water weighed
over time showed that, for approximately first 50 hours, the rates of drying of 0.4 mL and 0.2 mL
filter samples were identical. The drying data (APPENDIX-C) indicated that both 0.4 mL and
0.2 mL samples dehydrated similarly until the 50 hours (m0.4 = -3.938; m0.2 = -3.704). After
that, the 0.4 mL samples had an accelerated dehydration compared to 0.2 mL filters until the end
of 100 hours (m0.4 = -4.119; m0.2 = -0.0563). Although the 0.2 mL samples were dry for a
longer duration (approximately 50 hours more) than the 0.4 mL sample, the latter filter had an
accelerated rate of dehydration between 50 –100 hours period (Fig. C. 1) which may have
influenced the measured recovery by reducing the culturability of spores.
3.4.7. Microscopy
Porous media offers more space for attachment of spores and are likely to entrap them
over time leading to lower recovery. HVAC filters were evaluated using SEM before and after
autoclaving to check for any differences caused by the autoclaving process. Moistened HVAC
filter coupons identical to the ones used in this study were observed using light microscope for
tangible variations caused during drying over time. The results from SEM as well as light
microscopy indicated that there were no discernable physical changes of the HVAC filters due to
autoclaving or drying respectively.
SEM was also used to evaluate clumping. Images observed using SEM of B.
thuringiensis spores with and without Tween-20 indicated that there was a reasonable amount of
clumping or clustering present in both the samples (Fig. 10(a) & 10(b)). However, the nature of

48

clumps varied remarkably in both samples (Fig. 10(c) & 10(d)). In presence of Tween-20 a
relatively lower number of clumps were observed of smaller size (Fig. 10(e)). Whereas, in the
absence of Tween-20, clumps were dense, thickly layered, and strongly coiled to one another
(Fig. 10(f)). Clumping of spores was thus confirmed to exist at a higher degree in the absence of
the surfactant. Though a distinct conclusion could not be drawn between the presence and
absence of surfactant on clumping, it could be evidently stated that the degree of agglomeration
or clustering was greater in the absence of surfactants. This supports the results presented in
Section 3.4.4 which used CFUs to evaluate clumping.

49

A

B
A

B

C

D

E

F
E

F

Figure 10. SEM images of B. thuringiensis spores (A, C, E) with addition of Tween-20 and (B,
D, F) without the addition of Tween-20 (More images using SEM are presented in APPENDIX
D)
50

3.4.8. Statistical Analyses:
Results of Univariate GLM analyses showed that the independent variables had a
significant effect on the response variable (log-reduction). Main effects and interaction effects of
all three independent variables, species (B. anthracis and B. thuringiensis), sampling time (0, 24,
48, and 168 hours), and moisture contents (0.4 mL and 0.2 mL) were highly significant (p <
2

0.001), with R value of 0.826. The analysis of the entire data set provided statistical evidence of
the significance between the log-reduction of both the species. With the inclusion of the
interaction with time this significance increases, indicating that time plays a major role in the
recovery of Bacillus spores.
There was no significant difference in log-reduction of B. thuringiensis due to addition of
Tween-20 (p = 0.651). However there was a high significant difference on including the
interaction between time and surfactant (p < 0.001), once again indicating that time strongly
influences the recovery. Results from one-tailed t-tests showed that there was an overall
statistical decrease in the log-reduction of Bacillus spores regardless of the species over time (p <
0.001). Results from statistical analyses are presented in APPENDIX-E.

51

3.5. Discussion
3.5.1. Comparison of Normalization Factors
Recovery studies in the past have used multiple normalization factors to calculate their
recovery rates. Commonly used denominators in survival, inactivation, or persistence studies are
the 0 hour recovery, because total duration of such tests are over long periods such as 1000
hours, 60 days, and 40 days (Masago et al., 2008; Abad et al., 1994; Tripp 1960). Some, timeindependent recovery studies used the recovery from non-porous control surfaces in order to
eliminate the losses or biases involved during actual sampling of porous surfaces (Calfee et al.,
2013; Calfee et al., 2014; Edmonds et al., 2009). However, most recovery rates are computed by
normalizing with respect to the applied concentration of spores for short-term studies (Kim et al.,
2008; Krauter et al., 2005; Buttner et al., 2004; Wang et al., 2001; Burton et al., 2005). The
differences in results and subsequent inferences on using different normalization techniques
especially for surrogate species are highly significant (Murali & Mitchell, 2014). In the
normalized recovery rates (Table 2), the projected increase in the case of B. thuringiensis when
compared to B. anthracis, appears as if B. anthracis spores were lower when compared to B.
thuringiensis at all time points. It is possible that such normalization might lead to an inference
that B. thuringiensis was increasing or growing above the actual applied inoculum concentration.
The recovery rates (Table 2) show that B. anthracis had 0 hour recovery close to 100% because
the applied concentration and the 0 hour recovery for B. anthracis did not vary significantly. On
the other hand, B. thuringiensis had greater than 70% loss (0.2 mL treatment) at the 0 hour
recovery. This high initial decrease in recoverability of B. thuringiensis spores was morphed in
the normalized recovery rates leading to a projected apparent rise. Hence, it is acceptable to use 0
hour recoveries or recoveries from reference coupons for normalization for a short-term study as
52

long as it does not differ from the applied concentration statistically. Referencing to a 100%
initial recovery for persistence studies might lead to incorrect inferences about the trend of
decay. Since this is a recovery study, hereafter the discussion will be pertained only to recovery
rates or log-reductions based on applied concentrations.
3.5.2. Recovery of Bacillus spores over Time
Table 2, Figures 1 and 2 show the recovery rates and log-reductions over time for all the
treatments respectively. Bacillus anthracis was found to gradually decrease over time while B.
thuringiensis was found to consistently increase followed by a decrease with the appearance of a
“bump” in all trials. Inherent time-dependent reduction was anticipated for the spores of both the
species (Kim et al., 2008; Stanley et al., 2008). An increase after the 24 hours recovery for B.
thuringiensis spores was found to be similar to a previously published extraction study, in which
an increase in recovery of B. thuringiensis spores, after 32 minutes of extraction from inoculated
autoclaved HVAC filters was reported, though the applied or increased percentages were
unreported (Solon et al., 2012). It was suggested that the filter media by itself could act as a
substrate for microbial proliferation.
Besides the above mentioned work, studies of B. thuringiensis on filters or porous media
over time are limited for comparison purposes. Nevertheless, there are studies in which B.
thuringiensis spores were tested for survival and germination in natural and sterile (autoclaved)
soil samples. Considering the lack of filter-spores studies, studies on sterile soil samples with
similar physical conditions are the closest to compare time-dependent recovery. A similar timedelayed “bump” was observed in B. thuringiensis spores -- on sterile Kohnan and Arakawa soil
samples followed by a decrease (Akiba, 1986); and on Varna soil samples amended with proteins
after 24 hours (Saleh et al., 1970). It was concluded that B. thuringiensis spores were capable of
53

germinating in soil; however the subsequent decreases were unexplained. These results have
suggested that the germination and proliferation of B. thuringiensis spores even in sterile soil is
likely.
Considering the observations from these past studies, the possible reasons for the increase
in B. thuringiensis could be due to germination of the spores over time and/or proliferation on
the filter media. However lack of nutrients, sterility and safety levels during the experiments
warrant delving further into other possible intricate factors such as (1) clumping; (2) adhesion
leading to aggregated detachment from the filters followed by disintegration of clumps during
mechanical extraction, and (3) varying moisture content on the filters over time. Clumping and
moisture were tested and addressed in this study.
3.5.2.1. Clumping in B. thuringiensis spores
Bacillus thuringiensis spores have the tendency to clump (Dulmage et al., 1969).
Presence of peripheral appendages and exosporium as in case of B. thuringiensis spores, have
been found to aid their capability to adhere to solid surfaces (Tufts et al., 2013; Husmark, 1993).
Bacillus thuringiensis spores are stronger adherent to the surface and to themselves than B.
anthracis spores due to the presence of hairy naps on their outermost layer (Tufts et al., 2013).
Our results were in close relation to the results of Sugimoto et al. (1996), in which a “bump” was
observed in the recovery of dried B. stearothermophilus spores (morphologically similar to B.
thuringiensis spores) from aluminum strips. It was reported that the external appendages on B.
stearothermophilus spores caused an enhanced adhesion on the aluminum surface leading to
lower recovery. The inconsistency in the recovery was unexplained. Therefore, it was suspected
that the combined effects of adhesion to the filters and clumping of B. thuringiensis might have
led to disintegration of B. thuringiensis spores in clusters which decreased over time as the
54

moisture content decreased due to drying. Test for clumping revealed that irrespective of the
surfactant Tween-20, a “bump” was observed in log-reduction of B. thuringiensis spores (Fig. 5).
However, the SSE values of the mean plate counts showed that recoveries were more consistent
in the presence of Tween-20. This could mean that clumping was not as pronounced as it was,
prior to addition of Tween-20. Lesser clumping might have led to a more stable recovery
resulting in a lower SSE when a linear model was used to evaluate the severity of the bump.
Moreover, visual evidence of SEM imaging showed that there was a greater degree of clumping
or clustering of B. thuringiensis spores in the absence of surfactants. Microscopic images of two
samples were not conclusive enough to determine if clumping caused the significant increase in
the recovery of B. thuringiensis spores. Although addition of Tween-20 did not completely
eliminate clumping, it reduced clumping significantly by discretizing a large numbers of spores
leading to a more consistent recovery. Hence it is could be said that, in presence of Tween-20,
recoveries were more consistent, but clumping or aggregation was observed in both the samples
using SEM, it is concluded that clumping could not be solely accountable for significant rise in
the recovery of B. thuringiensis spores.

55

3.5.2.2. Comparison of drying rates of different moisture contents
A past recovery study of B. subtilis spores from polycarbonate filters reported an
increased spore recovery and culturability under higher relative humidity (85%); this greater
culturability of B. subtilis spores was attributed to greater amount of water loaded per unit filter
(Wang et al., 2001).
Transfer efficiency of B. thuringiensis spores from porous and non-porous media such as
cotton, polyester, and stainless steel to fingers, increased with increasing relative humidity. It
was directly correlated that wetness of the spores facilitated better transfer. In other words, a low
relative humidity could result in stronger adhesion between the B. thuringiensis spores and
fomites resulting in lesser transfer efficiency (Lopez et al., 2013).
Adhesion of B. subtilis spores increased upon drying of the inoculum on sterilized
stainless steel surface (Nanasaki et al., 2010). Although the underlying role of moisture content
over time, upon recovery is clearly not understood, it is possible that varied moisture levels could
directly or indirectly correlate to recovery of spores by affecting the spore drying and adhesion
over time. Based on previously published results on recovery of spores from porous and nonporous surfaces, the adhesion of spores on the surface of the filter was expected to strengthen
upon drying resulting in a lower recovery. The hypothesis was that, decreasing moisture or water
content loaded per filter would result in lower recovery of spores over time. However, the
observed results did not support this hypothesis completely. Log-reductions (Fig. 6 and 7) for the
first 48 hours from the reduced moisture samples (0.2 mL) were statistically lower than that of
higher moisture samples (0.4 mL) as expected for B. thuringiensis; but for B. anthracis spores,
there were no statistical differences in the log-reduction. Also, in contrast to the hypothesis,
filters with lower moisture (0.2 mL) had higher recovery in the 168 hours for B. anthracis and B.
56

thuringiensis spores (Fig. 6 & 7) which leads to the possibility of changes in viability or
culturability.
Ishihara et al. (1994) found a strong linear correlation between the percentage of viability
and percentage of free water content present in the B. subtilis spores, when desiccated for up to
24 hours. After 24 hours, it was not feasible to draw a correlation because the viability had
reached, and stayed at 0% till the end of the experiment (Ishihara et al., 1994). This strong
correlation could explain the influence of free water content to the viability of spores in this
study.
Although the true mechanism of the contribution of moisture content to the recovery is
unclear, since both the filters were in equal moisture content at the 168 hours, it is suggested that,
the rate of dehydration may have affected the recovery rather than the degree of dehydration.
Based on the dehydration results, filter samples with greater moisture content (0.4 mL)
dehydrated at an accelerated pace between 50 hours- 168 hours. In the light of the drying rates
th

and the culturable counts obtained for B. anthracis and B. thuringiensis at the 168 hour, it is
hypothesized that the culturability of spores may be linearly correlated to the rate of dehydration
rather than to the state of being dry. This results in decline in measured recovery on speedier
dehydration rates.

57

3.5.3. Surrogate Suitability
From our results, we conclude that B. thuringiensis and B. anthracis spores behaved
dissimilarly over time on HVAC filters. Gradual and significant reduction in B. anthracis spores
over time was not reflected in the behavior of B. thuringiensis spores. Effect of moisture and rate
of dehydration on adhesion phenomena, presence of appendages, and clumping could account for
this variability in the recovery of B. thuringiensis spores. Although similarities were established
in the past reviews between the surrogate and B. anthracis spores, (Greenberg et al., 2010; Tufts
et al., 2013) they were not tested experimentally to date on filter media. Bishop (2014) reported
that B. anthracis Sterne spores and B. thuringiensis spores behave differently in response to the
germinants in non-sterile soil microcosms. Germinated vegetative forms of B. anthracis strains
were not able to persist in soil samples while B. thuringiensis spores remained persistent on
germination (Bishop, 2014). From the results observed in this study, it is concluded that B.
thuringiensis spores are not a suitable surrogate for B. anthracis spores on porous media (Fig. 1
and 2).

58

3.5.4. Summary
Based on the experimental conditions in which the tests were performed and the observed
results of this study, the following conclusions are made for the time-dependent recovery of B.
anthracis and B. thuringiensis spores from HVAC filters:
(i)

Time had a significant effect on recovery regardless of the species used. The results
of culturable counts indicate that there was an eventual statistical reduction in
recovery after 168 hours in both the species when compared to the original applied
concentrations of spores.

(ii)

Bacillus anthracis Sterne and B. thuringiensis spores had dissimilar time-dependent
behavior on HVAC filters. Bacillus anthracis spores consistently and statistically
decreased with time whereas B. thuringiensis spores constantly had a ‘bump’ at the
24 hours or 48 hours recovery in all trials. Hence, irrespective of time, from the
experimental conditions and observed results B. thuringiensis spores could not be an
ideal experimental surrogate for B. anthracis spores. It is suggested that B. anthracis
Sterne strain could be the best suitable surrogate for pathogenic B. anthracis.

(iii)

Variation in moisture did not affect the recovery of B. anthracis significantly for the
first 48 hours, and for B. thuringiensis, lower moisture content resulted in lower logreduction until 48 hours. For both the species the log-reductions at the 168 hours
reversed the expected trend for lower water content samples. Based on the results of
drying tests it is suggested that the rate of dehydration could influence the recovery of
spores than the degree of dehydration.

59

(iv)

From the evidences of all tests the consistent ‘bump’ observed in B. thuringiensis
spores might be due to individual or combined effects of clumping, moisture content,
and rate of dehydration.

60

3.6. Concluding Remarks
Our study quantified the time-depended recovery of B. anthracis Sterne and B.
thuringiensis HD 2-61 spores from HVAC filters. Recoveries over different moisture content
were also evaluated for the effect of water content on recovery over time. Bacillus thuringiensis
spores constantly had a ‘bump’ at the 24 hours or 48 hours recovery in all trials, and B. anthracis
statistically reduced over time. The abnormal increase in the recovery of B. thuringiensis could
be due to individual or combined effects of clumping, moisture content, and rate of dehydration.
The results in this study are based on culturable counts. No experiments were carried out
attempting to estimate the non-culturable viable spores in the sample. Regardless of the species,
over time, eventually there was a significant decrease in the log-reduction. Results from
preliminary experiments using real-time PCR (qPCR) quantification on recovery of B. anthracis
Sterne spores using sporulation, inoculation, and extraction techniques identical to this study
indicated that there was a significant difference between the culture-based and molecular-based
recovery over time (APPENDIX-F). Since, qPCR enumerates both live as well as dead counts in
the sample this difference could be attributed to spores: that are viable yet not culturable, in
dormant stage, or large amounts of dead spores. Considering this high and significant difference
between the two quantification methods over time as well as the remarkable durability of
viability in spores over decades, it would be worthwhile to explore the true viable counts over
time for persistence and risk assessments.
Our study was carried out on new, sterile filters, extending the recovery study using the
presence of dust particles and commonly found debris in indoor filters are encouraged for a more
realistic setup of indoors. Our results are expected to be successfully replicable under identical

61

sporulation and experimental procedures, and suggest that experiments with aerosolized
moisture-free application are warranted to test the aerosol surrogacy of B. thuringiensis spores.
3.7. Acknowledgements
We thank Dr. Evangelyn Alocilja and Dr. Joan Rose for useful discussion. We thank
Andrew Bruce and Carol Flegler for assistance in data collection.

62

CHAPTER 4: DISCUSSION AND CONCLUSIONS
The final chapter of this thesis is devoted to discussion of the three research hypotheses
and main conclusions, followed by pertinent suggestions for future work.
Three research hypotheses were formulated for this study. For all three hypotheses, the
entity of interest was the amount of Bacillus spores recovered per filter or the log- reduction, as
per the context. In the following sections of this chapter, the main findings are reviewed for each
hypothesis followed by corresponding discussion.
4.1. Hypothesis I
Bacillus spores will have a time-dependent reduction in recovery from HVAC filters
This hypothesis was fully supported for both the species at both the moisture levels
tested. According to the observed culture-based results, there was an overall time-dependent
reduction in both the species when compared to their respective inoculated concentrations
(APPENDIX-B). The reduction trends observed in this study were comparable to that observed
in Kim et al. (2008). The relationship between the log-reduction of B. anthracis spores and time
was statistically decreasing over time while in contrast, a consistently increasing yet decreasing
trend was observed for B. thuringiensis spores over time. The results indicate that, B. anthracis
spores had a greater reduction over time when compared to B. thuringiensis spores, signifying
that there would be a larger fraction of B. anthracis spores that was capable of persisting on the
filter surface over time.
It is important to regard the fact that, these results are a reflection of culture-based data
and not true viability of spores. Spores are known for maintaining their viability and persistence
in environmental matrices over decades. Therefore, it is warranted that studies in the future,
account for spores over time beyond their culturability, highlighting their true viability.
63

4.2. Hypothesis II
Bacillus anthracis and B. thuringiensis spores will have identical time-dependent
recovery from HVAC filters
This hypothesis was not supported by this study for the variables, time and moisture
content. Regardless of the overall reduction observed in both the species, the reduction trends of
B. anthracis and B. thuringiensis spores differed remarkably over time. Bacillus anthracis spores
showed a consistently declining curve over time, whereas in case of B. thuringiensis the
reduction curves of all trials were consistently increasing followed by a decrease leading to an
appearance of a ‘bump’ like structure.
There has been only a handful of Bacillus spores’ recovery work carried out to date,
among which only one (to our knowledge) was using B. thuringiensis spores from HVAC filters
(Solon et al., 2012). The authors observed a similar increase after 32 minutes of extraction from
inoculated sterile HVAC filters, but the applied or increased percentages were unreported.
Therefore it could be said that, a similar recovery pattern was previously observed yet, it is hard
to extrapolate over time because of lack of time-dependent studies.
Similar abnormal increase was observed for B. thuringiensis spores in the past when
recovered from sterile soil matrices under comparable physical experimental and incubation
conditions. Saleh et al. (1970) and Akiba (1986) observed an increase followed by a decrease in
B. thuringiensis spores, recovered from Varna soil samples and Kohnan, Arakawa soil samples
respectively after approximately 24 hours or 48 hours.
Based on above mentioned studies, the suggested possible reasons for the increase in B.
thuringiensis could be due to germination of the spores over time, or proliferation of spores on
the filter media. But germination alone could not have accounted for an increase in B.

64

thuringiensis spores. Germination ought to be followed by proliferation in order to produce such
a bump. However lack of nutrients, sterility and safety levels during the experiments warrant
delving further into other possible intricate factors such as clumping and varying adhesion
leading to aggregated detachment from the filters followed by disintegration of clumps during
mechanical extraction, varying moisture content on the filters over time.
It has to be noted that clumping of B. thuringiensis spores has been reported since 1969
(Dulmage et al., 1969). Besides clumping, there were some published works in the past
pertaining to the variation in morphology between B. anthracis and B. thuringiensis spores – it
was suggested that the presence of hairy filamentous appendages on the periphery of B.
thuringiensis spores enhance their capability to adhere on surfaces (Tufts et al., 2013; Husmark,
1993).
It is interesting to note that in a 4 week study (Sugimoto et al., 1996) using B.
stearothermophilus (morphologically comparable to B. thuringiensis) and B. subtilis (does not
possess appendages or exosporium), dried and recovered from aluminum strips, B. subtilis
showed a consistent decline whereas B. stearothermophilus had an increase followed by a
decrease. Presence of hairy filamentatious structures (appendages) on the periphery of Bacillus
spores have to date, been positively correlated to the adhesion on solid surfaces (Tufts et al.,
2013; Husmark, 1993; Ronner et al., 1990; Sugimoto et al., 1996; Husmark & Ronner, 1990). It
is therefore convincing to suggest that, presence of appendages and diminishing moisture over
time might be instrumental in the increased recovery of B. thuringiensis spores.
Although addition of an emulsifier such as Tween-20 to the spore solution minimized
clumping to a great extent when compared to the samples in absence of Tween-20, there was
evidences of clumped or clustered spores in both the samples observed using SEM. Hence it is

65

could be said that, in presence of Tween-20, recoveries were consistent, however as clumping or
aggregation was observed in both the samples using SEM, it is concluded that clumping could
not be solely accountable for significant rise in the recovery of B. thuringiensis spores.
Therefore, significant increase in the reduction of B. thuringiensis spores might not have resulted
solely due to clumping; however clumping may have been one of the contributory factors.
Hence, the abnormal increase in the recovery of B. thuringiensis spores could be due to
multiple factors such as germination – which needs further true account of viability to confirm,
clumping, effects of adhesion, and varying moisture levels during drying.
Overall, the complete reason contributing to the reduction trend of B. thuringiensis spores
could not be addressed from this study, however it can be said that B. thuringiensis spores are
not ideal experimental surrogates for B. anthracis spores based on this data.

66

4.3. Hypothesis III
Measured recovery of Bacillus spores will be lower when water content per HVAC filter
is less
This hypothesis was partially supported by the study. It was anticipated that reduced
amount of water loaded per filter would lead to lower recovery over time. With regard to B.
anthracis spores, the change in water content did not make a significant difference in the
reduction value up to 48 hours, while for B. thuringiensis spores, the hypothesis was somewhat
supported until 48 hours. In both the species, the 168 hour reduction data failed to support the
hypothesis. In other words, filters with lower water content had a higher recovery when
compared to the filter samples with higher water content. This abnormal reverse in trend under
lowered moisture content was not observed in the past. Studies such as (Moritz et al., 2001;
Wang et al., 2001; Lopez et al., 2013; Nanasaki et al., 2010) reported an increase in recovery or
culturability from porous and non-porous matrices at a higher relative humidity or higher water
content loaded per filters.
It is suspected that, besides the primary variable for this hypothesis which is the moisture
content per filter, time could also play a chief role in influencing the recovery which was not
previously evaluated. Results (APPENDIX-C) from the tests that were carried out in order to test
for dehydration rates of different water contents (0.4 mL and 0.2 mL) loaded per filter, led to
some interesting inferences. It was found that, although there was higher amount of water per
filter as in case of 0.4 mL inoculation volume, the corresponding dehydration rate after 50 hours
was significantly accelerated when compared to that of the filter loaded with 0.2 mL water.
Ishihara et al. (1994) demonstrated that the viability of B. subtilis spores, desiccated for
96 hours, declined drastically as the percentage of free water in the spores decreased. Free water
67

is that fraction of water present inside the spore that is available for microbial germination and
growth. Also, free water is capable of diffusing with the water macromolecules present in a
given matrix, via semi-permeable outer-most layer called, exosporium. Swelling of spores when
present in a solution and shrinking of spores when dried were reported in the past, which can be
attributed to the free flow of water molecules through the exosporium (Abel-Santos, 2012). It is
therefore reasonable to assume that, the percentage of free water present inside a spore could be
interrelated to the water present outside the spore. In the study conducted by Ishihara et al.
(1994), the viability of B. subtilis spores declined with high correlation to that of free water
content in the spore up to 24 hours. But, it was not feasible to compare the rate of reduction in
th

free water with respect to viability after the 24 hour, because the viability reached 0% at the
th

24 hour reading (Ishihara et al., 1994). It could be speculated that, the rate of ambient
dehydration influences the free water of the spores, and thereby influencing viability of spores,
significantly.
Therefore, besides the sole phenomena of ‘being dried’ or ‘degree of dehydration’, it is
suggested that time could also play a significant role in affecting the culturable fraction of spores
which are reflected in the recovery. Hence, based on the results from hypothesis I and
supplemental drying tests, it is hypothesized that, higher the rates of dehydration of Bacillus
spores could lead to lower recovery, regardless of the initial moisture content.

68

4.4. Conclusions
Based on the results of this work as well as previously mentioned works herein the
following conclusions could be drawn:
(i) Regardless of the species, the culturable counts of Bacillus spores extracted from
HVAC filters significantly reduced over time. Statistical tests determined that,
time is the most affecting factor in the log-reduction.
(ii) Although both B. anthracis and B. thuringiensis spores reduced over time, their
corresponding reduction trends were remarkably dissimilar. The abnormal
increase in B. thuringiensis spores during 24 hours and subsequent decrease in all
trials could be the result of individual or combined effects of clumping or
clustering, varying moisture content, rate of drying, and delayed germination, but
not limited to these.
(iii) Bacillus thuringiensis spores are not an ideal experimental candidate for B.
anthracis in time-dependent studies.
(iv) Lower moisture content did not significantly lower the recovery of B. anthracis
spores up to 50 hours (p = 0.404), but it did result in lower reduction for B.
thuringiensis spores as expected up to 48 hours (p < 0.001). However lower
moisture content yielded higher recovery for both the species at the 168 hours. It
is therefore suggested from the results of supplemental tests that, lower moisture
content need not lead to a lower recovery, but accelerated dehydration rates could
result in a lower recovery.

69

4.5. Recommendation for Future Work
This study has limitations which would be worthwhile to address in future studies.
Recommendations for future work are as follows:
(i) Extending the recovery study using inoculation via aerosolization of spores, in
absence of moisture to test the aerosol suitability of B. thuringiensis as a potential
surrogate for B. anthracis spores
(ii) Extending the recovery study using the presence of dust particle and commonly
found debris in indoor filters is encouraged for a more realistic setup of indoors.
(iii)Carrying out recovery experiments with different attenuated strains of B.
anthracis spores, and different types of filters used commercially and
domestically.
(iv) Using molecular-based techniques and advanced microscopy are suggested in
order to have a more credible account of true viability, along with culture-based
quantification techniques.

70

APPENDICES

71

APPENDIX A: SUMMARY OF EXPERIMENTAL PLANNING AND DESIGN

72

Based on the objectives and hypotheses, this study, will quantify the recovery of B.
anthracis and B. thuringiensis spores over duration of 168 hours from HVAC filter coupons
under two different moisture conditions. Results from these experiments will be instrumental in
comparing the effects of sampling time and moisture content on the recovery of the two species
as well as the experimental convergence of B. thuringiensis spores and B. anthracis spores.
Therefore, the independent variables in this work are: sampling time, species, and moisture
content. The dependent variable is recovery which is described by log-reduction.
Table A. 1 clearly presents the timeline adopted for each task in this experiment. The
timeline is represented in terms of ‘day’ which does not correspond to any actual calendar date.
Figure A. 1, shows the schematics for a single time point. Triplicate filters, marked as ‘A’, ‘B’,
and ‘C’, are eluted in a centrifuge tube at every time point. The elution processes are sequentially
represented in the adjacent box (Fig. A. 1). These processes correspond to Days 16-18 and 23 in
Table A. 1. Following the elution, plate-culturing method of quantification was used to
enumerate the recovery at all time points. Triplicate plates for every filter resulted in nine
recovery values at every time step. A total of 56 filters and 168 plates along with 16 negatives, a
total of 184 plates were evaluated.

73

Table A. 1. Generalized timeline for recovery experiments: planning module for future use
Day*
Task
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16
a
Prepare BA and
b

BT vegetative cells
Spread BA and BT
cells on separate
TSA slants
Sporulation of BA
and BT (waiting
period)
Extract the BA & BT
spores from the
slants using sterile
Type-I water to yield
master stock
Quantify using serial
dilution plating
Count the plates
from Day 14
Prepare stock
solutions of desired
concentrations from
the master stock
Cut, autoclave, and
store HVAC filter
coupons
Spike 0.4 mL and
0.2 mL of BA stock
solution to different
filters in triplicates
& store in desiccator
74

17

18

19

20

21

22

23

24

Table A. 2. (cont’d) Generalized timeline for recovery experiments: planning module for future use
Elute coupons for 0
hour extractions
Quantify the elution
using serial dilution
plating
Count the plates
from Day 16
Elute coupons for 24
hour extractions
Quantify the elution
using serial dilution
plating
Count the plates
from Day 17
Elute coupons for 48
hour extractions
Quantify the elution
using serial dilution
plating
Count the plates
from Day 18
Elute coupons for
168 hour extractions
Quantify the elution
using serial dilution
plating
Count the plates
from Day 23





1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23
*All the experiments were carried in between 12/26/2013 to 1/24/2014
a, Bacillus anthracis (BA)
b, Bacillus thuringiensis (BT)
Blue box represents the beginning of experiments for BT (Day 20 to Day 28), procedures identical to the beginning Day 16
75

`
24

Figure A. 1. (L) Overall schematics representing sequential processes, inoculation, elution, and quantification involved in the recovery
experiments for a given sampling time. (R) Series of mechanical processes, sonication, vortexing, and shaking involved in the elution
stage.

76

APPENDIX B: APPLIED AND RECOVERY DATA COLLECTED FROM BACILLUS
ANTHRACIS AND BACILLUS THURINGIENSIS SPORES AT DIFFERENT
MOISTURE LEVELS OVER TIME

77

Table B. 1. Mean concentrations of master stocks of Bacillus spores
Species
B. anthracis
B. thuringiensis

Mean plate
counts
78.67
96.67

Conc, spores/
mL
7.87E+07
9.67E+07

Table B. 2. Concentrations of stock solutions of Bacillus spores needed for 0.4 mL inoculation
CM,
spores/
Species
mL
B. anthracis
7.87E+07
B. thuringiensis 9.67E+07

VApplied/
filter, mL
0.4
0.4

CD,
spores/
filter
1.00E+06
1.00E+06

V1,
mL
10
10

C1,
spores/
mL
2.50E+06
2.50E+06

Vstock,
mL
0.318
0.259

Vwater,
mL
9.682
9.741

Table B. 3. Concentrations of stock solutions of Bacillus spores needed for 0.2 mL inoculation
CM,
spores/
Species
mL
B. anthracis
7.87E+07
B. thuringiensis 9.67E+07

VApplied/
filter, mL
0.2
0.2

CD,
spores/ filter
2.00E+06
2.00E+06

V1,
mL
10
10

C1,
spores/
mL
1.00E+07
1.00E+07

Vstock,
mL
1.271
1.034

Vwater,
mL
8.729
8.966

where CM = Concentration of master stock, CD = Desired concentration to be applied on filters,
C1 = Concentration of stock solutions, VApplied = Vol. of inoculant applied on each filter, Vstock
= Vol. of stock solution in 10 mL total stock solution, Vwater = Vol. of water in 10 mL total
stock solution
Equations:
(i) V1 = Vstock + Vwater = 10 mL
(ii) C1 = V

CD

(iii)Vstock =

V1 ∗C1

…. (5)
…. (6)

Applied

…. (7)

CM

(iv) Vwater = 10 − Vstock

…. (8)

78

Table B. 4. Summary of plate counts and recovered concentrations of B. anthracis (0.4 mL)

Filter
A

B

C

A

B

C

A

B

C

A

B

C

Time
(hrs)
0
0
0
0
0
0
0
0
0
24
24
24
24
24
24
24
24
24
48
48
48
48
48
48
48
48
48
168
168
168
168
168
168
168
168
168

Plates
A1
A2
A3
B1
B2
B3
C1
C2
C3
A1
A2
A3
B1
B2
B3
C1
C2
C3
A1
A2
A3
B1
B2
B3
C1
C2
C3
A1
A2
A3
B1
B2
B3
C1
C2
C3

Plate
recovery
58.00
43.00
51.00
58.00
51.00
48.00
78.00
63.00
58.00
41.00
37.00
39.00
38.00
36.00
38.00
63.00
53.00
57.00
32.00
25.00
25.00
26.00
28.00
21.00
44.00
29.00
40.00
30.00
29.00
43.00
85.00
51.00
69.00
66.00
72.00
48.00

Spores
recovered
(Spores/ filter)
1.16E+06
8.60E+05
1.02E+06
1.16E+06
1.02E+06
9.60E+05
1.56E+06
1.26E+06
1.16E+06
8.20E+05
7.40E+05
7.80E+05
7.60E+05
7.20E+05
7.60E+05
1.26E+06
1.06E+06
1.14E+06
6.40E+05
5.00E+05
5.00E+05
5.20E+05
5.60E+05
4.20E+05
8.80E+05
5.80E+05
8.00E+05
6.00E+04
5.80E+04
8.60E+04
1.70E+05
1.02E+05
1.38E+05
1.32E+05
1.44E+05
9.60E+04

79

Table B. 5. Summary of plate counts and recovered concentrations of B. thuringiensis (0.4 mL)

Filter
A

B

C

A

B

C

A

B

C

A

B

C

Time
(hrs)
0
0
0
0
0
0
0
0
0
24
24
24
24
24
24
24
24
24
48
48
48
48
48
48
48
48
48
168
168
168
168
168
168
168
168
168

Plates
A1
A2
A3
B1
B2
B3
C1
C2
C3
A1
A2
A3
B1
B2
B3
C1
C2
C3
A1
A2
A3
B1
B2
B3
C1
C2
C3
A1
A2
A3
B1
B2
B3
C1
C2
C3

Plate
recovery
27.00
39.00
28.00
39.00
31.00
29.00
39.00
47.00
36.00
77.00
86.00
89.00
52.00
46.00
46.00
62.00
60.00
50.00
48.00
50.00
56.00
34.00
48.00
30.00
45.00
51.00
26.00
212.00
207.00
218.00
182.00
203.00
196.00
215.00
208.00
199.00

Spores
recovered
(Spores/ filter)
5.40E+05
7.80E+05
5.60E+05
7.80E+05
6.20E+05
5.80E+05
7.80E+05
9.40E+05
7.20E+05
1.54E+06
1.72E+06
1.78E+06
1.04E+06
9.20E+05
9.20E+05
1.24E+06
1.20E+06
1.00E+06
9.60E+05
1.00E+06
1.12E+06
6.80E+05
9.60E+05
6.00E+05
9.00E+05
1.02E+06
5.20E+05
4.24E+05
4.14E+05
4.36E+05
3.64E+05
4.06E+05
3.92E+05
4.30E+05
4.16E+05
3.98E+05

80

Table B. 6. Summary of plate counts and recovered concentrations of B. anthracis (0.2 mL)

Filter
A

B

C

A

B

C

A

B

C

A

B

C

Time
(hrs)
0
0
0
0
0
0
0
0
0
24
24
24
24
24
24
24
24
24
48
48
48
48
48
48
48
48
48
168
168
168
168
168
168
168
168
168

Plates
A1
A2
A3
B1
B2
B3
C1
C2
C3
A1
A2
A3
B1
B2
B3
C1
C2
C3
A1
A2
A3
B1
B2
B3
C1
C2
C3
A1
A2
A3
B1
B2
B3
C1
C2
C3

Plate
recovery
120.00
122.00
109.00
105.00
120.00
113.00
124.00
103.00
104.00
87.00
75.00
78.00
88.00
90.00
103.00
80.00
87.00
105.00
81.00
65.00
62.00
66.00
58.00
64.00
64.00
50.00
43.00
39.00
37.00
37.00
35.00
33.00
34.00
50.00
47.00
40.00

Spores
recovered
(Spores/ filter)
2.40E+06
2.44E+06
2.18E+06
2.10E+06
2.40E+06
2.26E+06
2.48E+06
2.06E+06
2.08E+06
1.74E+06
1.50E+06
1.56E+06
1.76E+06
1.80E+06
2.06E+06
1.60E+06
1.74E+06
2.10E+06
1.62E+06
1.30E+06
1.24E+06
1.32E+06
1.16E+06
1.28E+06
1.28E+06
1.00E+06
8.60E+05
7.80E+05
7.40E+05
7.40E+05
7.00E+05
6.60E+05
6.80E+05
1.00E+06
9.40E+05
8.00E+05

81

Table B. 7. Summary of plate counts and recovered concentrations of B. thuringiensis (0.2 mL)

Filter
A

B

C

A

B

C

A

B

C

A

B

C

Time
(hrs)
0
0
0
0
0
0
0
0
0
24
24
24
24
24
24
24
24
24
48
48
48
48
48
48
48
48
48
168
168
168
168
168
168
168
168
168

Plates
A1
A2
A3
B1
B2
B3
C1
C2
C3
A1
A2
A3
B1
B2
B3
C1
C2
C3
A1
A2
A3
B1
B2
B3
C1
C2
C3
A1
A2
A3
B1
B2
B3
C1
C2
C3

Plate
recovery
39.00
42.00
34.00
64.00
52.00
48.00
35.00
29.00
38.00
71.00
65.00
63.00
53.00
54.00
67.00
47.00
63.00
60.00
60.00
46.00
55.00
70.00
70.00
61.00
80.00
87.00
53.00
65.00
67.00
72.00
57.00
40.00
62.00
54.00
79.00
60.00

Spores
recovered
(Spores/ filter)
7.80E+05
8.40E+05
6.80E+05
1.28E+06
1.04E+06
9.60E+05
7.00E+05
5.80E+05
7.60E+05
1.42E+06
1.30E+06
1.26E+06
1.06E+06
1.08E+06
1.34E+06
9.40E+05
1.26E+06
1.20E+06
1.20E+06
9.20E+05
1.10E+06
1.40E+06
1.40E+06
1.22E+06
1.60E+06
1.74E+06
1.06E+06
1.30E+06
1.34E+06
1.44E+06
1.14E+06
8.00E+05
1.24E+06
1.08E+06
1.58E+06
1.20E+06

82

APPENDIX C: RECOVERY DATA COLLECTED FROM SUPPLEMENTARY
EXPERIMENTS TO TEST FOR CLUMPING AND DRYING RATES

83

Table C. 1. Summary of plate counts and recovered concentrations of B. thuringiensis without
Tween-20 (0.4 mL)
Spores
Time
Plate
recovered
Filter
(hrs)
Plates recovery (Spores/ filter)
A
0
A1
45.00
9.00E+05
0
A2
54.00
1.08E+06
0
A3
51.00
1.02E+06
B
0
B1
68.00
1.36E+06
0
B2
67.00
1.34E+06
0
B3
63.00
1.26E+06
C
0
C1
54.00
1.08E+06
0
C2
66.00
1.32E+06
0
C3
55.00
1.10E+06
A
24
A1
53.00
1.06E+06
24
A2
52.00
1.04E+06
24
A3
50.00
1.00E+06
B
24
B1
47.00
9.40E+05
24
B2
48.00
9.60E+05
24
B3
39.00
7.80E+05
C
24
C1
76.00
1.52E+06
24
C2
85.00
1.70E+06
24
C3
81.00
1.62E+06
A
48
A1
139.00
2.78E+06
48
A2
156.00
3.12E+06
48
A3
137.00
2.74E+06
B
48
B1
193.00
3.86E+06
48
B2
184.00
3.68E+06
48
B3
190.00
3.80E+06
C
48
C1
157.00
3.14E+06
48
C2
184.00
3.68E+06
48
C3
171.00
3.42E+06
A
168
A1
54.00
1.08E+06
168
A2
64.00
1.28E+06
168
A3
72.00
1.44E+06
B
168
B1
177.00
3.54E+05
168
B2
161.00
3.22E+05
168
B3
192.00
3.84E+05
C
168
C1
114.00
2.28E+05
168
C2
N/A
N/A
168
C3
116.00
2.32E+05

84

Table C. 2. Summary of plate counts and recovered concentrations of B. thuringiensis with
Tween-20 (0.4 mL)
Spores
Time
Plate
recovered
Filter
(hrs)
Plates recovery
(Spores/ filter)
A
0
A1
52.00
1.04E+06
0
A2
61.00
1.22E+06
0
A3
44.00
8.80E+05
B
0
B1
61.00
1.22E+06
0
B2
48.00
9.60E+05
0
B3
47.00
9.40E+05
C
0
C1
65.00
1.30E+06
0
C2
75.00
1.50E+06
0
C3
62.00
1.24E+06
A
24
A1
77.00
1.54E+06
24
A2
65.00
1.30E+06
24
A3
56.00
1.12E+06
B
24
B1
110.00
2.20E+06
24
B2
148.00
2.96E+06
24
B3
168.00
3.36E+06
C
24
C1
129.00
2.58E+06
24
C2
82.00
1.64E+06
24
C3
81.00
1.62E+06
A
48
A1
137.00
2.74E+06
48
A2
146.00
2.92E+06
48
A3
164.00
3.28E+06
B
48
B1
59.00
1.18E+06
48
B2
71.00
1.42E+06
48
B3
88.00
1.76E+06
C
48
C1
78.00
1.56E+06
48
C2
92.00
1.84E+06
48
C3
103.00
2.06E+06
A
168
A1
90.00
1.80E+06
168
A2
112.00
2.24E+06
168
A3
98.00
1.96E+06
B
168
B1
54.00
1.08E+06
168
B2
49.00
9.80E+05
168
B3
55.00
1.10E+06
C
168
C1
63.00
1.26E+06
168
C2
52.00
1.04E+06
168
C3
66.00
1.32E+06

85

Table C. 3. Summary of weights of filters loaded with 0.4 mL and 0.2 mL water content
Volume of water/ filter = 0.4 mL
Volume of water/ filter = 0.2 mL
Weight,
Weight,
Weight,
Date
Time Filters
mg
Filters
mg
Date
Time Filters
mg
A,
B,
A,
25-Aug 11:00 blank
330
blank
339
25-Aug 11:00 blank
345.8
26-Aug 11:00 A
855
B
816.1
26-Aug 11:00 A
612
26-Aug 17:00 A
827.8
B
772.7
26-Aug 17:00 A
581.4
27-Aug 11:00 A
745.4
B
650.8
27-Aug 11:00 A
496.1
27-Aug 17:00 A
724.3
B
617
27-Aug 17:00 A
476.7
28-Aug 11:00 A
649.5
B
491
28-Aug 11:00 A
406.2
28-Aug 17:00 A
625.5
B
450.6
28-Aug 17:00 A
384.6
29-Aug 11:00 A
572.8
B
346.4
29-Aug 11:00 A
346.2
29-Aug 17:00 A
549.4
B
337.3
29-Aug 17:00 A
336.1
30-Aug 11:00 A
503.4
B
337.1
30-Aug 11:00 A
335.5
30-Aug 17:00 A
457.3
B
336.8
30-Aug 17:00 A
334.9
31-Aug 11:00 A
380.7
B
337.2
31-Aug 11:00 A
335.3
31-Aug 17:00 A
352
B
336.8
31-Aug 17:00 A
334.9
1-Sep
11:00 A
328.3
B
336.2
1-Sep
11:00 A
334.9
1-Sep
17:00 A
328.1
B
339.9
1-Sep
17:00 A
334.8
2-Sep
11:00 A
328.4
B
336.4
2-Sep
11:00 A
334.9
2-Sep
17:00 A
328.1
B
336.2
2-Sep
17:00 A
334.2
3-Sep
11:00 A
328.4
B
337.7
3-Sep
11:00 A
334.6
3-Sep
17:00 A
327.8
B
336.7
3-Sep
17:00 A
334.6
The highlighted row is the 0-hour measured weights.

86

Filters

Weight,
mg

B, blank
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B

319.9
596.5
561
480.5
465.1
400.2
379.1
337.6
323.6
319.1
314.6
318.5
318.2
318.1
318.2
317.7
317.5
318.9
317.7

Drying Weights vs. Time
900
800
0.4ml (A)

Weight of filters, mg

700

0.4ml (B)

600

0.2ml (A)

500

0.2ml (B)

400
300
200
100
0
0

24

48

72

96
120
Time, hours

144

168

192

Figure C. 1. Weight of filters (mg) as a function of time for filters loaded with 0.4 mL and 0.2 mL volumes of water

87

APPENDIX D: SCANNING ELECTRON MICROSCOPY

88

Procedure for Scanning Electron Microscopy (SEM)
SEM was used to observe the tangible effects of autoclaving on HVAC filters, and to
observe the nature of B. thuringiensis spores, in the presence and absence of Tween-20.
In order to examine the effects of autoclaving on filter samples, fresh and autoclaved
HVAC filters (identical to experimental dimensions) were subjected to SEM side by side. Pieces
of filter samples were mounted on aluminum stubs using high vacuum carbon tabs (SPI Supplies,
Westchester, PA). The samples were coated with platinum to an approximate thickness of 8 nm
in a Quorum Technologies Q150T Turbo Pimped Coater (Quorum Technologies, Laughton, East
Essex, England). Samples were examined in a Joel JSM-6610LV scanning electron microscope
(JOEL Ltd., Tokyo, Japan).
In order to observe the B. thuringiensis spores with and without the presence of Tween20, small quantities of corresponding samples of B. thuringiensis spores were mixed with an
equal quantity of 4% glutaraldehyde buffered with 0.1 M sodium phosphate at pH 7.4. Fixation
was allowed to proceed for one-half hour at 4oC. One drop of 1% Poly-L-Lysine (Sigma Aldrich
P1399) was placed on a plastic petri dish and a 12 mm round glass cover slip was placed on top
of the drop and allowed to stand for 5 minutes. The coverslip was removed which previously
faced down. The suspension was allowed to settle for 5 minutes. The coverslip was then gently
washed with several drops of water and placed in a graded ethanol series (25%, 50%, 75%, and
95%) for 5 minutes in each step and with 5 minutes changes in 100% ethanol (Klopmparens et
al., 1986). Samples were critical point dried in a Leica EM CPD300 (Leica Microsystems,
Vienna, Austria) using liquid carbon dioxide as the transitional fluid. Samples were mounted on
aluminum stubs using epoxy glue- System Three Quick Cure 5 (System Three Resins, Inc.,
Auburn, WA). Samples were coated with iridium to an approximate thickness of 5.5 nm in a

89

Quorum Technologies Q150T Turbo Pimped Coater (Quorum Technologies, Laughton, East
Essex, England) for 60 s. Samples were examined in a Joel JSM-7500F (cold field emission
electron emitter) scanning electron microscope (JOEL Ltd., Tokyo, Japan). Images obtained
using SEM is presented in Figures C. 1 through C. 18.

90

Figure D. 1. Scattered clusters of B. thuringiensis spores without Tween-20 (200x)

Figure D. 2. Spores of B. thuringiensis with Tween-20 (200x) showing lesser clustering as
compared to the absence of Tween-20
91

Figure D. 3. Joint clumps of B. thuringiensis spores without Tween-20 (1000x)

Figure D. 4. Clumps of B. thuringiensis spores with Tween-20 (1000x)
92

Figure D. 5. Images of B. thuringiensis spores without Tween-20 (1900x)

Figure D. 6. Images of B. thuringiensis spores with Tween-20 (1900x) showing an overall lesser
degree of attachment
93

Figure D. 7. Spores of B. thuringiensis without Tween-20 (2500x)

Figure D. 8. Spores of B. thuringiensis with Tween-20 (2500x)

94

Figure D. 9. Dense clump of B. thuringiensis spores without Tween-20 (3000x)

Figure D. 10. Spores of B. thuringiensis with Tween-20 (3000x)

95

Figure D. 11. Spores of B. thuringiensis without Tween-20 (8000x) with strong clustering
observed at higher magnifications

Figure D. 12. Images of B. thuringiensis spores with Tween-20 (8000x)
96

Figure D. 13. Surface of clumped B. thuringiensis spores without Tween-20 (18000x)

Figure D. 14. Separated spores of B. thuringiensis with Tween-20 (18000x)

97

Figure D. 15. Coiled spores of B. thuringiensis without Tween-20 (20000x)

Figure D. 16. Coiled spores of B. thuringiensis with Tween-20 (20000x) similar to Fig. C. 16

98

Figure D. 17. Connection between two B. thuringiensis spores without Tween-20 (55000x)
observed at very high magnification

Figure D. 18. Spores of B. thuringiensis with Tween-20 (55000x)
99

APPENDIX E: STATISTICAL ANALYSES

100

Frequency

Univariate General Linear Model Analysis on Complete Data Set (N=144)

Log-reduction (LR)
Figure E. 1. Normal distribution of the response variable, log-reduction (LR) for a combined data
set of both B. anthracis and B. thuringiensis using both moisture levels at all sampling times

101

Table E. 1. Descriptive statistics of the dependent variable of the complete data set
Dependent Variable: TLR
Species
B. anthracis

B. thuringiensis

Total

Std.
Deviation
4.24886E-01
1.86080E-01
3.34482E-01
1.95165E-01
1.12075E-01
1.69561E-01
3.42830E-01
1.57414E-01
2.65918E-01

Mean
-3.6870E-01
-2.1723E-01
-2.9296E-01
-1.7246E-01
-2.9461E-01
-2.3353E-01
-2.7058E-01
-2.5592E-01
-2.6325E-01

0.4 mL
0.2 mL
Total
0.4 mL
0.2 mL
Total
0.4 mL
0.2 mL
Total

N
36
36
72
36
36
72
72
72
144

Table E. 2. Tests of Between-Subjects effects (Time, Species, Treatment)
Dependent Variable: TLR
Source
Corrected Model

Type III Sum of
Squares
a
8.354
.916
.674

df
6

Intercept
1
Species*
1
Treatment
Species * Time
2.043
1
Treatment * Time 1.520
1
Species
.506
1
Treatment
.600
1
Time
3.983
1
Error
1.758
137
Total
20.091
144
Corrected Total
10.112
143
a. R Squared = .826 (Adjusted R Squared = .819)

102

Mean
Square
1.392

F
108.502

Sig.
.000

.916
.674

71.381
52.510

.000
.000

2.043
1.520
.506
.600
3.983
.013

159.174
118.427
39.409
46.724
310.389

.000
.000
.000
.000
.000

Table E. 3. Parameter estimates for complete data set
Dependent Variable: TLR

Parameter
Intercept
[Species=1.00] *
[Treatment=1.00]
[Species=1.00] *
[Treatment=2.00]
[Species=2.00] *
[Treatment=1.00]
[Species=2.00] *
[Treatment=2.00]
[Species=1.00] *
Time
[Species=2.00] *
Time
[Treatment=1.00]
* Time
[Treatment=2.00]
* Time
[Species=1.00]

B
-.346
.338

Std. Error
.024
.036

t
-14.288
9.272

Sig.
.000
.000

95% Confidence
Interval
Lower Upper
Bound Bound
-.394
-.298
.266
.410

.299

.032

9.347

.000

.235

.362

.313

.032

9.797

.000

.250

.376

-.003

.000

-11.173

.000

-.003

-.002

.001

.000

3.395

.001

.000

.001

-.003

.000

-10.882

.000

-.004

-.003

[Species=2.00]

0

a

0

a

0

a

0

a

[Treatment=1.00] 0a
[Treatment=2.00] 0a
Time

a

0
a. This parameter is set to zero because it is redundant.

103

Figure E. 2. Scatter plot matrix of standard residuals, observed, and predicted values of the
dependent variable (LR) of the complete data set

104

Univariate General Linear Model Analysis on Surfactant Data Set (N=71)
Table E. 4. Descriptive statistics of the dependent variable of surfactant data set
Dependent Variable: TLR
Tween
No Tween
Yes Tween
Total

Mean
1.9673E-03
9.6443E-03
5.8599E-03

Std. Deviation
2.42890E-02
1.15472E-02
1.91875E-02

N
35
36
71

Table E. 5. Tests of Between-Subjects effects (Time, Tween-20)
Dependent Variable: TLR
Type III Sum of
Source
Squares
df
a
Corrected Model .006
3
Intercept
.005
1
Tween
6.103E-05
1
Time
.003
1
Tween * Time
.002
1
Error
.020
67
Total
.028
71
Corrected Total .026
70
a. R Squared = .234 (Adjusted R Squared = .199)

Mean
Square
.002
.005
6.103E-05
.003
.002
.000

F
6.805

Sig.
.000

17.256
.207
9.681
7.884

.000
.651
.003
.007

Table E. 6. Parameter estimates for surfactant data set
Dependent Variable: TLR

Parameter
Intercept
[Tween=1.00]
[Tween=2.00]

B
.010
.003

Std. Error
.004
.006

t
2.619
.455

Sig.
.011
.651

95% Confidence
Interval
Lower
Upper
Bound
Bound
.002
.018
-.009
.014

.827
.007

-9.810E-05 7.868E-05
.000
-5.193E-05

a

0
Time
-9.709E-06
4.428E-05
-.219
[Tween=1.00] .000
6.397E-05
-2.808
* Time
[Tween=2.00] 0a
* Time
a. This parameter is set to zero because it is redundant.

105

Figure E. 3. Scatter plot matrix of standard residuals, observed, and predicted values of the
dependent variable (LR) of the surfactant data set

106

APPENDIX F: SUMMARY OF REAL-TIME PCR FOR DIFFERENT MOISTURE
LEVELS TO QUANTIFY BACILLUS ANTHRACIS

107

Table F. 1. Forward and Reverse Primer Sequences for B. anthracis Sterne spores*
Position
DNA Sequence
Forward, 5’-3’
AGAAGTGCATGCGTCGTTCTT
Reverse, 5’-3’
GCGACCGTACTTGAATTCGAA
*Primers were obtained from Intergrated DNA Technologies®
Table F. 2. Composition of Reaction Mixture
Component
Volume
DNA extracted from Bacillus anthracis Sterne spores 5 µL
solution*
Roche Lightcycler® 480 SYBR Green I Master
10 µL
Forward primer
1 µL
Reverse primer
1 µL
RNA-ase free wáter
3 µL
*DNA was extracted using heat-inactivation method. Spore solutions
were oven-heated at 100oC for 30 minutes, followed by centrifugation at
11000xg for 5 minutes
Table F. 3. Summary of Lightcycler® Operation Cycles
Operation Cycles
Pre-incubation
Amplification
(45 cycles)
Melting curve
Cooling

Temperatures, time
95oC, 5 minutes
95oC, 10 s
55oC, 20 s
72oC, 10 s
95oC, 5 s
65oC, 1 hour
97oC, continuous
40oC, 10 s

108

Table F. 4. Summary of applied and recovered concentrations of B. anthracis Sterne spores enumerated using qPCR
Volume of water/ filter = 0.4 mL
Volume of water/ filter = 0.2 mL
Applied concentration
(spores/ filter)

Filter
A
B
C
A
B
C
A
B
C
A
B
C

Time
(hrs)
0
0
0
24
24
24
48
48
48
168
168
168

Copies/
reaction
4.91E+04
4.52E+04
7.16E+04
4.39E+04
4.85E+04
6.52E+04
1.94E+05
2.73E+05
2.51E+05
1.95E+05
5.45E+04
1.71E+05

Applied concentration
(spores/ filter)

2.34E+09
Copies/ filter
(spores/
filter)
1.96E+08
1.81E+08
2.86E+08
1.76E+08
1.94E+08
2.61E+08
7.76E+08
1.09E+09
1.00E+09
7.80E+08
2.18E+08
6.84E+08

Log
reduction
-1.08E+00
-1.11E+00
-9.12E-01
-1.12E+00
-1.08E+00
-9.53E-01
-4.79E-01
-3.31E-01
-3.67E-01
-4.77E-01
-1.03E+00
-5.34E-01

Filter
A
B
C
A
B
C
A
B
C
A
B
C

109

Time
(hrs)
0
0
0
24
24
24
48
48
48
168
168
168

Copies/
reaction
1.33E+05
1.38E+05
1.29E+05
1.12E+05
1.06E+05
1.24E+05
3.51E+05
5.12E+05
2.84E+05
3.44E+05
2.95E+05
3.08E+05

7.58E+09
Copies/ filter
(spores/
filter)
5.32E+08
5.52E+08
5.16E+08
4.48E+08
4.24E+08
4.96E+08
1.40E+09
2.05E+09
1.14E+09
1.38E+09
1.18E+09
1.23E+09

Log
reduction
-1.15E+00
-1.14E+00
-1.17E+00
-1.23E+00
-1.25E+00
-1.18E+00
-7.32E-01
-5.68E-01
-8.24E-01
-7.41E-01
-8.08E-01
-7.89E-01

Figure F. 1. Log-concentration of B. anthracis Sterne spores as a function of time for 0.4 mL volume of inoculation quantified using
culture-based and molecular-based techniques. Error bars indicate standard deviation from the mean.

110

Figure F. 2. Log-concentration of B. anthracis Sterne spores as a function of time for 0.2 mL volume of inoculation quantified using
culture-based and molecular-based techniques. Error bars indicate standard deviation from the mean

111

REFERENCES

112

REFERENCES
Abad, FX, RM Pinto, and Albert Bosch. 1994. “Survival of Enteric Viruses on Environmental
Fomites.” Applied and Environmental … 60 (10): 3704–10.
http://aem.asm.org/content/60/10/3704.short.
Abel-Santos, Ernesto. 2012. Bacterial Spores: Current Research and Applications. Norfolk, UK:
Caister Academic Press.
Akiba, Y. 1986. “VI. Germination of Bacillus Thuringiensis Spores in the Soil.” Applied
Entomology and Zoology 21 (1): 76–80. http://pdf.lookchem.com/pdf/22/b80c768f-ead641a0-8060-45a5751e78ca.pdf.
Barakat, Lydia A, Howard L Quentzel, John A Jernigan, David L Kirschke, Kevin Griffith,
Stephen M Spear, Katherine Kelley, et al. 2002. “Fatal Inhalational Anthrax in a 94Year-Old Connecticut Woman.” Jama 287 (7): 863–68.
http://jama.jamanetwork.com/article.aspx?articleid=194657.
Bishop, A. H. 2014. “Germination and Persistence of Bacillus Anthracis and Bacillus
Thuringiensis in Soil Microcosms.” Journal of Applied Microbiology, August.
doi:10.1111/jam.12620.
Block, SM. 2001. “The Growing Threat of Biological Weapons The Terrorist Threat Is Very
Real, and It’s about to Get Worse. Scientists Should Concern Themselves before It's
Too.” American Scientist.
http://scholar.google.com/scholar?hl=en&btnG=Search&q=intitle:The+Growing+Threat
+of+Biological+Weapons#0.
Borio, Luciana, Dennis Frank, V Mani, and C Chiriboga. 2001. “Death due to BioterrorismRelated Inhalational Anthrax: Report of 2 Patients.” Jama 286 (20).
http://jama.jamanetwork.com/article.aspx?articleid=194409.
Brachman, PS, AF Kaufman, and FG Dalldorf. 1966. “Industrial Inhalation Anthrax.”
Bacteriological Reviews 30 (3): 646–57.
http://www.ncbi.nlm.nih.gov/pmc/articles/PMC378258/.
Bush, LM, and BH Abrams. 2001. “Index Case of Fatal Inhalational Anthrax due to Bioterrorism
in the United States.” New England Journal of Medicine 345 (22): 1607–10.
http://www.nejm.org/doi/full/10.1056/nejmoa012948.
Buttner, MP, and Patricia Cruz. 2004. “Evaluation of the Biological Sampling Kit (BiSKit) for
Large-Area Surface Sampling.” Applied and … 70 (12). doi:10.1128/AEM.70.12.7040.
Calfee, M Worth, Laura J Rose, Stephen Morse, Dino Mattorano, Matt Clayton, Abderrahmane
Touati, Nicole Griffin-Gatchalian, Christina Slone, and Neal McSweeney. 2013.
113

“Comparative Evaluation of Vacuum-Based Surface Sampling Methods for Collection of
Bacillus Spores.” Journal of Microbiological Methods 95 (3): 389–96.
doi:10.1016/j.mimet.2013.10.015.
Calfee, M Worth, Laura J Rose, Jenia Tufts, Stephen Morse, Matt Clayton, Abderrahmane
Touati, Nicole Griffin-Gatchalian, Christina Slone, and Neal McSweeney. 2014.
“Evaluation of Sampling Methods for Bacillus Spore-Contaminated HVAC Filters.”
Journal of Microbiological Methods 96 (January): 1–5.
doi:10.1016/j.mimet.2013.10.012.
Canter, Dorothy a. 2007. “Addressing Residual Risk Issues at Anthrax Cleanups: How Clean Is
Safe?” Journal of Toxicology and Environmental Health. Part A 68 (11-12): 1017–32.
doi:10.1080/15287390590912621.
Carrera, M, R O Zandomeni, J Fitzgibbon, and J-L Sagripanti. 2007. “Difference between the
Spore Sizes of Bacillus Anthracis and Other Bacillus Species.” Journal of Applied
Microbiology 102 (2): 303–12. doi:10.1111/j.1365-2672.2006.03111.x.
Carrera, Monica, J Kesavan, R Zandomeni, and JL Sagripanti. 2005. “Method to Determine the
Number of Bacterial Spores within Aerosol Particles.” Aerosol Science and Technology,
no. August 2014: 37–41. doi:10.1080/02786820500352098.
Chensue, Stephen W. 2003. “Exposing a Killer: Pathologists Angle for Anthrax.” The American
Journal of Pathology 163 (5): 1699–1702. doi:10.1016/S0002-9440(10)63526-2.
Christopher, G. W.; Cieslak, T. J.; Pavlin, J. A.; Eitzen, E. M., Jr. 1997. “Biological Warfare: A
Historical Perspective.” JAMA : The Journal of the American Medical Association 278
(5): 412–17.
http://scholar.google.com/scholar?hl=en&btnG=Search&q=intitle:Biological+warfare.+A
+historical+perspective#0.
Clark Burton, Nancy, Atin Adhikari, Sergey a Grinshpun, Richard Hornung, and Tiina Reponen.
2005. “The Effect of Filter Material on Bioaerosol Collection of Bacillus Subtilis Spores
Used as a Bacillus Anthracis Simulant.” Journal of Environmental Monitoring : JEM 7
(5): 475–80. doi:10.1039/b500056d.
Cole, LA. 2010. “Anthrax as a Weapon of War and Terrorism.” Bacillus Anthracis and Anthrax.
http://onlinelibrary.wiley.com/doi/10.1002/9780470891193.ch15/summary.
Cote, C K, J Bozue, N Twenhafel, and S L Welkos. 2009. “Effects of Altering the Germination
Potential of Bacillus Anthracis Spores by Exogenous Means in a Mouse Model.” Journal
of Medical Microbiology 58 (Pt 6): 816–25. doi:10.1099/jmm.0.008656-0.
Dang, JL, Karen Heroux, and John Kearney. 2001. “Bacillus Spore Inactivation Methods Affect
Detection Assays.” Applied and Environmental Microbiology 67 (8): 3665–70.
doi:10.1128/AEM.67.8.3665.
114

Day, Judy, Avner Friedman, and Larry S Schlesinger. 2011. “Modeling the Host Response to
Inhalation Anthrax.” Journal of Theoretical Biology 276 (1). Elsevier: 199–208.
doi:10.1016/j.jtbi.2011.01.054.
Day, Thomas G. 2003. “The Autumn 2001 Anthrax Attack on the United States Postal Service:
The Consequences and Response.” Journal of Contingencies and Crisis Management 11
(3): 110–17. doi:10.1111/1468-5973.1103004.
Dixon, T C, M Meselson, J Guillemin, and P C Hanna. 1999. “Anthrax.” The New England
Journal of Medicine 341 (11): 815–26. doi:10.1056/NEJM199909093411107.
Dragon, D C, and R P Rennie. 1995. “The Ecology of Anthrax Spores: Tough but Not
Invincible.” The Canadian Veterinary Journal. La Revue Vétérinaire Canadienne 36 (5):
295–301.
http://www.pubmedcentral.nih.gov/articlerender.fcgi?artid=1686874&tool=pmcentrez&r
endertype=abstract.
Dragon, D.C., R.P. Rennie, and B.T. Elkin. 2001. “Detection of Anthrax Spores in Endemic
Regions of Northern Canada.” Journal of Applied Microbiology 91 (3): 435–41.
doi:10.1046/j.1365-2672.2001.01389.x.
Dulmage, HT, JA Correa, and AJ Martinez. 1970. “Coprecipitation with Lactose as a Means of
Recovering the Spore-Crystal Complex of Bacillus Thuringiensis.” Journal of
Invertebrate Pathology 15: 15–20.
http://www.sciencedirect.com/science/article/pii/0022201170900935.
Edmonds, Jason M, Patricia J Collett, Erica R Valdes, Evan W Skowronski, Gregory J Pellar,
and Peter a Emanuel. 2009. “Surface Sampling of Spores in Dry-Deposition Aerosols.”
Applied and Environmental Microbiology 75 (1): 39–44. doi:10.1128/AEM.01563-08.
Farnsworth, James E., Sagar M. Goyal, Seung Won Kim, Thomas H. Kuehn, Peter C. Raynor,
M. a. Ramakrishnan, Senthilvelan Anantharaman, and Weihua Tang. 2006.
“Development of a Method for Bacteria and Virus Recovery from Heating, Ventilation,
and Air Conditioning (HVAC) Filters.” Journal of Environmental Monitoring 8 (10):
1006. doi:10.1039/b606132j.
Fritze, D, and R Pukall. 2001. “Reclassification of Bioindicator Strains Bacillus Subtilis DSM
675 and Bacillus Subtilis DSM 2277 as Bacillus Atrophaeus.” International Journal of
Systematic and Evolutionary Microbiology 51 (Pt 1): 35–37.
http://www.ncbi.nlm.nih.gov/pubmed/11211269.
Graham-Smith, G S. 1941. “Further Observations on the Longevity of Dry Spores of B.
Anthracis.” The Journal of Hygiene 41 (5-6): 496.
http://www.pubmedcentral.nih.gov/articlerender.fcgi?artid=2239232&tool=pmcentrez&r
endertype=abstract.

115

Graham-Smith, GS. 1930. “The Longevity of Dry Spores of B. Anthracis.” Journal of Hygiene,
213–15. http://journals.cambridge.org/abstract_S0022172400010391.
Greenberg, David L, Joseph D Busch, Paul Keim, and David M Wagner. 2010. “Identifying
Experimental Surrogates for Bacillus Anthracis Spores: A Review.” Investigative
Genetics 1 (1): 4. doi:10.1186/2041-2223-1-4.
Hachisuka, Y, S Kozuka, and M Tsujikawa. 1984. “Exosporia and Appendages of Spores of
Bacillus Species.” Microbiology and Immunology 28 (5): 619–24.
http://onlinelibrary.wiley.com/doi/10.1111/j.1348-0421.1984.tb00714.x/abstract.
He, Jin, Xiaofeng Luo, Shouwen Chen, Lili Cao, Ming Sun, and Ziniu Yu. 2003.
“Determination of Spore Concentration in Bacillus Thuringiensis through the Analysis of
Dipicolinate by Capillary Zone Electrophoresis.” Journal of Chromatography A 994:
207–12. http://www.sciencedirect.com/science/article/pii/S0021967303004229.
Herzog, Amanda B, S Devin McLennan, Alok K Pandey, Charles P Gerba, Charles N Haas, Joan
B Rose, and Syed a Hashsham. 2009. “Implications of Limits of Detection of Various
Methods for Bacillus Anthracis in Computing Risks to Human Health.” Applied and
Environmental Microbiology 75 (19): 6331–39. doi:10.1128/AEM.00288-09.
Hong, Tao, and Patrick L Gurian. 2012. “Characterizing Bioaerosol Risk from Environmental
Sampling.” Environmental Science & Technology 46 (12): 6714–22.
doi:10.1021/es300197n.
Hong, Tao, Patrick L Gurian, Yin Huang, and Charles N Haas. 2012. “Prioritizing Risks and
Uncertainties from Intentional Release of Selected Category A Pathogens.” PloS One 7
(3): e32732. doi:10.1371/journal.pone.0032732.
Hong, Tao, Patrick L Gurian, and Nicholas F Dudley Ward. 2010. “Setting Risk-Informed
Environmental Standards for Bacillus Anthracis Spores.” Risk Analysis : An Official
Publication of the Society for Risk Analysis 30 (10): 1602–22. doi:10.1111/j.15396924.2010.01443.x.
Husmark, U, and U Rönner. 1990. “Forces Involved in Adhesion of Bacillus Cereus Spores to
Solid Surfaces under Different Environmental Conditions.” The Journal of Applied
Bacteriology 69 (4): 557–62. http://www.ncbi.nlm.nih.gov/pubmed/2127266.
Husmark, U. 1993. “Adhesion Mechanisms of Bacterial Spores to Solid Surfaces”. Chalmers
University of Technology and SIK, The Swedish Institute for Food Research, Göteborg,
Sweden.
Ilstrup, DM. 1990. “Statistical Methods in Microbiology.” Clinical Microbiology Reviews 3 (3):
219–26. http://cmr.asm.org/content/3/3/219.short.

116

Inglesby, TV, DA Henderson, JG Bartlett, MS Ascher, and et al. 1999. “Anthrax as a Biological
Weapon: Medical and Public Health Management.” Jama 281 (18): 1735–46.
http://archderm.jamanetwork.com/article.aspx?articleid=189876.
Inglesby, TV, T O’Toole, DA Henderson, JG Bartlett, and et al. 2002. “Anthrax as a Biological
Weapon, 2002: Updated Recommendations for Management.” Jama 287 (17).
http://jama.jamanetwork.com/article.aspx?articleid=194886.
Ishihara, Yoshimi, Jiro Takano, S Mashimo, and M Yamamura. 1994. “Determination of the
Water Necessary for Survival of Bacillus Subtilis Vegetative Cells and Spores.”
Thermochimica Acta 235 (2): 153–60.
http://www.sciencedirect.com/science/article/pii/004060319485159X.
Jenkins, Peggy L, Thomas J Phillips, Elliot J Mulberg, and Steve P Hui. 1992. “Activity Patterns
of Californians: Use of and Proximity to Indoor Pollutant Sources.” Atmospheric
Environment. Part A. General Topics 26 (12): 2141–48. doi:10.1016/09601686(92)90402-7.
Jernigan, J a, D S Stephens, D a Ashford, C Omenaca, M S Topiel, M Galbraith, M Tapper, et al.
2001. “Bioterrorism-Related Inhalational Anthrax: The First 10 Cases Reported in the
United States.” Emerging Infectious Diseases 7 (6): 933–44.
doi:10.3201/eid0706.010604.
Katz, E, and a L Demain. 1977. “The Peptide Antibiotics of Bacillus: Chemistry, Biogenesis, and
Possible Functions.” Bacteriological Reviews 41 (2): 449–74.
http://www.pubmedcentral.nih.gov/articlerender.fcgi?artid=414008&tool=pmcentrez&re
ndertype=abstract.
Kim, Seung Won, Peter C. Raynor, Thomas H. Kuehn, Sagar M. Goyal, M. a. Ramakrishnan,
Senthilvelan Anantharaman, and James E. Farnsworth. 2008. “Optimizing the Recovery
of Surrogates for Bacterial Bioterrorism Agents from Ventilation Filters.” CLEAN - Soil,
Air, Water 36 (7): 601–8. doi:10.1002/clen.200700182.
Klomparens, Karen L., Stanley L. Flegler, and Gary R. Hooper. 1986. Procedures for
Transmission and Scanning Electron Microscopy for Biological and Medical Science: A
Laboratory Manual. Ladd Research Industries.
Krauter, Paula, and Arthur Biermann. 2007. “Reaerosolization of Fluidized Spores in Ventilation
Systems.” Applied and Environmental Microbiology 73 (7): 2165–72.
doi:10.1128/AEM.02289-06.
Krauter, Paula, Arthur Biermann, and Lloyd D. Larsen. 2005. “Transport Efficiency and
Deposition Velocity of Fluidized Spores in Ventilation Ducts.” Aerobiologia 21 (3-4):
155–72. doi:10.1007/s10453-005-9001-z.

117

Lai, AC K, and WW Nazaroff. 2000. “Modeling Indoor Particle Deposition from Turbulent Flow
onto Smooth Surfaces.” Journal of Aerosol Science 31 (4): 463–76.
http://www.sciencedirect.com/science/article/pii/S0021850299005364.
Lopez, Gerardo U, Charles P Gerba, Akrum H Tamimi, Masaaki Kitajima, Sheri L Maxwell, and
Joan B Rose. 2013. “Transfer Efficiency of Bacteria and Viruses from Porous and
Nonporous Fomites to Fingers under Different Relative Humidity Conditions.” Applied
and Environmental Microbiology 79 (18): 5728–34. doi:10.1128/AEM.01030-13.
Manchee, RJ, and MG Broster. 1994. “Formaldehyde Solution Effectively Inactivates Spores of
Bacillus Anthracis on the Scottish Island of Gruinard.” Applied and … 60 (11): 4167–71.
http://aem.asm.org/content/60/11/4167.short.
Manchee, RJ, MG Broster, and J Melling. 1981. “Bacillus Anthracis on Gruinard Island.” Nature
297 (19): 254–55. http://www.nature.com/nature/journal/v294/n5838/abs/294254a0.html.
Masago, Yoshifumi, Tomoyuki Shibata, and JB Rose. 2008. “Bacteriophage P22 and
Staphylococcus Aureus Attenuation on Nonporous Fomites as Determined by Plate
Assay and Quantitative PCR.” Applied and Environmental … 74 (18). American Society
for Microbiology (ASM): 5838–40. doi:10.1128/AEM.00352-08.
Maus, R, A Goppelsröder, and H Umhauer. 2000. “Survival of Bacterial and Mold Spores in Air
Filter Media.” Atmospheric Environment 35 (1): 105–13.
http://www.sciencedirect.com/science/article/pii/S1352231000002806.
Meselson, M, J Guillemin, M Hugh-Jones, a Langmuir, I Popova, a Shelokov, and O
Yampolskaya. 1994. “The Sverdlovsk Anthrax Outbreak of 1979.” Science (New York,
N.Y.) 266 (5188): 1202–8. http://www.ncbi.nlm.nih.gov/pubmed/7973702.
Minett, FC, and MR Dhanda. 1941. “Multiplication of B. Anthracis and Cl. Chauvei in Soil and
Water.” Ind J Vet Sci Anim Husb, no. 11: 308–28.
http://deriv.nls.uk/dcn6/7525/75258330.6.pdf.
Mittal, Himanshu, SR Parks, and Thomas Pottage. 2011. “Survival of Microorganisms on HEPA
Filters.” Journal of the American … 16 (3): 163–66.
http://www.absa.org/abj/abj/ABJ2011v16n3.pdf#page=37.
Möritz, M, H Peters, B Nipko, and H Rüden. 2001. “Capability of Air Filters to Retain Airborne
Bacteria and Molds in Heating, Ventilating and Air-Conditioning (HVAC) Systems.”
International Journal of Hygiene and Environmental Health 203 (5-6): 401–9.
doi:10.1078/1438-4639-00054.
Murali, Bharathi, and Jade Mitchell. 2014. “The Effect of Recovery on Modeling Inactivation of
Bacillus Spores on HVAC Filters.” In Ames, D.P., Quinn, N.W.T., Rizzoli, A.E. (Eds.),
Proceedings of the 7th International Congress on Environmental Modelling and

118

Software, June 15-19, San Diego, California, USA, 1492–99.
http://www.iemss.org/sites/iemss2014/papers/iemss2014_submission_143.pdf.
Nanasaki, Yusuke, Tomoaki Hagiwara, Hisahiko Watanabe, and Takaharu Sakiyama. 2010.
“Removability of Bacterial Spores Made Adherent to Solid Surfaces from Suspension
with and without Drying.” Food Control 21 (11). Elsevier Ltd: 1472–77.
doi:10.1016/j.foodcont.2010.04.016.
Nicholson, WL, and N. Munakata. 2000. “Resistance of Bacillus Endospores to Extreme
Terrestrial and Extraterrestrial Environments.” Microbiology and … 64 (3): 548–72.
doi:10.1128/MMBR.64.3.548-572.2000.
Noris, Federico, Jeffrey a. Siegel, and Kerry a. Kinney. 2011. “Evaluation of HVAC Filters as a
Sampling Mechanism for Indoor Microbial Communities.” Atmospheric Environment 45
(2). Elsevier Ltd: 338–46. doi:10.1016/j.atmosenv.2010.10.017.
Redmond, C, M J Pearce, R J Manchee, and B P Berdal. 1998. “Deadly Relic of the Great War.”
Nature 393 (6687): 747–48. doi:10.1038/31612.
Riedel, Stefan. 2004. “Biological Warfare and Bioterrorism: A Historical Review.” Proceedings
(Baylor University. Medical Center) 17 (4): 400–406.
http://www.pubmedcentral.nih.gov/articlerender.fcgi?artid=1200679&tool=pmcentrez&r
endertype=abstract.
Riedel, Stefan. 2005. “Anthrax: A Continuing Concern in the Era of Bioterrorism.” Proceedings
(Baylor University. Medical Center) 18 (3): 234–43.
http://www.pubmedcentral.nih.gov/articlerender.fcgi?artid=1200731&tool=pmcentrez&r
endertype=abstract.
Rönner, U, U Husmark, and A Henriksson. 1990. “Adhesion of Bacillus Spores in Relation to
Hydrophobicity.” Journal of Applied … 69 (Atcc 12980): 550–56.
http://onlinelibrary.wiley.com/doi/10.1111/j.1365-2672.1990.tb01547.x/full.
Sagripanti, J-L, M Carrera, J Insalaco, M Ziemski, J Rogers, and R Zandomeni. 2007. “Virulent
Spores of Bacillus Anthracis and Other Bacillus Species Deposited on Solid Surfaces
Have Similar Sensitivity to Chemical Decontaminants.” Journal of Applied Microbiology
102 (1): 11–21. doi:10.1111/j.1365-2672.2006.03235.x.
Saleh, S M, R F Harris, and O N Allen. 1970. “Fate of Bacillus Thuringiensis in Soil: Effect of
Soil pH and Organic Amendment.” Canadian Journal of Microbiology 16 (8): 677–80.
http://www.ncbi.nlm.nih.gov/pubmed/4921875.
Setlow, Peter. 2003. “Spore Germination.” Current Opinion in Microbiology 6 (6): 550–56.
doi:10.1016/j.mib.2003.10.001.

119

Sextro, RG, and DM Lorenzetti. 2002. “Modeling the Spread of Anthrax in Buildings.”
Proceedings: Indoor Air …. http://energy.lbl.gov/ie/pdf/LBNL-49537.pdf.
Solon, Ian. 2010. “The Extraction of a Bacillus Anthracis Surrogate from Pleated HVAC Filter
Samples”. Drexel University. http://idea.library.drexel.edu/handle/1860/3205.
Solon, Ian, PL Gurian, and Hernando Perez. 2012. “The Extraction of a Bacillus Anthracis
Surrogate from HVAC Filters.” Indoor and Built Environment.
doi:10.1177/1420326X11420307.
Spencer, R C. 2003. “Bacillus Anthracis.” Journal of Clinical Pathology 56 (3): 182–87.
http://www.pubmedcentral.nih.gov/articlerender.fcgi?artid=1769905&tool=pmcentrez&r
endertype=abstract.
Stanley, Nicholas J, Thomas H Kuehn, Seung Won Kim, Peter C Raynor, Senthilvelan
Anantharaman, M a Ramakrishnan, and Sagar M Goyal. 2008. “Background Culturable
Bacteria Aerosol in Two Large Public Buildings Using HVAC Filters as Long Term,
Passive, High-Volume Air Samplers.” Journal of Environmental Monitoring : JEM 10
(4): 474–81. doi:10.1039/b719316e.
Sternbach, George. 2003. “The History of Anthrax.” The Journal of Emergency Medicine 24 (4):
463–67. doi:10.1016/S0736-4679(03)00079-9.
Sugimoto, EE. 1996. “Recovery of Bacterial Spores Dried on Aluminium Strips.” Journal of
Applied Bacteriology 80: 147–52. http://onlinelibrary.wiley.com/doi/10.1111/j.13652672.1996.tb03202.x/abstract.
Tripp, MR. 1960. “Mechanisms of Removal of Injected Microorganisms from the American
Oyster, Crassostrea Virginica (Gmelin).” Biological Bulletin 119 (2): 273–82.
http://www.jstor.org/stable/1538928.
Tufts, Jenia a M, M Worth Calfee, Sang Don Lee, and Shawn P Ryan. 2013. “Bacillus
Thuringiensis as a Surrogate for Bacillus Anthracis in Aerosol Research.” World Journal
of Microbiology & Biotechnology 30 (5): 1453–61. doi:10.1007/s11274-013-1576-x.
Van Cuyk, Sheila, Alina Deshpande, Attelia Hollander, David O Franco, Nerayo P Teclemariam,
Julie a Layshock, Lawrence O Ticknor, Michael J Brown, and Kristin M Omberg. 2012.
“Transport of Bacillus Thuringiensis Var. Kurstaki from an Outdoor Release into
Buildings: Pathways of Infiltration and a Rapid Method to Identify Contaminated
Buildings.” Biosecurity and Bioterrorism : Biodefense Strategy, Practice, and Science 10
(2): 215–27. doi:10.1089/bsp.2011.0081.
Wang, Zheng, Tiina Reponen, and SA Grinshpun. 2001. “Effect of Sampling Time and Air
Humidity on the Bioefficiency of Filter Samplers for Bioaerosol Collection.” Journal of
Aerosol Science 32: 661–74.
http://www.sciencedirect.com/science/article/pii/S0021850200001087.
120

Waring, M S, and J a Siegel. 2008. “Particle Loading Rates for HVAC Filters, Heat Exchangers,
and Ducts.” Indoor Air 18 (3): 209–24. doi:10.1111/j.1600-0668.2008.00518.x.
Weis, Christopher P, Anthony J Intrepido, Aubrey K Miller, Patricia G Cowin, Mark a Durno,
Joan S Gebhardt, and Robert Bull. 2002. “Secondary Aerosolization of Viable Bacillus
Anthracis Spores in a Contaminated US Senate Office.” JAMA : The Journal of the
American Medical Association 288 (22): 2853–58.
http://www.ncbi.nlm.nih.gov/pubmed/12472327.
Wilkening, DA. 2006. “Sverdlovsk Revisited: Modeling Human Inhalation Anthrax.”
Proceedings of the National Academy of … 103 (20).
http://www.pnas.org/content/103/20/7589.short.
Woloszyn, Monika, Targo Kalamees, Marc Olivier Abadie, Marijke Steeman, and Angela Sasic
Kalagasidis. 2009. “The Effect of Combining a Relative-Humidity-Sensitive Ventilation
System with the Moisture-Buffering Capacity of Materials on Indoor Climate and Energy
Efficiency of Buildings.” Building and Environment 44 (3): 515–24.
doi:10.1016/j.buildenv.2008.04.017.

121