Citation
Novel method of bioremediation and characterization of bacterial communities on arsenic-impacted museum collections

Material Information

Title:
Novel method of bioremediation and characterization of bacterial communities on arsenic-impacted museum collections
Creator:
Subotic, Sladjana ( author )
Place of Publication:
Denver, Colo.
Publisher:
University of Colorado Denver
Publication Date:
Language:
English
Physical Description:
1 electronic file (93 pages) : ;

Thesis/Dissertation Information

Degree:
Master's ( Master of science)
Degree Grantor:
University of Colorado Denver
Degree Divisions:
Department of Integrative Biology, CU Denver
Degree Disciplines:
Biology

Subjects

Subjects / Keywords:
Bioremediation ( lcsh )
Bioinformatics -- Statistical methods ( lcsh )
Arsenic -- Environmental aspects ( lcsh )
Genre:
bibliography ( marcgt )
theses ( marcgt )
non-fiction ( marcgt )

Notes

Review:
Metal-based pesticides, including toxic arsenic salts, were widely used prior to the 1970s for the prevention of rodent and insect damage to museum collections. An estimated 80% of collections housed in U.S. and Canadian museums have exogenous metals present, and the use of bacteria in the removal of these metals is of interest. Research shows that Rhodopseudomonas palustris , a metabolically versatile proteobacterium, is able to volatilize arsenic via methylation resulting in the conversion of the metal into a gas which can be collected and removed. In this study, R. palustris tolerated concentrations of arsenic up to 250 ppm and showed the ability to remove up to 78% of arsenic from a starting concentration of 10 ppm soluble arsenic within 19 days. In order to optimize the potential application of bacterial volatilization of arsenic as a remediation technology for arsenic-treated museum specimens, the presence of in situ bacteria on the surface of museum collections needs to be addressed. High throughput sequencing revealed diverse bacterial communities associated with the museum specimens examined at the Denver Museum of Nature and Science. Bacterial community composition seemed strongly influenced by several factors including type of materials comprising the specimen, the presence of arsenic, and which museum Collection the specimen was housed in. Organisms of high abundance across the items sampled included the genera Ralstonia , Sediminbacterium , Acinetobacter and the family Enterobacteriaceae . Additionally, testing of surface-associated arsenic revealed a wide range of concentrations on arsenic-impacted items at the Denver Museum of Nature and Science. This study is the first to utilize high throughput sequencing techniques to characterize the bacterial communities associated with museum collections.
Bibliography:
Includes bibliographical references.
System Details:
System requirements: Adobe Reader.
Statement of Responsibility:
by Sladjana Subotic.

Record Information

Source Institution:
University of Colorado Denver Collections
Holding Location:
Auraria Library
Rights Management:
All applicable rights reserved by the source institution and holding location.
Resource Identifier:
985625558 ( OCLC )
ocn985625558
Classification:
LD1193.L45 2016m S93 ( lcc )

Downloads

This item has the following downloads:


Full Text
NOVEL METHOD OF BIOREMEDIATION AND CHARACTERIZATION OF
BACTERIAL COMMUNITIES ON ARSENIC-IMPACTED MUSEUM
COLLECTIONS
by
SLADJANA SUBOTIC B.S., University of Colorado Denver, 2012
A thesis submitted to the Faculty of the Graduate School of the University of Colorado in partial fulfillment of the requirements for the degree of Master of Science Biology
2016


2016
SLADJANA SUBOTIC ALL RIGHTS RESERVED
11


This thesis for the Master of Science degree by Sladjana Subotic has been approved for the Biology Program by
Timberley M. Roane, Chair Annika C. Mosier
Alan M. Vajda


Subotic, Sladjana (M.S. Biology)
Novel Method of Bioremediation and Characterization of Bacterial Communities on Arsenic-
Impacted Museum Collections
Thesis directed by Associate Professor Timberley M. Roane
ABSTRACT
Metal-based pesticides, including toxic arsenic salts, were widely used prior to the 1970s for the prevention of rodent and insect damage to museum collections. An estimated 80% of collections housed in U.S. and Canadian museums have exogenous metals present, and the use of bacteria in the removal of these metals is of interest. Research shows that Rhodopseudomonaspalustris, a metabolically versatile proteobacterium, is able to volatilize arsenic via methylation resulting in the conversion of the metal into a gas which can be collected and removed. In this study, R. palustris tolerated concentrations of arsenic up to 250 ppm and showed the ability to remove up to 78% of arsenic from a starting concentration of 10 ppm soluble arsenic within 19 days. In order to optimize the potential application of bacterial volatilization of arsenic as a remediation technology for arsenic-treated museum specimens, the presence of in situ bacteria on the surface of museum collections needs to be addressed. High throughput sequencing revealed diverse bacterial communities associated with the museum specimens examined at the Denver Museum of Nature and Science. Bacterial community composition seemed strongly influenced by several factors including type of materials comprising the specimen, the presence of arsenic, and which museum Collection the specimen was housed in. Organisms of high abundance across the items sampled included the genera Ralstonia, Sediminbacterium, Acinetobacter and the family Enterobacteriaceae. Additionally, testing of surface-associated arsenic revealed a wide range of concentrations on arsenic-impacted items at the Denver Museum of Nature and Science.
IV


This study is the first to utilize high throughput sequencing techniques to characterize the
bacterial communities associated with museum collections.
The form and content of this abstract are approved. I recommend its publication.
Approved: Timberley M. Roane
v


ACKNOWLEDGMENTS
There are many people I would like to thank for their role and involvement in this project. First off, thank you to my advisor Dr. Timberley Roane, who allowed me the opportunity to work in her lab, has encouraged me to grow as a scientist, and has been an amazing mentor. One of the best decisions I ever made was to take her microbiology course that opened the door to many amazing opportunities. Thank you to my committee members Dr. Annika Mosier and Dr. Alan Vajda without all of your contribution and expertise I would not be where I am today.
Thank you to the amazing individuals at the Denver Museum of Nature and Science who made a huge part of this project possible. First off, thank you to Richard Busch, education collection manager, who helped us get in touch with many of the other collection managers and supported our project from the beginning. Without Rich, this project would not have been possible. I would also like to acknowledge and thank Jeff Stephenson, zoology collection manager and Melissa Bechhoefer, anthropology collection manager and NAGPRA coordinator who allowed us to work with their collections for the completion of this project. I would like to thank several individuals for their help with sample collection: Anna Nguyen, Helen Dupree, Kelsey Foster, and Nancy Moreno-Huizar. A thank you to Bhargavi Ramanathan for support and help through the writing process.
I would also like to thank Jeffrey Boon and the University of Colorado Shared Analytical Services laboratory for help with arsenic quantification and method development.
Finally, there are a few people that deserve a special thank you and to whom I will always be grateful for their help and support. Joshua Sackett, you have been an amazing friend and colleague from the start of my program and even after you moved away I could always count on you not just for help but for advice and support. Thank you for everything
vi


you have done to help me from data analysis to the numerous conversations of encouragement throughout this process. Adrienne Narrowe, without you I couldnt have gotten through the hardest part of this thesis. Thank you for investing your time and knowledge to help me navigate through the data analysis and for providing me with advice and support. A huge thank you to Dr. Miller for help with data and for allowing the use of his server and resources for data analysis. Munira Lantz and Anna Nguyen, I will always be grateful for the amazing support system you two have been. Thank you!
I would finally like to thank my family my parents, brother and husband who have always supported my goals and never doubted me. I am grateful to have you on my team.
Lastly, I would like to thank the National Park Service, National Center for Preservation Technology and Training (Grant #P13AP00081), for funding this project.
Vll


TABLE OF CONTENTS
CHAPTER
I. PROJECT OVERVIEW AM) OBJECTIVES........................................... 1
II. BACKGROUND: USE OF ARSENIC VOLATILIZATION AND DEVELOPMENT OF
ARSEMC QUANTIFICATION METHODS.................................................3
Overview....................................................................3
Bacterial Volatilization of Arsenic.........................................3
Quantification Methods Available for Arsenic................................7
Rhodopseudomonaspalustris and Arsenic.......................................8
NAGPRA and Museum Pesticides...............................................10
III. CONDUCTED STUDY: APPLICATION OF ARSENIC VOLATILIZATION BY
Rhodopseudomonas palustris...................................................14
Abstract...................................................................14
Introduction...............................................................14
Methods....................................................................16
Confirmation of Arsenic Removal by R palustris...........................16
Development of an Arsenic Application and Detection Method for Simulated Museum Materials............................................................16
Method 1: Quantification of Arsenic From Paper Using Microwave Digestion.17
Method 2: Quantification of Arsenic From Paper Using Acid Digestion....17
Method 3: Quantification of Arsenic from Paper Using Boiling Reflux Digestion .... 18
Graphite Furnace Atomic Absorption Spectrophotometry.....................18
Results....................................................................19
Confirmation of Arsenic Removal by R palustris...........................19
Development of an Arsenic Application and Detection Method for Simulated Museum Materials............................................................22
viii


Method 1 Results: Quantification of Arsenic from Paper Using Microwave Digestion ...................................................................22
Method 2 Results: Quantification of Arsenic from Paper Using Acid Digestion.23
Method 3 Results: Quantification of Arsenic from Paper Using Boiling Reflux Digestion.................................................................24
Discussion...................................................................25
IV. BACKGROUND: MICROBIOLOGY OF THE DENVER MUSEUM OF NATURE
AND SCIENCE.....................................................................26
Overview.....................................................................26
Preservation Methods..........................................................26
Microbiology of Museums.......................................................27
16S rDNA and Sequencing.......................................................28
Denver Museum of Nature and Science...........................................31
V. CONDUCTED STUDY: MICROBIOLOGY OF THE DENVER MUSEUM OF
NATURE AND SCIENCE..............................................................34
Abstract.....................................................................34
Introduction.................................................................34
Methods......................................................................35
Sampling Methodology........................................................35
Bacterial DNA Collection Method.............................................36
Dust Sample Collection....................................................36
Arsenic Detection and Quantification........................................36
Bacterial Community Characterization........................................37
Computational Analyses......................................................39
Results......................................................................39
IX


Presence of Arsenic on Specimens
39
Arsenic Impact on Alpha Diversity...................................41
Collection Impact on Alpha Diversity................................41
Material Impact on Alpha Diversity...................................42
Impact of Arsenic on Bacterial Communities...........................44
Impact of Collection on Bacterial Communities........................47
Impact of Material on Bacterial Communities..........................51
Community Analysis...................................................58
Dust at the Denver Museum of Nature and Science......................59
Discussion.............................................................64
VI. CONCLUSION AND FUTURE DIRECTIONS.....................................69
SUPPLEMENTAL TABLES AM) FIGURES..........................................71
REFERENCES...............................................................79
x


CHAPTER I
PROJECT OVERVIEW AND OBJECTIVES
Prior to the 1970s, in order to preserve museum specimens, chemicals such as naphthalene, paradichlorobenzene, carbon tetrachloride, and metal salts were applied to cultural artifacts, as well as to materials found in herbarium, educational, and zoological collections. These chemical pesticides were used to reduce damage caused by insects and rodents during storage and display. More recently identified is the long-term persistence of metal-based preservatives and, thus, the continued toxicity of treated specimens. Among the metal-based chemicals applied was sodium arsenite (NaAs02), which is linked to cardiovascular diseases, neurological disorders, cancers, and, in some cases, death. The use of pesticides, and specifically arsenic, in the preservation of museum collections became a recognized issue of public concern with the 1990 enactment of the Native American Graves Protection and Repatriation Act, NAGPRA. NAGPRA requires all federally funded agencies to return cultural artifacts to original tribal owners. Early repatriation efforts found that pesticide-treated artifacts could not be returned to tribal owners without health risks due to the toxicity of the used, and still remaining, pesticide concentrations.
Previous work in the Roane lab looked at the use of metal-resistant bacteria in the removal of mercury from museum collections. In this work, the bacterium Cupriavidus metallidurans CH34 was capable of reducing material-associated (e.g. paper and textiles) mercury through volatilization. In working with the Arizona State Museum (Tucson, AZ) and the Smithsonian Institution (Washington D.C.) on mercury-impacted specimens, it was discovered that a variety of collection materials at the Denver Museum of Nature and Science (DMNS) are heavily impacted by arsenic. This project aimed to design the framework for the
1


bacterial removal of arsenic focusing on the development of arsenic quantification methods from a simulated museum material, paper, and to explore the presence of bacteria associated with collection specimens housed at the DMNS. The specific objectives of the work reported here were to:
Objective 1: Begin method development for the removal of arsenic from designated material types using the arsenic-resistant bacterium Rhodoposeudomonaspalustris CGA009 .
Objective 2: Identify the bacterial communities and arsenic presence on Denver Museum of Nature and Science specimens of different materials and storage collections.
2


CHAPTER II
BACKGROUND: USE OF ARSENIC VOLATILIZATION AND DEVELOPMENT OF ARSENIC QUANTIFICATION METHODS Overview
Microorganisms, in particular bacteria, have been studied for their unique abilities to survive in a variety of environments that can be toxic to other forms of life. Not only are they able to survive in these environments, many bacterial species are capable of breaking down toxic chemicals and are being utilized in remediation processes. Among these toxic chemicals are heavy metals, which were used in preservation processes in museums until the 1970s. Rhodopseudomonaspalustris is a bacterium capable of reducing arsenic concentrations through the process of volatilization. The long-term goal of this study is to develop a bacteria-based remediation technology to remove arsenic from impacted museum collections, especially those earmarked for repatriation under NAGPRA (Native American Graves Protection and Repatriation Act).
Bacterial Volatilization of Arsenic
Arsenic compounds, such as arsenite (AsC>3 ', As(III)) and arsenate (AsC>4 ', As(V)), are well-recognized as detrimental to human health. In general, ingesting > lmg of arsenic is considered toxic and can result in chronic renal failure, paralysis, coma and death (Odegaard 2005). At the cellular level, arsenic disrupts ATP production, increases hydrogen peroxide production, and disrupts the function of enzymes (Shen 2013). As(III), in particular, impairs the function of proteins by interacting with their sulfhydryl groups, while As(V) inhibits oxidative phosphorylation (Majumder 2013; Shen 2013). The combination of these events leads to cell death and, on a larger scale, can lead to organ failure.
3


While arsenic is toxic to all biological systems, some bacteria have developed physiological mechanisms of dealing with arsenic toxicity. As(III) typically enters bacterial cells through aquaglyceroporins, while As(V) can enter the cell through phosphate transporters (Figure 2.1). Once arsenic enters the cell, bacteria can use different mechanisms to resist and reduce the metals immediate toxicity. The ars operon consisting of several genes, found on chromosomal or plasmid DNA, code for membrane-associated transport proteins that pump arsenic out of the cell (coded for by the arsB gene); for an arsenate reductase to reduce As(V) to the less toxic As(III) (coded for by the arsC gene); and for S-adenosylmethionine methyltransferase that methylates arsenic (coded for the arsM gene) (Oremland 2003).
As(lll)
As uptake
As(V)
/PH 3- Irancnnrinre\
As(V)

(A(jUQQly/'Arr>ru%, As precipitation
As extrusion
As(lll)
GIpF1
Evaporation
Figure 2.1. A schematic representation of different resistance mechanisms found in bacteria to cope with arsenic (Slyemi 2012).
4


The methylation of arsenic is important to this project as the volatilization of arsenic into gas can decrease arsenic concentrations associated with a material. In the volatilization process, S-adenosylmethionine (SAM) serves as a methyl group donor and the methylation pathway consists of several reduction and methylation steps that produce dimethyl- and trimethyl-arsine gases (Figure 2.2; Bentley 2002; Paez-Espino 2009; Slyemi 2012; Vahter 2002). For remediation purposes, arsine gases can be collected and disposed of effectively reducing the associated concentration of arsenic.
While arsenic has long been considered an easily biomethylated metal/metalloid (Bentely 2002), only a few studies have looked at arsenic volatilization as a remediation technology. One bacterium of particular interest for arsenic removal is Rhodopseudomonas palustris CGA009. The arsenite S-adenosylmethionine methyltransferase enzyme has been well-characterized in R. palustris (Qin et al. 2006; Yuan et al.2008). Qin et. al (2006) showed that when E. coli cells were genetically transformed with the arsM gene from R. palustris and incubated with As(III) for 18 hours, there was a 4,600 pg decrease in the amount of arsenic in the culture. Furthermore, gaseous products were trapped on filters and analyzed showing that the amount of volatilized arsenic correlated with the decreased concentration of arsenic in solution. In another study, the arsM gene from cyanobacteria was cloned into an E. coli vector conferring arsenic resistance to the otherwise non-resistant A. coli up to 100 pM As(III) (Yin 2011). Another study looking at arsenic-contaminated soil showed that genetically engineered bacteria expressing the arsM gene isolated from R. palustris could remove 2.2-4.5% (approximately 42 mg/kg) of arsenic during a 30 day period (Liu et al. 2011). Preliminary studies done in the Roane lab with R. palustris containing the arsM gene, showed bacterial growth in media amended with arsenic concentrations as high as 250 ppm.
5


/
o
II
As(V)
oxidation
\
OH
As(lll)
/ \
SAM
O
OH OH OH
Arsenate
reduction , ^ methylation
OH OH Arsen ite
> As(V)
R-SH
reduction
SAM
O
II
/ I ^
OH CH3OH
Monomethylarsonic acid (MMA)
H
reduction
> As(V) methylation ^ j1 ^
CH3 ch3 OH
Dimethylarsinic acid (DMAA)
As(lll)
/ \
CH,
CH3
Dimethylarsine
(DMA)
CH3
R-SH
SAM

reduction
methylation
As(lll)

/
CH3
\
reduction
CH,
Trimethylarsine
(TMA)
R-SH O
II
As(V)
7 I x CH3 CH3 CH3
Trimethylarsine oxide (TMAO)
Figure 2.2.The process of methylation of As(III) or As(V) by prokaryotic organisms. SAM= S-adenosylmethionine, R-SH= thiol containing compounds (Slyemi 2012).
Bacteria have long been used to mitigate chemical contamination due to their ability to transform chemicals. For example, bacterial remediation has become a cost-effective accepted strategy for the degradation of hydrocarbons, diesel, and naphthalene (Fuentes 2014; Chen 2015; Huang 2016). For metal-based contaminants, the focus of the work presented here, several studies show the potential for bacteria to transform metal-based chemicals. For example, a study done by Dash et al. (2014) showed that Bacillus thuringiensis PW-5, isolated from the Odisha Coast, resisted concentrations of mercury as high as 50 ppm HgCC and volatilized 90% of the mercury. In another study, Cupriavidus metallidurans MSR33 removed 100% of HgCC in solution following the addition of
6


thioglycolate, which provides the thiol groups needed for the methylation (Rojas 2011). In a field-based study with selenium, the ability of in situ soil bacteria to remove selenium from agriculturally-contaminated waters through methylation was examined. When treated with a protein amendment, such as gluten or casein, 68-88% of the total soil selenium was removed via volatilization from the first 15 cm of soil over a period of 100 months (Flury 1997).
Quantification Methods Available for Arsenic
Various methods exist for quantifying arsenic in human samples (i.e. hair, urine, blood) as well as environmental samples such as soil, however, a standard method does not exist for materials encountered in the museum setting (i.e. paper, feather, fur). Numerous analytical instruments can be used for sensitive and accurate determination of arsenic, such as Graphite Furnace Atomic Absorption Spectrophotometry (GFAA), Inductively Coupled Plasma Atomic Emission Spectrometry (ICP-AES), Inductively Coupled Plasma Mass Spectrometry (ICP-MS), Flame Atomic Absorption Spectrophotometry (FAAS), among others, with each instrument having advantages and disadvantages (Allen 1997; Nishimura 2010; Ulusoy 2013).
Required for all analytical quantification methods is the digestion of the material, which is generally done in a solution of one or more acids. For example, microwave digestion is used when quantifying arsenic in food (i.e. rice) and sediments, wherein the sample is dissolved in acid and/or hydrogen peroxide during heating in a microwave (Allen 1997; Nishimura 2010; Pantuzzo 2009; Zhou 1997). The microwave is thought to provide a more even and intense heating than other methods, such as digesting the sample in acid during boiling on a hot plate (Boutakhrit 2004; dos Santos 2013). Additionally, the EPA has methods for digesting biological samples using acids (i.e. nitric acid, hydrochloric acid),
7


combined with other chemicals such as chloroform or methanol (Faroon 2002). This said, there are no studies available addressing the recovery and quantification of arsenic from materials likely to be encountered in the museum setting, including paper, hide, feathers, and textiles. This observation necessitated the development of an arsenic digestion method, part of objective 1 of this project.
Rhodopseudomonas palustris and Arsenic Rhodopseudomonaspalustris CGA009 is a non-pathogenic, facultatively photosynthetic bacterium belonging to the a-Proteobacteria (Larimer 2004). R. palustris was first identified in 1907 by Austrian scientist Hans Molisch and further characterized in 1944 (van Neil 1944). The rod-shaped, Gram negative, purple non-sulfur bacterium is ubiquitous and can be found in abundance in swine waste lagoons, pond water, and marine sediments (Larimer 2004). R palustris has the ability to use different types of metabolisms including photoheterotrophy, photosynthesis, chemoheterotrophy, and chemoautotrophy. R palustris is also capable of metabolizing aerobically and anaerobically using a variety of electron donors, e.g. H2 (Figure 2.3). Under aerobic conditions, R. palustris can utilize chemoautotrophy (use of inorganic electron donors to fix C02 into biomass), especially under low-nutrient conditions.
8


Lignin monomers other organic compounds
Ttnosulfarte. Hj, other inorganic etedron donors
HvO
Aerobic
Ch*moh*t*rotrophic
growth
Ch#m autotrophic growth
Anaerobic
Photoheterotrophic
growth
bgrm monomers, other organic compounds
Photoautotrophlc growth
Ttvos^tate Mj, other morgamr
rtrctmn donors
Figure 2.3. Diagram showing the four metabolisms utilized by R palustris (Larimer 2004).
R. palustris is well known for its metabolic versatility. It has the necessary genes for carbon dioxide fixation, nitrogen fixation, and under anaerobic conditions it can generate energy from light via photosynthesis (Larimer 2004). R palustris also has the ability to degrade organic compounds (i.e. sugars, lignin monomers), major components of plant and animal wastes. R palustris can additionally dehalogenate and degrade chlorinated benzoates and chlorinated fatty acids (Egland 2001; McGrath 1997). The optimal temperature range for R.palustris is 25-37C, and it can survive in a pH range of 5-9 with optimum being neutral pH (6.8-7). The metabolic flexibility of R palustris makes it a good candidate for the arsenic remediation in the museum setting where environmental conditions may be unpredictable and nutrient availability limited.
R. palustris is also commonly studied for its ability to survive in the presence of arsenic (Qin 20066; Yin 2011; Zhao 2015). The presence of the ars operon allows the organism to extrude arsenic out of the cell, carry out oxidation/reduction reactions with arsenic, and volatilize arsenic through methylation. In particular interest in this study is arsenic volatilization for the removal of arsenic.
9


NAGPRA and Museum Pesticides
In the 1980s museums were introduced to a new pest management technique called Integrated Pest Management (IPM). At this time, the widespread use of toxic chemicals as pesticides applied to specimens in museums was being fully discovered and acknowledged. The negative health impacts of these chemicals (ranging from organics to metals) were also being recognized (Odegaard 2005). IPM-based storage methods, use of insect hormones to prevent reproduction, and application of tightly controlled environmental conditions, e.g. temperature or oxygen availability, were adopted as non-chemical alternatives to toxic pesticides. In IPM, collection areas are frequently monitored using sticky traps to identify insect presence and to determine the specific IPM response needed for the appropriate protection of a specimen or entire collection. While IPM has been largely successful in allowing museums to control material deterioration, older chemically-treated specimens which make up a majority of museum collections, present a large public health issue due to the presence of the still toxic originally-applied pesticides.
Metal-based pesticides, in particular, show long-term persistence as metals are not degradable. Chemical forms of arsenic and mercury are found on collections ranging from ethnographic, zoological, herbarium, and even fine and decorative arts (Sirois 2001). Work done by Sirois, at the Analytical Research Laboratory in Ontario, Canada, showed a shocking 81% of the natural history collections from five Canadian museums tested positive for the presence of arsenic. Additionally, cultural items tested at the Arizona State Museum showed arsenic readings ranging from 3.17 pg/cm to as high as 3,132 pg/cm (Odegaard 2005). The previous use and associated toxicity concerns with arsenic are documented in museums throughout the U.S. and Canada, and are expected to exist worldwide.
10


Early methods (prior to the 1970s) of material preservation for collection and storage purposes frequently applied arsenic as a 9% solution of sodium arsenite. Objects were often dipped in or sprayed with the solution (Odegaard 2005). Arsenic was also applied as a powder of sodium arsenite, the presence of which can still be seen on some collections today as a white powder (Figure 2.4). Unfortunately, accurate and consistent records of the use of arsenic, frequency of use, concentrations used, and method of application were not regularly kept and are therefore not readily available for the identification and characterization of arsenic use on museum specimens. This has forced the need for non-destructive quantitative methods for arsenic detection on collection materials. A frequently used, despite a number of limitations, technology for the detection of arsenic on museum specimens is X-ray fluorescence (XRF). While XRF can accurately detect the presence of arsenic, its ability to quantify arsenic concentrations and/or identify forms of arsenic associated with materials from wood to pottery to textiles remains in question. XRF is additionally expensive and not readily available in most museums. Because of this, most museums have to assume any item collected prior to the 1970s is impacted by toxic chemical preservatives, including arsenic.
11


Figure 2.4. Arsenic in the form of a white powder, sodium arsenite, seen between feathers of a headdress (Odegaard 2005).
One of the impacted collections of most concern, as having been frequently treated with arsenic salts, is ethnographic (cultural) specimens. In 1990, Congress enacted the Native American Graves Protection and Repatriation Act (NAGPRA), which requires the return of cultural items to lineal descendants and affiliated Indian tribes. These items include human remains, funerary objects, sacred objects, and objects of cultural patrimony. Sacred objects are defined as specific ceremonial objects which are needed by traditional Native American religious leaders for the practice of traditional Native American religions by their present day adherents (Odegaard 2005). Mr. Leigh Kuwanwisiwma, from the Hopi Cultural Preservation Office, was unaware of the toxicity of sacred objects until 1995 when he was working on the repatriation of Katsina Friends, a Hopi spiritual artifact. Kuwanwisiwma says these objects have life and spirit. They are just like your son, your mother. Its part of our living human community that has been contaminated with poison (Kraker 2001). The Hopi, the first tribe to physically repatriate items, such as Katsina Friends, were initially unaware of
12


the contamination and potential toxicity risk. The Hopi repatriated approximately sixty items which were returned to cultural use in traditional ceremonies. As several tribal leaders became ill, tribes were told of the presence of chemical pesticides and tribal leaders were confirmed as having mercury and arsenic poisoning. A moratorium was placed on the physical return of items to tribes across the U.S. and Canada until the items are tested clean.
Current approaches for the removal of pesticides, such as arsenic, include washing, vacuuming, UV light degradation, and containment; however, most cleaning methods show limited effectiveness (Spencer 2000). For example, the use of compressed air to blow out chemicals resulted in a maximum removal of 40% of the pesticides (Glastrup 2001). Methods, such as washing, cause deterioration of fragile artifacts. Light can cause fading in dyes or changes in the materials which are often not acceptable for conservation. Vacuuming seemed initially promising but much like the compressed air method, pesticides remained on the objects even after vacuuming. In addition, vacuuming could be destructive. Another proposed method specific to mercury and arsenic was the application of alpha lipoic acid (ALA) for mercury-treated and dihydrolipoic acid (DHLA) for arsenic-treated materials (Cross 2010). While the use of ALA and DHLA showed potential, their use requires the immersion of objects in liquid, resulting in possible deterioration and damage. Physical containment has also been proposed for treated collections, but is not realistic for item re-use under NAGPRA. Finally, specific to cultural collections is consideration of tribal beliefs, as certain methods, such as exposing spiritual items to heat or UV light, may not be acceptable.
13


CHAPTER III
CONDUCTED STUDY: APPLICATION OF ARSENIC VOLATILIZATION BY
Rhodopseudomonas palustris Abstract
Metal-based pesticides, including toxic arsenic salts, were widely used prior to the 1970s for the prevention of rodent and insect damage to museum collections. An estimated 80% of collections housed in U.S. and Canadian museums have exogenous metals present, and the use of bacteria in the removal of these metals is of interest. Research shows that Rhodopseudomonas palustris, a metabolically versatile proteobacterium, is able to volatilize arsenic via methylation resulting in the conversion of the metal into a gas which can be collected and removed. In this study, R. palustris tolerated concentrations of arsenic up to 250 ppm and showed the ability to remove up to 78% of arsenic from a starting concentration of 10 ppm soluble arsenic within 19 days. To show that R. palustris could be as effective at reducing arsenic concentrations from simulated museum materials, a method for arsenic quantification from collection materials was developed. Boiling reflux digestion showed the highest recovery of arsenic at greater than 80%, while other methods, e.g. microwave digestion, showed less than 10% recovery on average.
Introduction
Arsenic is a highly toxic metal/metalloid historically used as a preservative for ethnographic and zoological collections. Still present today on historically-treated artifacts, arsenic continues to pose a health risk. Commonly applied as sprays, pastes, dips and powders, arsenic is a non-degradable persistent pesticide (Odegaard 2005). While arsenic effectively protects materials from microbial, insect and rodent damage, with the recognition
14


of its human toxicity in the 1970s, its use as a museum preservative has been replaced with safer integrated pest management practices.
Current mitigation for arsenic on museum collections includes vacuuming and washing (Spencer 2000). These methods, however, are not widely used because of their harsh impact on the material; their cultural insensitivity; and/or their ineffectiveness on a broad range of material types (Spencer 2000). The use of bacterial arsenic removal methods is being actively investigated as a possible remediation technology. The use of bacteria has several advantages, including their potential for controlled growth; their ability to facilitate volatilization and physical removal of arsenic; and their cultural acceptability (Roane 2010). Presented in this study is the early identification and characterization of the bacterium Rhodopseudomonaspalustris in the removal of arsenic from arsenic-amended solutions. Removal of arsenic from solution was used to demonstrate early potential for the bacteriums ability to remove arsenic from materials such as paper and textiles.
R. palustris possesses the arsM gene, providing the organism with the ability to volatilize arsenic. The volatilization of arsenic, which converts the metal into a gaseous form, has several advantages from a remediation standpoint. Gaseous arsenic can be collected and disposed of properly, and provides an avenue for the reduction of material-associated arsenic concentrations, especially important given the non-degradability of metals such as arsenic. Reduction of the arsenic concentrations associated with specimens throughout the museum will provide for safer handling. Driven by the toxicity of arsenic initially identified in the repatriation of tribal artifacts to the Hopi Tribe under NAGPRA (Native American Graves Protection and Repatriation Act, enacted in 1990), active research is looking into the development of an effective, broadly acceptable, method for arsenic remediation in this
15


unique setting. The objectives of the work presented here were (1) to characterize the initial ability of R. palustris to volatilize and remove arsenic, and (2) to develop an analytical method for arsenic quantification, previously non-existent for materials found in the museum environment, to provide the framework for future studies addressing the ability of R. palustris to remove arsenic from museum materials.
Methods
Confirmation of Arsenic Removal by R. palustris
In order to determine the arsenic removal capabilities of Rhodopseudomonas palustris, initial studies focused on testing the removal from arsenic-amended tryptic soy broth. Tryptic soy broth was amended with 10 ppm arsenic (17.3 ppm sodium arsenite, As(III)), and inoculated with >105 R. palustris cells. No inoculum, arsenic-amended control media were used to monitor sources of abiotic loss of arsenic. Incubation on a rotary shaker (150 RPM) at 25C was performed. Broth samples were prepared for arsenic quantitative analysis by diluting to a 40 ppb concentration in a 2.5% nitric acid solution to a total volume of 10 mL, and then read on a Graphite Furnace Atomic Absorption Spectrophotometer (GFAA, PerkinElmer, Waltham, MA; described below).
Development of an Arsenic Application and Detection Method for Simulated Museum Materials
To develop a reproducible method for arsenic recovery and quantification from arsenic-treated museum-mimicking materials (e.g. paper), several methods of material digestion were evaluated.
16


Method 1: Quantification of Arsenic From Paper Using Microwave Digestion
In order to determine the concentration of arsenic associated with a museum material (i.e. filter paper as used in this project), it was necessary to first develop an effective method of digesting the material for the GFAA. Initial work looked at a method utilizing a microwave digestion system using nitric acid, a method previously used in the Roane lab for mercury quantification from filter paper and textiles. For the included arsenic work, filter paper was dipped in a solution of 10 ppm arsenic (17.3 ppm sodium arsenite) and allowed to air dry. The amended paper was then placed in a microwave safe Teflon container and 10 mL of concentrated nitric acid was added. The microwave (Remote Microwave Systems, Floyd Inc., Lake Wylie, SC) was set to 100% power for 2 minutes and 20 seconds. Samples were diluted as needed for GFAA analysis and arsenic quantification.
Because arsenic loss was detected above, power and time on the microwave were adjusted to optimize arsenic recovery. Parameters included 25% power for 2 minutes 20 seconds and 25% power for 1 minute 10 seconds.
Method 2: Quantification of Arsenic From Paper Using Acid Digestion
Instead of dipping filter paper into a 10 ppm arsenic solution, 1 mL of the solution was pipetted onto the paper and allowed to air dry. In order to digest the paper prior to GFAA analysis, 10 mL of concentrated nitric acid was placed in a 50mL test tube and placed on a vortexor for 10 minutes at 14,000 RPM. Following vortexing, the tubes were centrifuged for 2 minutes and diluted for analysis on the GFAA. Additionally, 25 ppm and 50 ppm arsenic solutions were applied to filter paper to test whether increasing the initial arsenic concentration increased the amount of arsenic recovered.
17


Method 3: Quantification of Arsenic from Paper Using Boiling Reflux Digestion
A known concentration of an arsenic standard was applied to Whatman 3 filter papers by pipetting 1 mL directly to the surface of the paper. The filter paper was transferred to a 100 mL test tube and 7.5 mL of concentrated nitric acid was added in addition to a boiling chip. The test tube was placed on a hot plate and the temperature kept at ~95C. The test tube was covered with a glass bulb and was allowed to boil anywhere between 3-10 hours. MiliQ water was added to the test tube to bring the volume up to 100 mL. This diluted the amount of arsenic to fit on the standard curve of the GFAA.
Graphite Furnace Atomic Absorption Spectrophotometry
Prior to analysis, control flasks were created with known concentrations of arsenic in order to create a standard calibration curve to be used for calculating arsenic concentrations in sample flasks. An arsenic standard (10,000 ppm arsenic) was used and diluted to seven concentrations (0 ppb (100 mL of 2.5% nitric acid), 5 ppb, 10 ppb, 25 ppb, 50 ppb, 75 ppb, and 100 ppb) to be used for the creation of a standard curve and the calculation of arsenic concentrations in the samples. Calculated concentrations of arsenic were added to a solution of 2.5% nitric acid and a total volume of 100 mL.
GFAA uses a graphite-coated tube, known as a cuvette, to vaporize the sample to allow for the absorption of light at wavelengths characteristic of the element of interest (195 nm for arsenic). Sample (20 pL) is deposited onto the cuvette, which gets heated to 2,500C in order to vaporize and atomize the sample. The atoms then absorb the light and the instrument provides a numerical value that can be used to create a standard curve for calibration and sample determinations.
18


Results
Confirmation of Arsenic Removal by R. palustris
Rhodopseudomonaspalustris was able to grow in tryptic soy broth amended with arsenic concentrations as high as 250 ppm. To assess removal of arsenic, flasks were incubated on a rotary shaker at 25C for 7 days, 10 days, 15 days, and 19 days to determine whether arsenic removal increased with time (Table 3.1).
Table 3.1. Arsenic concentrations remaining in individual flasks with (1)7?. palustris in medium only, (2) 10 ppm arsenic-amended medium, (3) R. palustris with 10 ppm arsenic in medium. The standard deviation is in parentheses and the mean is in ppm of arsenic. BDL= below detectable limit (<5 ppb).
Flask Arsenic detected ( 7 days) Arsenic detected (10 days) Arsenic detected (14 days) Arsenic detected (15 days) Arsenic detected (19 days)
R. palustris only BDL BDL BDL BDL BDL
Arsenic only 4.72(1.73) 6.72(.82) 8.7 3 (. 71) 9.97(.53) 132(59)
R. palustris & Arsenic 2.81(.53) 1.74(1.41) BDL 4.73(1.71) 1.58(1.1)
% Removal 40% 74% 100% 53% 78%
After 7 days of incubation, arsenic concentrations in treatment flasks showed 40% removal of arsenic (Table 3.1). Following 10 days of incubation, the removal of arsenic increased to 74%. The 15 day flasks showed 53% removal of arsenic while the 19 day incubation showed the highest removal at 78% (Figure 3.1). One study incubated for 14 days showed 100% removal of arsenic (Table 3.1). Arsenic-only control flasks contained an average of 8.73 ppm arsenic while treatment flasks showed 0 ppm arsenic present upon analysis with GFAA for this particular 14 day experiment (Table 3.1). It is important to note
19


that each experiment was done independently, rather than one set of flasks quantified at different time periods, so analytical variation is to be expected.
20


Arsenic Removal by R. palustris
B
Control Treatment
Control vs Treatment
a CO
s
<0

u i o
D
I
5
I
|
S
fl
O
u
<
Control Treatment
Control vs treatment
Figure 3.1. Arsenic removal showing control flasks (arsenic only) and treatment flasks (arsenic and R. palustris in broth). (A) Removal after 7 days. (B) Removal after 10 days. (C) Removal after 15 days. (D) Removal after 19 days. 14 day data is not shown here.


Development of an Arsenic Application and Detection Method for Simulated Museum Materials
While R. palustris showed potential for arsenic removal from broth, the ability of R. palustris to remove arsenic from solid material such as those likely to be encountered in the museum setting needed to be demonstrated. To do this, however, a new arsenic quantification and digestion method needed to be developed. The percent loss for arsenic was calculated based on the expected calculation in the digestate, which included the paper digested in 10 mL of concentrated nitric acid.
Method 1 Results: Quantification of Arsenic from Paper Using Microwave Digestion
When the microwave (Remote Microwave Systems) was set to 100% power for 2 minutes and 20 seconds results showed that no arsenic remained on the paper, while the expected concentration was 1 ppm in the digestate solution. When the power on the microwave was reduced to 50% with a digestion time of 2 minutes and 20 seconds, results continued to show a 100% loss or zero recovery of the arsenic (Table 3.2). Power and time on the microwave were adjusted (25% for 2 minutes 20 seconds and 25% for 1 minute 10 seconds) and evaluated. While both times showed higher recovery for arsenic, the detected arsenic levels were less than those predicted and arsenic loss was between 90% and 100% (Table 3.2). Consistency was also problematic with each of the method modifications examined as can be seen by the high standard deviations (Table 3.2).
22


Table 3.2. Summary of mean arsenic detection for paper swatches under different microwave digestion parameters examined. Values reported are mean standard deviation.
Arsenic expected 100% power 2 min 20 sec 50% power 2 min 20 sec 25% power 2 min 20 sec 25% power 1 min 20 sec
1 ppm 0.022ppm (0.037) BDL 0.07 ppm (0.111) 0.039ppm (0.074)
% Loss 97.8% 100% 93% 96.1%
Method 2 Results: Quantification of Arsenic from Paper Using Acid Digestion
An acid digestion method was tested in order to determine if higher amounts of arsenic could be recovered from the material. This method did not show better recovery as the average amount quantified was 0.094 ppm with a standard deviation of 0.020 ppm even though 1 ppm was expected in the digestate solution and 90.6% loss of arsenic was observed (Table 3.3).
The applied concentration of arsenic was increased to 25 ppm while all other parameters remained the same, and the expected concentration was 2.5 ppm. This increased the concentration recovered from the paper, however, there was still a large amount of inconsistency from one experiment to the next, as seen by the large standard deviation and 50% loss of arsenic was observed (Table 3.3). When the applied concentration was increased to 50 ppm (an expected concentration in the digestate of 5 ppm), the observed concentration did not generally increase from what was observed with 25 ppm and loss increased to 90.9% (Table 3.3).
23


Table 3.3. Summary of mean arsenic detection for paper swatches when various concentrations of arsenic were applied and digested in acid by vortexing. Values reported are mean standard deviation.
Measured Arsenic Measured Arsenic Measured Arsenic
Arsenic Expected 1 ppm 2.5 ppm 5 ppm
Arsenic mean 0.094 ppm (0.020) 1.25 ppm (0.76) 0.455 ppm(0.135)
% Loss 90.6% 50% 90.9%
Method 3 Results: Quantification of Arsenic from Paper Using Boiling Reflux Digestion
Digestion by boiling was the final method tried that showed better consistency from one sample to the next as well as concentrations that matched those applied to the material (filter paper). The standard deviations were smaller than those observed with the other digestion methods and percent loss of arsenic was 20% or less (Table 3.4).
Table 3.4. Summary of arsenic detection for paper swatches when various concentrations of arsenic were applied and digested by boiling reflux digestion. Values reported are mean standard deviation.
Measured Measured Measured Measured
Arsenic Arsenic Arsenic Arsenic
Arsenic Expected 0.030 ppm 0.045 ppm 0.060 ppm 0.075 ppm
Arsenic 0.0240 ppm 0.039 ppm 0.055 ppm 0.065 ppm
Mean (0.00544) (0.0041) (0.00198) (0.0019)
% Loss 20.4% 13.1% 8.3% 12.9%
24


Discussion
Experiments in broth showed that R palustris could remove 100% of arsenic within 14 days at 25C. Showing that R palustris had potential as a remediation organism, the study shifted to applying arsenic to simulated museum materials (i.e. materials that museum collections are composed of). The work here attempted to develop a standard method for the quantification of arsenic associated with materials such as paper, fur, and feather. The method of acid digestion via boiling was the most effective of those tried at recovering the mathematically predicted amount of arsenic. Future work will include testing the developed method on other material types, e.g. textiles, for recovery of expected arsenic concentrations for eventual use for monitoring the ability of R. palustris to remove arsenic from museum specimens.
Because of the limitations of current arsenic removal methods, and the resulting harsh treatment of materials, arsenic removal from museum specimens is not widely performed. Additionally, for cultural items earmarked for repatriation under NAGPRA, many tribes find the currently available methods culturally unacceptable. A bacterial-based approach has been accepted by many tribal leaders as it is perceived as a more natural method to heal artifacts that have been poisoned by chemicals (in particular metals). While much work remains to be done, Rhodopseudomonaspalustris is an organism for consideration for arsenic removal from museum specimens.
25


CHAPTER IV
BACKGROUND: MICROBIOLOGY OF THE DENVER MUSEUM OF NATURE
AND SCIENCE Overview
Studies characterizing microorganisms in the museum setting are heavily focused on organisms that contribute to degradation of material, and, to date, a wide scale characterization of bacterial communities associated with museum materials has not been done. This study aimed to elucidate these bacterial communities associated with different material types housed in different Collections Departments within the museum using high-throughput sequencing, a technique that has not been utilized in museum studies. Additionally, the use of arsenic in past preservation methods has caused its persistence on individual collections items, another potential variable in shaping the resulting microbial communities. Knowledge of existing microorganisms on museum specimens will contribute to our understanding of material degradation but will also help shape the use of bacteria in pesticide removal for the safer handling of collections and for the repatriation of chemically-treated artifacts under NAGPRA.
Preservation Methods
Taxidermy is the process of preparing, stuffing, and mounting the skins of animals for display. Through the 1970s a common preservative used was arsenic, usually in the form of sodium arsenite. For taxidermy applications, the arsenic powder was dissolved by boiling in a solution of water and alcohol (added in unknown amounts). In addition, arsenic powder was combined with a flour paste, among other chemicals, to create a papier-mache mixture that was used as an adhesive throughout the preparation process. To prepare bird or mammal
26


mounts, the first step included skinning the animal and washing the skin in a solution of ammonia water, followed by alcohol, and then in a solution of water and benzene (Pray 1923). Following drying, arsenic solutions were applied to the inner and outer surfaces of the skin depending on the specimen, in the form of liquid or powder. The skin was then stuffed with wire and wood wool and mounted for display (Pray 1923).
Microbiology of Museums
Work done surrounding the microbiology of the museum setting is heavily focused on organisms that contribute to biodeterioration and, in particular, the deterioration associated with important historical documents. These documents are often parchment (collagen) or paper (cellulose) making them targets for microbial degradation (Lech 2016). The fungi are common degraders, including members of the genera Aspergillus, Penicillium, Cladosporium and Trichoderma (Lech 2016; Sterflinger 2012). An assessment of parchment paper documents in Krakow Poland revealed the presence of these common fungi as well as various bacterial strains including Bacdlus spp., Micrococcus spp., Staphylococcus spp., and Pseudomonas spp.
Other studies looking at wood, textiles, and stone (i.e. fossils) revealed the presence of various bacterial and fungal species. A study looking at wood at a museum in Auschwitz showed that fungi such as Aspergillus, Penicillium and Cladosporium were present, as were various bacterial species, including Bacillus spp. (Kozirog 2014). Textiles, as well as stone, sampled across museums in India and Poland also detected common fungal deterioration culprits, such as Aspergillus, as well as bacteriaMicroccous spp., Staphyloccocus spp., Pseudomonas spp., Bacillus spp., and Actinobacteria (Biswas 2013; Blyskal 2014;
Sterflinger 2013).
27


Other studies did an overall characterization of microorganisms found in the museum setting by obtaining samples from the air. Interestingly, these studies found most of the same organisms as those found on the surfaces of wood, textiles, and stone with the addition of Acinetobacter spp. and Lactobacillus spp. (Gauzere 2014; Skora2015).
16S rDNA and Sequencing
Much of the work done characterizing the microbiology of the museum setting has relied on techniques like culturing, Denaturing Gradient Gel Electrophoresis (DGGE), and 454 Sequencing. The current availability of high-throughput sequencing allows for a more comprehensive analysis of the microorganisms present on museum specimens. High-throughput sequencing targets the 16S rDNA gene which is highly conserved across bacteria and is commonly used for constructing phylogenies and identifying bacterial communities in a wide variety of environments. While the gene is found in all bacterial species, it consists of variable regions specific to genus and species allowing for identification upon sequencing (Figure 4.1; Clarridge 2004).
28


VI V? V3 V4 V5 VO I VO VO
!/*)
YlrZ,
CZ..JZ-ZI__. .3
l*. .4H A,
" H n!^
_^y r m :
at -vo
"> (1 /Os-
^ ^ fora
X^.j $yP J is/^r
cf^3 i^P-h=r t-'.v
|£fe* fi^sw
1J
cti -jj-
Jj
-*> 7
-QS.' r
Jf
//'sit
O' s /
1^
Figure 4.1. Folding pattern of the 16S rRNA gene and the conserved (green) and variable (V) regions. Image adapted from: http://rna.ucsc.edu/rnacenter/xma/xma_gallery.html and http://www.al imetrics.net/en/index.php/dna-sequence-analysis)
Illumina DNA sequencing technology allows for the consecutive sequencing of multiple DNA molecules in a mixed sample (Caporaso et al. 2011). The hypervariable V4 region of the 16S rDNA is amplified using a forward primer, annealing to a conserved region of the gene, and a reverse primer containing a 12-nucleotide barcode, a 3 complement of the Illumina adaptor and reverse primer pad and linker (Figure 4.2). Each DNA molecule in a sample library (a library is a sample containing multiple sequences) requires the use of a different barcode on the reverse primer, allowing for the separation and reading of multiple DNA sequences.
29


V4
P5 SBSF {515F) | 46 bp overlap -| 150 bp index 806R index

150 bp 254 bp -------
SBSR
(806R)
P7
Figure 4.2. V4 region is amplified using F515 and 806R primers. Paired 150 base pair sequencing gives a 254 bp fragment with a 46 bp overlap. Figure adapted from Illumina application note.
The library is prepared for loading on the MiSeq Sequencing System and a process called bridge amplifications of the DNA strands results in the formation of millions of unique clusters with thousands of copies of one amplicon (Figure 4.3). Following this step, DNA templates are copied using fluorescently-labeled nucleotides which are then interpreted by a detector as specific nucleotides (Figure 4.4; The Illumina HiSeq 2000 Sequencing Technology 2015).
?. iiliili iltiiui ill P
:

Figure 4.3. Cluster generation, the first step of sequencing on the Illumina MiSeq flow cell. Source: Illumina Inc.
30


Figure 4.4. Clusters on the flow cell are sequenced simultaneously using sequencing-by-synthesis technology. Fluorescently-labeled nucleotides are incorporated into the synthesis one by one and a detector interprets the wavelength of each fluorophore as a specific nucleotide.
Sequencing data generated from high-throughput 16S rDNA sequencing can be analyzed with QIIME (Quantitative Insights Into Microbial Ecology) and R (Caporaso et al. 2010; R Core Team 2014; Oksanen et al. 2015). QIIME allows for the filtering of sequencing reads and assigning taxonomy. Additionally, it can be used to perform a variety of phylogenetic and statistical analyses including diversity determinations within samples (alpha diversity) and diversity among samples (beta diversity).
Denver Museum of Nature and Science The Denver Museum of Nature and Science, established in 1900, is located in Denver, Colorado, and currently houses 1.4 million items across 7 Collections/Departments. These include Anthropology, Library and Archives, Education, Geology, Health Sciences, Paleontology and Zoology. The items in each Collection are stored in rooms where humidity is not controlled and varies depending on the relative humidity of the outside air.
Temperature of the Collections areas are kept at a standard room temperature of approximately 25C. In 2014, a new portion of the building was completed and many of the
31


Departments moved to these new storage areas where humidity and temperature can now be monitored.
Figure 4.5. Hide sampled from Zoology and Education Collections. (A) Item Z2717, a jaguar hide from Zoology. (B) Item PB6183, coyote hide from Education.
This study included 150 samples (individual swabs) from specimens (bird mounts, animal mounts, hides) housed in Zoology and Education Collections and 13 dust samples (Figure 4.5; Figure 4.6).
32


Figure 4.6. Bird and animal mounts sampled in the Education and Zoology Collections. (A) Item PB0248 from Education. (B) Item ZB 1078, a woodpecker from Zoology. (C) Item PB4275, a red fox from Education. (D) Item ZM2480, a howler monkey from Zoology.
33


CHAPTER V
CONDUCTED STUDY: MICROBIOLOGY OF THE DENVER MUSEUM OF
NATURE AND SCIENCE Abstract
High throughput sequencing revealed diverse bacterial communities associated with the museum specimens examined at the Denver Museum of Nature and Science. Bacterial community composition seemed strongly influenced by several factors including type of materials comprising the specimen, the presence of arsenic, and which museum Collection the specimen was housed in. Organisms of high abundance across the items sampled included the genera Ralstonia, Sediminbacterium, Acinetobacter and the family Enterobacteriaceae. Additionally, testing of surface-associated arsenic revealed a wide range of concentrations on arsenic-impacted items at the Denver Museum of Nature and Science. This study is the first to utilize high throughput sequencing techniques to characterize the bacterial communities associated with museum collections.
Introduction
Microorganisms play an important historic and current role in artifact quality and longevity. While known to participate in the degradation of materials, a comprehensive study of the microbial communities present on collection items throughout the museum setting is not available. Studies show the presence of fungal (i.e. Aspergillus, Cladosporium) and bacterial {Bacillus, Staphylococcus, Pseudomonas) species; however, few studies have looked at bacteria associated with different types of specimens and hypothesized driving factors that shape these communities. The aim of this study was to use high-throughput 16S rDNA sequencing techniques in a diversity study of the bacterial communities present on
34


zoological specimens at the Denver Museum of Nature and Science (Denver, CO). The Denver Museum of Nature and Science houses 1.4 million specimens across 7 Departments, with many of the specimens arsenic-treated, providing a unique opportunity to evaluate how early arsenic preservation treatment has shaped the resulting bacterial communities. This study aimed to characterize not only the bacteria found on these specimens, but to also determine which factors have the greatest influence on surface-associated bacterial diversity. Several factors were investigated, including the presence of arsenic, material type, and the museum Collection/Department the specimen belonged to. The associated museum Collection can be important because often specimens belonging to different Collections are stored in different ways and frequency of handling of items can vary. Elucidating the communities that exist on museum specimen collections could answer questions about effectiveness of preservation and storage methods. This study, in the long term, will contribute to the development of bacterial-based remediation methods for the removal of early-used toxic chemical preservatives on some museum specimens.
Methods
Sampling Methodology
In order to compare the influence of arsenic presence, material type, and Collection storage/handling on bacterial communities, individual items were chosen that were arsenic-impacted and non-impacted, represented different material types (hides, animal (non-bird) mounts, bird mounts), and were from different Collections Departments (Zoology and Education) within the Denver Museum of Nature and Science. Table 5.1 summarizes the number of items within each Collection of different material types as well as the range of arsenic detected on these items. Arsenic testing (described below) was performed for each
35


bacterial sample collected. Overall, 150 individual samples for DNA extraction were collected from specimens across the two Collections and 13 dust samples throughout the museum.
Bacterial DNA Collection Method
To collect bacterial samples from museum specimens, Whatman 42 filter paper, marked into four parts, was dipped into a solution of 0.1% glycerophosphate buffer. The damp paper was rubbed on the item in an area covering approximately 7 cm. The filter paper was immediately cut and a fourth of the paper was placed in 50 mL 0.1% glycerophosphate buffer for arsenic testing while the remaining paper was placed in a petri plate, bound with parafilm, and placed in a cooler with dry ice. Each specimen was sampled in 2-3 different locations to assess bacterial diversity differences within individual specimens.
Dust Sample Collection
Dust samples across the museum were also collected for bacterial and arsenic analyses. Thirteen dust samples were collected representing both public spaces and collection storage spaces. Dust was collected by wiping the top of door frames with damp Whatman 42 filter paper (as above) in each of thirteen spaces.
Arsenic Detection and Quantification
As described above, a Whatman 42 filter paper was wiped along the sampling surface to collect bacterial DNA as well as surface-associated arsenic residues. To wash the arsenic off the sample paper, the sample paper was then suspended in 50 mL 0.1% glycerophosphate. The one fourth filter paper and buffer were vortexed for 5 minutes using the MO BIO Vortex adapter for 50 mL tubes (MO BIO Laboratories, Inc., Carlsbad, CA). Following vortexing, the solution was placed in a bottle and approximately 200 mL of MiliQ water was added.
36


Arsenic testing was performed on site upon collection using the LaMotte Arsenic Test Kit (LaMotte Company, Chestertown, MD). The LaMotte arsenic test kit instructions were followed to determine the arsenic concentration in the solution. The LaMotte testing kit can detect arsenic concentrations from 4 ppb to 2,000 ppb, and is used in a variety of applications for the quantification of aqueous arsenic.
Bacterial Community Characterization
In the laboratory, DNA was extracted from the sample paper using the MO BIO PowerMax Soil DNA Isolation Kit following manufacturers protocol (MO BIO Laboratories, Inc., Carlsbad, CA). Following extraction, DNA was concentrated following the kit instructions with the modifications below. Sodium chloride (5M, 250 pL) was added to each sample and mixed by inverting. Ethanol (200 proof, 10.4 milliliters) was added to each sample, mixed by inverting and left on ice for one hour. Samples were centrifuged for 30 minutes at 0C and supernatant was removed. Approximately 20 mL of 70% ethanol was added and samples were centrifuged for 3 minutes at 4C. Supernatant was removed and samples were left in a hood overnight. Following drying, 100 pL of sterile molecular biology grade water (Thermo Fisher Scientific, Inc., Waltham, MA) was added to resuspend the resulting DNA pellet. Polymerase chain reaction (PCR) was performed using primers to target the V4 variable region of the 16S rRNA gene. The primers included F515NS4 (5' AAT GAT ACG GCG ACC ACC GAG ATC TAC ACT ATG GTA ATT GTG TGY CAG CMG CCG CGG TAA 3) and 806R (5 CAA GCA GAA GAC GGC ATA CGA GAT XXXXXXXXXXXX AGT CAG TCA GCC GGA CTA CHV GGG TWT CTA AT 3) each containing 5 overhangs necessary for Illumina high-throughput sequencing. The reverse primer contained a 12-nucleotide Golay barcode, represented by X, that was specific for each
37


sample allowing for multiplexing during the sequencing process (Caporaso et al. 2012). The PCR reaction mixture contained 10 pL 5 PRIME HOT Master Mix (5 Prime, Gaithersburg, MD) [final reaction concentrations 0.5U Taq DNA polymerase; 22.5 mM KC1; 1.25 mM Mg2+; and 100 pM of each dNTP]; 200 nmol/L of F515NS4; 200 nmol/L of 806R; 200 ng of bovine serum albumen (BSA) (New England BioLabs, Inc., Ipswich, MA); molecular biology grade water (Thermo Fisher Scientific, Inc., Waltham, MA); and template DNA to a total volume of 25 pL. The thermocycler protocol included a 3 minute denaturation at 94C; 30 cycles of a 45 second denaturation at 94C; a 60 second annealing step at 50C; and a 90 second extension step at 72C; followed by a ten minute final extension step at 72C; and storage at 4C.
To ensure successful amplification and the correct product size, agarose gel electrophoresis was performed using a 1% agarose gel stained with ethidium bromide. Once amplification was confirmed, two replicate PCR reactions were performed for concentration and purification with the Zymo DNA Clean and Concentrator Kit (Zymo Research, Irvine, CA), following the included protocol. The cleaned and concentrated DNA was quantified using the Qubit High Sensitivity dsDNA Assay Kit and Qubit 3.0 Fluorometer (Thermo Fisher Scientific, Inc., Waltham, MA). The triplicate PCR reactions for each sample were pooled in equimolar ratios to achieve 20 nanograms of DNA. Pooled samples were concentrated and re-quantified again to obtain a final library concentration in ng/pL.
Libraries were sent to the Genomics and Microarray Core at the University of Colorado Denver Anschutz Medical Campus (Aurora, CO) for Illumina MiSeq 2x150 paired-end sequencing.
38


Computational Analyses
Phylogenetic analyses were performed with QIIME (Quantitative Insights Into Microbial Ecology) (Caporaso et al. 2010). Paired end sequences were joined using the join_paired_ends.py command, which uses the fastq-join method and a specified minimum overlap score of 30 was given (Aronesty 2011). Joined sequences were filtered (split libraries fastq.py command) and sequences with a Phred quality score of 29 were kept, which corresponds to a sequencing error rate of 0.1%. Operational taxonomic units (OTUs) were picked using the pick open reference.py where reads were clustered together based on 97% or greater sequence similarity. This command also assigned taxonomy to OTUs using the August 2013 Greengenes bacterial and archaeal 16S rRNA database (DeSantis 2006). Chimeric sequences were identified using DECIPHER Find Chimeras Web tool and were removed from the dataset prior to further analysis (Wright 2012). Alpha diversity was performed in QIIME using the alpha diversity.py command. R was used for the generation of heat maps and the phyloseq package in R was used for the generation of PCoA plots and statistical testing (McMurdie 2015).
Results
Presence of Arsenic on Specimens
Table 5.1 summarizes the types of specimens sampled and the corresponding arsenic concentrations. Each of the Educational animal mounts and hides tested below the detectable limit (4 ppb) for arsenic with the exception of specimen PB5758A, which was the top of the back of a red wolf hide testing at 60 ppb (Supplemental Table SI). Each of the 15 bird mount samples collected from the Education Collection showed arsenic concentrations ranging from 20 to 2,000 ppb (the maximum detectable concentration (Supplemental Table S2).
39


From the Zoology Collection, animal mounts and hides, with and without arsenic, were examined. Arsenic ranges seen were from 20 1,000 ppb for animal mounts, while hides ranged from 20-80 ppb (Supplemental Table S3). Additionally, 5 animal mounts and 14 hides from the Zoology Collection tested negative for the presence of arsenic (Supplemental Table S3). Bird mount samples in the Zoology Collection where arsenic was detected ranged from 50 ppb to over 2,000 ppb (Supplemental Table S4). The bird mount samples from the Education Collection in addition to the arsenic-impacted Zoology samples showed the highest arsenic concentrations of all the items tested. Additionally, 27 bird mount samples were taken from the Zoology Collection where arsenic was not detected.
Table 5.1. Summary of specimens sampled and the corresponding arsenic ranges. BDL= below detectable limit (4 ppb).
Associated Collection/Department Material type sampled Arsenic Detection
Education 5 animal mount BDL
14 hide BDL
1 hide Detected
15 bird mount Detected
Zoology 15 animal mount BDL
4 hide BDL
23 animal mount Detected
15 hide Detected
27 bird mount BDL
31 bird mount Detected
40


Arsenic Impact on Alpha Diversity
Alpha diversity, the diversity within samples, within arsenic-treated and non-arsenic treated bacterial communities was measured using the Chaol estimate, Observed OTU richness, Faiths phylogenetic diversity and Shannon Index metrics. Chaol and Observed OTU richness measure the species richness, while Shannon Index measures species diversity. Faiths phylogenetic diversity correlates phylogenetic relationships with species richness. All metrics were higher for arsenic free samples when compared to samples where arsenic was detected. A /-test comparing the average from each group showed a significant diversity difference exists between arsenic and no arsenic samples (p<0.01) (Table 5.2).
Table 5.2. Alpha diversity metrics for arsenic detected and non-detected samples where Chaol estimate, Observed OTU richness, Faiths phylogenetic diversity, and Shannon Index were performed and a /-test was done to test if the mean value for arsenic-detected samples was significantly different from non-arsenic detected samples. The data shown is the average of all samples belonging to each group with the standard deviation in parentheses.
Chaol Observed OTU Richness Faiths PD Shannon Index
Arsenic Present 542 (257) 293(166) 15(8) 3.3 (1.3)
Arsenic Absent 805 (466) 467 (303) 24(14) 4.3 (1.3)
/-test p-value p<0.01 p<0.01 p<0.01 p<0.01
Collection Impact on Alpha Diversity
Alpha diversity was measured within specimens found in the two Collections
(Education vs. Zoology) using the Chaol estimate, Observed OTU richness, Faiths phylogenetic diversity and Shannon Index metrics. Diversity was higher in the Zoology
41


Collection (Table 5.3). A /-test comparing the averages from each group showed significant diversity differences between the Zoology and Education Collections (p<0.01), except Faiths PD (p=0.05) (Table 5.3).
Table 5.3. Alpha diversity metrics for Zoology and Education Collection samples where Chaol estimate, Observed OTU richness, Faiths phylogenetic diversity, and Shannon Index were performed and a /-test was done to test if the mean value for each collection was significantly different. The data shown is the average of all samples belonging to each group with the standard deviation in parentheses.
Chaol Observed OTU Richness Faiths PD Shannon Index
Zoology 703(417) 399 (274) 20(13) 3.94 (1.49)
Education 506(187) 270 (102) 16(5) 3.19 (.603)
/-test p-value Zoology/Education p<0.01 p<0.01 p=0.050 p<0.01
Material Impact on Alpha Diversity
Alpha diversity comparing different source materials (hide, bird mounts, and animal mounts) was measured using the Chaol estimate, Observed OTU richness, Faiths phylogenetic diversity and Shannon Index metrics. Diversity measures were higher for samples where arsenic was not detected compared to the same material type where arsenic was detected (Table 5.4). When comparing arsenic-treated hide and animal mounts, a significant difference existed between the two materials for the Shannon index (p<0.01) and OTU richness (p=0.048) but in the absence of arsenic none of the alpha diversity metrics were significantly different between hide and animal mounts (p>0.01) (Table 5.4). When comparing hide and bird mounts, all metrics were significantly different between the two material types when arsenic was present (p<0.01) but were not significantly different when arsenic was not detected (p>0.01) (Table 5.4). The same pattern was observed when
42


comparing bird and animal mounts a significant difference existed when arsenic was
present (p<0.01) but not when arsenic was absent (p>0.01) (Table 5.4).
Table 5.4. Alpha diveristy metrics for all samples of the same material grouping where Chaol estimate, Observed OTU richness, Faiths phylogenetic diveristy, and Shannon Index were performed and a t-test was done to test if the mean value for each group was significantly different. The data shown is the average of all samples belonging to each group with the standard deviation in parentheses. D= arsenic detected (>20 ppb), ND = arsenic not-detected (<4 ppb).
Chaol Observed OTU Richness Faiths PD Shannon Index
HideD 795 (284) 498 (223) 24(10) 5.35 (1.34)
Hide 592 (296) 365 (213) 20(10) 3.98 (1.16)
Animal Mount15 625 (265) 313 (134) 17(6) 3.45 (.45)
Animal Mount 1008 (549) 567 (351) 27(14) 4.88 (1.59)
Bird Mount15 413 (129) 213 (58) 12(3) 2.60 (.480)
Bird Mount 797 (425) 461 (293) 26(16) 4.10 (1.09)
t-test p-value HideD/Animal Mount15 p=l p=0.048 p=0.160 p<0.01
t-test p-value Hide/Animal Mount p=0.127 p=0.698 p=l p=0.974
t-test p-value HideD/Bird Mount15 p<0.01 p<0.01 p<0.01 p<0.01
t-test p-value Hide/Bird Mount p=l p=l P=1 p=l
t-test p-value Bird Mount15/Animal Mount15 p<0.01 p<0.01 p<0.01 p<0.01
t-test p-value Bird Mount/Animal Mount P=1 P=1 P=1 p=0.868
43


Figure 5.1. Heatmap showing relative abundance of organisms where arsenic was detected (D; >20 ppb) and organisms where arsenic was not detected (ND; < 4ppb) across all museum specimens sampled. Dendrograms show hierarchical clustering (average linkage) of Bray-Curtis dissimilarity for samples on the y-axis. On the x-xis, samples are clustered according to similarity of abundance profiles. (A) Includes the two most abundant genera, Ralstonia and Sediminbacteriam. (B) has these genera removed.
Impact of Arsenic on Bacterial Communities


When comparing microbial communities among specimen samples where arsenic was detected (>20 ppb), less bacterial diversity was observed and the genus Ralstonia had the highest abundance, while the genus Sediminbacterium was highest in samples without arsenic. Because Ralstonia and Sediminbacterium were highly abundant, a heat map was generated without these genera to determine what other groups were highly abundant. A higher abundance of the family Enterobacteriaceae was seen when arsenic was not detected, while Acinetobacter had a high abundance in samples where arsenic was detected (Figure 5.1A; Figure 5.IB).
To evaluate the impact of arsenic presence on the specimen-associated bacterial communities, the arsenic and no arsenic-impacted specimens from the Zoology Collection were compared. Principal coordinate analysis (PCoA) plots of Bray Curtis dissimilarity showed that animal mount samples clustered together based on whether specimen-associated arsenic was present (Figure 5.2A). Axis 1 explained 54.4% of the variation observed among samples and samples arrayed along this axis (Figure 5.2A). Samples were compared from the same Collection (e.g. Education or Zoology) and of the same material type (e.g. bird mounts; animal mounts; hides) in order to detect the influence of the arsenic without the impact of the other variables being taken into consideration. Analysis of similarity (ANOSIM) showed that for animal mount samples, arsenic explained 96.3% of the variation observed within the microbial communities (p=0.001) (Table 5.5). When looking at bird mount samples only, the same pattern was observed, samples with detected arsenic clustered together while those without arsenic clustered (Figure 5.2B). The first axis explained 58.2% of the variation observed among the communities. ANOSIM showed that arsenic explained 87% of the variation observed in the bacterial communities (p<0.01)(Table 5.5).
45


Axis.2 [10.5%]
Arsenic
D
ND
B
(M -
fill
-0 25 0 00 0 25 0 90
Axis 1 {582%]
A/Wlc
D
NO
Figure 5.2. PCoAplot of Bray Curtis dissimilarity of (A) Zoology Collection animal mount samples with arsenic-impacted (D; >20 ppb) samples clustering together and non-impacted (ND; <4 ppb) samples clustering together. (B) Zoology Collection bird mount samples with arsenic-impacted (D; >20 ppb) samples clustering together and non-impacted (ND; <4 ppb) samples clustering together.
46


Table 5.5. ANOSIM tests for Bray Curtis dissimilarity comparing arsenic detected (>20 ppb) and non-detected (<4 ppb) Zoology bird and animal mounts to determine whether arsenic impacted bacterial communities.
ANOSIM R statistic P-value
Zoology bird mounts 0.8696 0.001
Zoology animal mounts 0.9630 0.001
Impact of Collection on Bacterial Communities
When comparing bacterial communities between the Education and Zoology Collection specimens, the genera Ralstonia and Sediminbacterium were abundant across both collections. Ralstonia was more abundant in the Zoology Collection, while Sediminbacterium was more abundant in the Education Collection samples (Figure 5.3A). The next most abundant groups were the family Bradyrhizobiaceae among specimens in the Education Collection, and Enterobacteriaceae and Acinetobacter among specimens in the Zoology Collection (Figure 5.3B)
47


Figure 5.3. Heatmap showing relative abundance of organisms In the Education and Zoology Collections. Dendrograms show hierarchical clustering (average linkage) of Bray-Curtis dissimilarity for samples on the y-axis. On the x-axis, samples are clustered according to similarity of abundance profiles. (A)Includes the two most abundant genera, Ralstonia and Sediminbacterium while (B) has these genera removed.


To evaluate the impact of Collection on the specimen -associated bacterial communities, arsenic unimpacted animal mounts, non-arsenic impacted hides and arsenic-impacted bird mounts from each Collection (Education vs. Zoology) were compared. PCoA plots of Bray Curtis dissimilarity showed microbial similarities among similar specimen types (animal mounts vs. hide) within their Collection location (Education vs. Zoology). For animal mount samples, Axis 1 explained 36.2% of the variation observed and samples arrayed along this axis (Figure 5.4A). All samples compared tested negative for arsenic and were of the same material type in order to detect the influence of the Collection without the impact of the other variables being taken into consideration. ANOSIM statistical testing showed that for animal mount samples, Collection explained 77.9% of the variation observed with the bacterial communities (p=0.001) (Table 5.6). When looking at hide samples that tested negative for arsenic, clustering was seen by collection type among the first axis which explained 55.5% of the variation in bacterial communities (Figure 5.4B). ANOSIM statistical testing showed that Collection type explained 80.7% of the variation (Table 5.6). Arsenic impacted bird mounts showed some clustering by Collection but many Education and Zoology samples also clustered together. The first axis of the PCoA explained 75.8% of the variation in bacterial communities, and ANOSIM showed that Collection type explained 19.8% of the variation (p=0.029) (Figure 5.5; Table 5.6).
49


Axis 2 [24.3%] W Axis.2 [16.1%]
-0.6
-0 4
-0.2
0.0
0.2
Axis.1 [36.2%]
Collection
Education
Zoology
Collection
Education
Zoology
Figure 5.4. PCoAplot of Bray Curtis dissimilarity (A) non-arsenic-impacted (<4 ppb) animal mount samples from the Education and Zoology Collections showing clustering by Collection type, and (B) non-arsenic-impacted (<4 ppb) hide samples from the Education and Zoology Collection showing clustering by Collection type.
50


Collection
Education
# Zoology
Figure 5.5. PCoAplot of Bray Curtis dissimilarity arsenic-impacted (>20ppb) bird mount samples from the Education and Zoology Collection.
Table 5.6. ANOSIM tests for Bray Curtis dissimilarity comparing animal mounts, hides, and bird mounts from Education and Zoology Collections to determine the influence of Collection on bacterial communities.
ANOSIM R statistic P-value
Animal mounts 0.778 0.001
Hide 0.807 0.001
Bird mounts 0.198 0.029
Impact of Material on Bacterial Communities
When comparing bacterial communities among material types (bird mount, animal mount and hide), the genus Ralstonia was most abundant on animal mounts, while Sediminbacterium was the most abundant on bird mount samples (Figure 5.6A). When these two highly abundant genera were removed from analysis, the family Enterobacteriaceae was
51


n
r^Tl ^7^
a>
01
'as
Cd
m r^Vl l^rnTPl

Animal Mount
Hide
Bird Mount
B
Bird Mount
Hide
Animal Mount
- ^ S EE t3 -2 c
O O O
O O O C
---S 2
F CN ;r _l i
3 E '_> .2 m O ^ tj =
O G '- -E CD CO O
S H O -g Qf £
y t3
.2 m OO O oOm
s *i s s- is o z.
V
O CO
I
(3
-9
| o'
2=. U. *- X >
^1 LU I I
oIS|J
a
o I ,
(O |
; o
f 5 f I
o
o co ro ra o
D C t3 3>
3 ro £ c
O t5 C 13 £
O Q.
-, -O E -g co
-2 |
oil Jo1 I IJ
) ra F
>3 3 = 3 t;
> o o o .2
> O O C U 3
) O O O o
) O O p o tJ
> o o t: o cd
u s a 2
2 I
t3
CD
O
2
i _j
1 -G |
- -E m CO
> 3
X. >,
5 8 2 a 1
2 2 a. £
- - o
CO 04
2 ^
3 0
Iso10 | o'
m c C
o | :
I 15 = 5 O w,
lo'J". ' '
o' <3 '
| o'

o o o
! i I
t) C O
o
_Q F -O
I .2 > £ 2
CL O
O
II,
Figure 5.6. Heatmap showing relative abundance of organisms of bird mounts, animal mounts, and hide. Dendrograms show hierarchical clustering (average linkage) of Bray-Curtis dissimilarity for samples on the y-axis. On the x-axis, samples are clustered according to similarity of abundance profiles. (A) Includes the two most abundant genera, Ralstonia and Sediminbacterium while (B) has these genera removed.
Ol
to
highest on bird mount samples while Acinetobacter was highest in animal mount samples


To evaluate the impact of material type on the specimen-associated bacterial communities Zoology non arsenic-impacted animal vs. bird mounts, Zoology arsenic-impacted animal vs. bird mounts, Zoology arsenic-impacted animal mount vs. hide, Zoology arsenic-impacted bird mount vs. hide, and Education non arsenic-impacted animal mount vs. hide were compared. PCoA plots of Bray Curtis dissimilarity showed microbial similarities among similar material types (hide, animal mount, bird mount) within their Collection location (Zoology). All samples compared tested negative for arsenic (<4 ppb) and were from the Zoology Collection in order to detect the influence of the material type without the impact of the other variables being taken into consideration. When comparing animal and bird mounts from the Zoology collection, axis one explained 34.8% of the variation and ANOSIM showed that material explained 80.1% of the variation observed within the bacterial communities (p=0.001) (Table 5.7). Bird and animal mount, arsenic-impacted (> 20 ppb) samples showed clustering by material type on a PCoA plot and the first axis explained 83.5% of the variation in bacterial communities. ANOSIM showed that material explained 97% of the variation (p=0.001) (Figure 5.7B; Table 5.7).
53


Axis.2 [7.4%] Axis.2 [12.2%]
0.4-r
0.2-
o.o-
-0.2-





# *
0.1 -
-0.1 -


-0.25 0.00 Axis.1 [34.8%] 0.25

#
(. W
%
* !%

Material
AnimalMount
BirdMount
Material
AnimalMount
BirdMount
0.00
Axis.1 [83.5%J
0.25
Figure 5.7. PCoA plot of Bray Curtis dissimilarity of (A) non-arsenic-impacted (<4 ppb) bird mount and animal mount samples from the Zoology Collection showing clustering by material and (B) arsenic-impacted (>20 ppb) animal and bird mount samples from the Zoology Collection showing clustering by material type.
54


When looking at hide and animal mount samples from the Zoology Collection that tested positive for arsenic, the first axis of the PCoA plot explained 62.4% of the variation in bacterial communities and ANOSIM statistical testing showed that material explained 94.6% of this variation (p=.001) (Figure 5.8A;Table 5.7). Hide and bird mount arsenic-impacted samples from the Zoology Collection showed clustering by material type and the first axis of the PCoA plot explained 47.7% of the variation in bacterial communities, while ANOSIM statistical testing showed that 74.3% of the variation was explained by material type (p=0.001)(Figure 5.8B; Table 5.7).
55


Axis.2 [25.9%] Axis.2 [13.6%]
Material
AnimalMount
Hide
Material
Bird Mount
Hide
Figure 5.8. PCoAplot of Bray Curtis dissimilarity of (A) arsenic-impacted (> 20 ppb) hide and animal mount samples from the Zoology Collection showing clustering by material type (B) arsenic-impacted (> 20 ppb) hide and bird mount samples from the Zoology Collection showing clustering by material type.
56


PCoA plot of Bray Curtis dissimilarity showed that hide and animal mount samples from the Education Collection (not impacted by arsenic), the first two axis of the PCoA plot explained 83.8% of the variation in bacterial communities (Figure 5.9). However, ANOSIM statistical testing showed that material type did not explain the variation seen in the bacterial communities suggesting another factor was driving the differences (p=0.443) (Table 5.7).
Material
Animal Mount
Hide
Figure 5.9. PCoA plot of Bray Curtis dissimilarity non-arsenic-impacted (<4 ppb) hide and animal mount samples from the Education Collection.
57


Table 5.7. ANOSIM tests for Bray Curtis dissimilarity comparing material types to determine whether material impacted bacterial communities.
ANOSIM R statistic P-value
Arsenic-Impacted
Zoology Hide & Animal Mount 0.946 0.001
Zoology Hide & Bird Mount 0.743 0.001
Zoology Animal Mount & Bird Mount 0.970 0.001
Arsenic Non-impacted
Zoology Animal Mount & Bird Mount 0.801 0.001
Education Animal Mount and Hide 0.000841 0.443
Community Analysis
Community analysis showed the consistent presence of the same genera in all arsenic-impacted samples, while different genera were consistently present in all non-arsenic impacted samples. While many groups of bacteria were found across all collected samples (i.e. Ralstonia, Acinetobacter, Enterobacteriaceae, Comamonadaceae), all samples where arsenic was not detected included Bradyrhizobiaceae, Pseudomonas, Staphyloccocns, Sediminbacterinm, and Sphingomonas (Table 5.8). Ralstonia was much more abundant in samples with arsenic (44%, as opposed to 6% in non-impacted samples), while Enterobacteriaceae was more abundant in non-impacted samples (12%, as opposed to 4% in impacted samples).
Interestingly, all specimen types (bird, animal, hide) across both Collections despite the presence of arsenic, had Ralstonia and families Enterobacteriaceae and
58


Comamonadaceae (Table 5.8). Geobacillus and Acinetobacter were unique to specimens within the Zoology Collection. Unique to Education Collections specimens were Sphingomonas and Pseudomonas and the families Caulobacteraceae and Bradyrhizobiaceae.
Table 5.8. Bacteria consistently present on arsenic-impacted samples, arsenic-absent samples and samples from the Zoology and Education Collections. All genera and families are higher than 1% abundance.
No Arsenic Arsenic Education Collection Zoology Collection
Ralstonia Enter obacteriaceae Bradyrhizobiaceae Acinetobacter Psttedomonas Staphylococcus Sediminbacterium Comamonadaceae Sphingomonas Ralstonia Enter obacteriaceae Acinetobacter Comamonadaceae Ralstonia Comamonadaceae Sphingomonas Enterobacteriaceae Bradyrhizobiaceae Caulobacteraceae Sediminbacterium Pseudomonas Ralstonia Enterobacteriaceae Acinetobacter Geobacillus Comamonadaceae
Dust at the Denver Museum of Nature and Science
Dust samples across the museum were collected in 13 spaces (Table 5.9). All samples tested below detectable limits (<4 ppb) for the presence of arsenic
59


Table 5.9. Dust samples collected from top of door frames in different use rooms across DMNS. Museum location is also indicated. BDL=below detectable limit (<4 ppb).
Sample ID Arsenic Area Use Door Location
DUST-3 BDL Collection Storage Figgins
DUST-4 BDL Collection Storage Old Zoology
DUST-5 BDL Collection Storage Old Zoology
DUST-6 BDL Collection Storage Fur Vault
DUST-15 BDL Collection Storage Education
DUST-17 BDL Collection Storage Education
DUST-18 BDL Public Explore Colorado
DUST-20 BDL Public Explore Colorado
DUST-24 BDL Public Expedition Health
DUST-25 BDL Public Discovery Zone
DUST-26 BDL Public Space Odyssey
DUST-27 BDL Public Gems and Minerals
DUST-30 BDL Public Anschutz Gallery
The bacterial communities in the dust revealed the abundance of the order Streptophyta as well as the genus Ralstonia (Figure 5.10). Also abundant were the genera Staphylococcus, Streptococcus, and Acinetohacter (Figure 5.11).
60


Color Key
0 0.2
0.4 0.6
Value
0.8

rW FT
TTm rVn
DUST-6
DUST-5
DUST-27
DUST-26
DUST-15
DUST-4
DUST-25
DUST-18
DUST-24
DUST-30
DUST-20
DUST-3
DUST-17
ro ro io co
f ip
& s g
% s,
I I
o
CO t-
s e
ra g E .2
§_ cu -Q
& t r o _
L! Q. N
t TS i= f S
l e ! ? I
E
3
c ra 1 s
h
wi1 1 <->,
JD
e s y 2
- -r
T,
K, *
I £
1 I
s| "l u. fc m
o O I *, 3
O I I o
O 1
CO E * in
3 5 3 ra
o 2 o c t3
o I 8 9fi I
o o in ra &
o o o % o o
O COO Sue
Tills!
{ I t 3 P s
t m 5 <
1
CJ O TO o ^ c
O e .Q O JS O
8 I 1 *, i 5
CBS ra o
s 1 I
o |
o '
I E % g E 3 J f I S| R1 1 I I o
Figure 5.10. Heatmap showing relative abundance, including the Order Streptophyta and genus Ralstonia, which were highly abundant in the majority of samples. Dendrograms show hierarchical clustering (average linkage) of Bray-Curtis dissimilarity for samples on the y-axis. On the x-axis, samples are clustered according to similarity of abundance profiles.
61


Color Key
DUST 20
DUST-17
DUST-15
DUST-1
DUST-6
DUST-24
DUST-1B
DUST-25
DUST-27
DUST-26
DUST 30
DUST-3
DUST-6
Figure 5.11. Heatmap showing relative abundance, excluding the Order Streptophyta genus Ralstonia. Dendrograms show hierarchical clustering (average linkage) of Bray-Curtis dissimilarity for samples on the y-axis. On the x-axis, samples are clustered according to similarity of abundance profiles.
Bacterial communities among the dust samples were explored using PCoA plots of Bray Curtis dissimilarity and showed that the first axis of the PCoA plot explained 42.4% of the variation and samples from public areas were closer to each other than to samples frorm collection areas (Figure 5.12). However, ANOSIM statistical testing showed that a significant difference did not exist between the collection storage area samples and the public areas (p=0.08) (Table 5.10).
62


Area
Collection
Public
Figure 5.12. PCoAplot of Bray Curtis dissimilarity of dust samples from the collection storage and public areas.
On


Table 5.10. ANOSIM tests for Bray Curtis dissimilarity comparing collection storage and public areas to determine whether area impacts bacterial communities.
ANOSIM R statistic P-value
Area 0.1495 0.08
Alpha diversity was measured using the Chaol estimate, Observed OTU richness, Faiths phylogenetic diversity and Shannon Index metrics. Chaol estimate was higher for public area samples while all other metrics were higher in the samples belonging to collection are samples. However, a t-test comparing the average from each group showed no significant difference (Table 5.11).
Table 5.11. Alpha diversity metrics for public and collection samples where Chaol estimate, Observed OTU richness, Faiths phylogenetic diversity, and Shannon Index were performed and a t-test was done to test if the mean value for each area was significantly different. The data is shown as the average of all dust samples belonging to each group with the standard deviation in parentheses.
Chaol Estimate Observed OTU Richness Faiths PD Shannon Index
Public 2660 (614) 1324 (295) 75 (13) 6.03 (1.04)
Collection 2425 (826) 1466 (463) 80 (23) 7.37(2.06)
t-test p-value 0.600 0.552 0.647 0.192
Discussion
This study characterized the bacterial communities found on specimens housed at the
Denver Museum of Nature and Science. Specimens examined included arsenic-impacted/unimpacted; animal mounts, bird mounts, and hides; and were housed in both
64


Education and Zoology Collections. The presence of arsenic, the material type, and the Collection likely combined to influence community presence and contribute to the differences observed among the bacterial communities. The presence of arsenic was more influential on animal mounts versus bird mounts and hides. Alpha diversity analyses (Chaol, OTU richness, Faiths Phylogenetic Diversity, and Shannon Index) were higher for samples where arsenic was absent (p<0.01), suggesting that arsenic decreases the bacterial diversity. Bacterial community analyses showed that four genera/families (e.g. Ralstonia, Acinetobacter, Comamonadaceae, and Enterobacteriaceae) were consistently associated with arsenic presence, and nine genera/families (e.g Ralstonia, Acinetobacter, Sphingomonas, Comamonadaceae) were associated with the absence of arsenic. Many bacteria do not possess the necessary resistance mechanisms to allow them to live in the presence of arsenic so decreased diversity on arsenic-impacted samples was expected. Acinetobacter and Ralstonia have genes for arsenic resistance (Cai 2009; Mergeay 2003). Non-impacted samples had higher abundances of human-associated organisms (Enterobacteriaceae family, Staphylococcus, and Pseudomonas) likely due to more handling of these items. While museums do not have records, many items acquired prior to 1970 are assumed to have been treated with arsenic and are therefore treated with caution. When comparing microbial communities among animal mounts with arsenic presence and absence as well as bird mounts with arsenic presence and absence, a significant difference existed
(p=0.001)
The Collection Department housing individual specimens additionally influenced bacterial community. Alpha diversity was higher with Zoology Collection samples than Education (p<0.01) except for Faiths Phylogenetic diversity (p=0.05). Microbiome analyses
65


identified two genera, Geobacillus and Acinetobacter, present on Zoology specimens and absent from Education specimens. Eight bacterial groups were indicative of specimens housed within the Education Collection, and Sphingomonas, Bradyrhizobiaceae, Caulobacteraceae, Sediminbacterium, and Pseudomonas were all unique to these specimens. Acinetobacter can be associated with humans as it can cause nosocomial infections as well as the genus Pseudomonas (Jawad 1998). Many of the other organisms found in Education and Zoology Collection samples are commonly associated with soil and water. When comparing bacterial communities on arsenic non-impacted animal mounts, hides and arsenic-impacted bird mounts, all were significantly different based on the Collection they belonged to
(p<0.01).
Diversity measures showed no significant differences between communities associated with hides and animal mounts. When arsenic was present, hide samples had significantly higher alpha diversity than bird mounts, but when arsenic was not detected, bird mount samples had higher alpha diversity (p<0.01). Bird and animal mounts were not significantly different when arsenic was absent. However, in the presence of arsenic, animal mounts were significantly more diverse (p<0.01). While the material type influenced bacterial diversity, the presence or absence of arsenic with the material had a larger effect. Bird and animal mount specimens from both the Zoology and Education Collections were stored on shelves. Hide samples from the Zoology Collection were stored in closed cabinets, while the hides from the Education Collection were stored in plastic bags and bins. These could be contributing factors to the decreased alpha diversity observed with Education Collection samples.
66


Interestingly, dust in the museum revealed that many of the same groups of bacteria as on the specimens including Ralstonia, Sphingomonas, Pseudomonas, Acinetobacter and family Enterobacteriaceae. Dust samples from public areas compared to collection storage areas did not show a significant difference. While many of the specimens are protected from dust in bags or in storage cabinets, dust could still be a potential source of the bacterial communities observed. Human related organisms (i.e. Enterobacteriaceae) are also observed on the specimens and some are handled more often than others, which could contribute to the surface-associated bacterial communities.
The presence of bacteria on museum specimens is of particular interest from a biodegradation standpoint as it is well know that bacteria and fungi are common culprits in material degradation in museums. Studies looking at the decay of wood collections in museum have identified many fungal and bacterial species (e.g. Cytophaga, Cellvibrio, Clsotridium, Bacdlus, Arthrobacter, Flavobacterium, Spirillum, Aspergillus, Cladosporium, Trichoderma) involved (Kim 2000; Gelbrich 2008; Kretschmar 2008). The genera Bacillus, Staphylococcus, Pseudomonas, Virgibacillus, and Micromonospora have been implicated in the deterioration of parchment paper (Krakova 2012). Studies characterizing bacteria associated with feather degradation on birds focused on bacteria isolated from feathers with keratinolytic activity (the degradation of keratin). The studies showed that bacteria from the Firmicutes phylum (i.e. Bacillus, Staphylococcus) and Actinobacteria and Proteobacteria (Kocuria, Micrococcus, Streptomyces, Pseudomonas) were common degradation culprits (Burtt 2004; Gunderson 2008; Thys 2004). Studies looking at fur seals revealed that the rectum contained species of Staphylococcus, Escherichia, Proteus, Acinetobacter, and Salmonella (Vedros 1982). Many of these same organisms were found in this study here and
67


could potentially contribute to degradation, although this study did not characterize the viability of the material-associated bacteria.
While other factors could also influence the bacterial communities on specimens across the Denver Museum of Nature and Science, this study showed the influence of the presence of arsenic, Collection storage, and material type.
68


CHAPTER VI
CONCLUSION AND FUTURE DIRECTIONS
The bacterial communities associated with museum collections have not been characterized to date and this study is the first of its kind attempting to elucidate how different factors have contributed to these communities. Future studies will expand on different material types and Collections such as including specimens from other Collections. For this study, while not enough data was collected for comparative analysis, preliminary data were collected from the Anthropology Collection at DMNS (Supplemental Table S5; Supplemental Figure SI). Additionally, high-throughput sequencing does not reveal whether bacteria present are viable and future studies can address the viability of organisms and therefore further elucidate the potential for degradation of material.
The presence of arsenic has been one of the influential factors and testing various objects has shown that a wide range of arsenic concentrations still persist on these collections even though the use of arsenic was discontinued decades ago. This makes the matter of remediation an even more pressing issue and future work will continue to elucidate methods for utilizing R. palustris for the removal of arsenic associated with museum collections. The bacterium will be applied to various materials to emulate those seen in the museum (i.e. paper, feather, fur) to determine if it will be successful as a remediation technology in the overall museum setting.
In addition, future work will explore collections in other museums as well as determining bacterial sources on specimens. This will include expanding the dust study to see if the overall museum microbiome is similar between dust and collection or if perhaps the
69


bacteria are inherent to the organisms prior to the taxidermy process. This work is just a small portion in an unexplored system that hopefully opens the door for future studies.
70


SUPPLEMENTAL TABLES AND FIGURES
Supplemental Table SI. Hide and animal mount samples from specimens in the Education Collection. Date acquired and origin of item information is unknown. BDL = below detectable limit (<4 ppb).
Sample Artifact Type Area Arsenic (ppb)
PB4275A Red fox Top back BDL
PB4275B Red fox Mid back BDL
PB4275C Red fox Low back BDL
PBBOBA Bobcat Top back BDL
PBBOBB Bobcat Mid back BDL
PB5669A Coyote hide Top back BDL
PB5669B Coyote hide Mid back BDL
PB5669C Coyote hide Low back BDL
PB6184A Coyote hide Top back BDL
PB6184B Coyote hide Mid back BDL
PB6184C Coyote hide Low back BDL
PB6183A Coyote hide Top back BDL
PB6183B Coyote hide Mid back BDL
PB5747C Arctic fox hide Low back BDL
PB5758A Red wolf hide Top back 60
PB5758B Red wolf hide Mid back BDL
PB5758C Red wolf hide Low back BDL
PB5753A Coyote hide Top back BDL
PB5753B Coyote hide Mid back BDL
PB5753C Coyote hide Low back BDL
71


Supplemental Table S2. Bird Mount samples from specimens in
the Education Collection. Date acquired and origin of item
information is unknown.
Sample Artifact Type Area Arsenic (ppb)
PB1119A Hooded merganser Back >2,000
PB1119B Hooded merganser Belly >2,000
PB1119C Hooded merganser Cloaca >2,000
PB3085A Buffle head duck Back 60
PB3085B Buffle head duck Belly 40
PB3085C Buffle head duck Cloaca 50
PB0248A Bird Back 60
PB0248B Bird Belly 500
PB0248C Bird Cloaca 1,500
PB3088A Bird Back 20
PB3088B Bird Belly 40
PB3088C Bird Cloaca 20
PB3091A Bird Back 250
PB3091B Bird Belly 425
PB3091C Bird Cloaca >2,000
72


Supplemental Table S3. Hide and animal mount samples from specimens in the Zoology
Collection with associated arsenic levels. UNK = unknown.
SamplelD Artifact Type Area Arsenic (ppb) Date Acquired Origin of item
DMNSZM2B Bison Mid back 150 2/14/1905 South Park, CO
DMNSZM2C Bison Low back 1000 2/14/1905 South Park, CO
ZM2397A Deer Top back too 10/11/2025 Brazil
ZM2397B Deer Mid back 60 10/11/2025 Brazil
ZM2397C Deer Low back 60 10/11/2025 Brazil
ZM2401A Deer Top back too 11/19/2025 Brazil
ZM2401B Deer Mid back too 11/19/2025 Brazil
ZM2401C Deer Low back too 11/19/2025 Brazil
ZM2480A Howler Monkey Top back too 4/3/1905 UNK
ZM2480B Howler Monkey Mid back 50 4/3/1905 UNK
ZM2480C Howler Monkey Low back 50 4/3/1905 UNK
ZM2479A Howler Monkey Top back 50 4/3/1905 UNK
ZM2479B Howler Monkey Mid back 50 4/3/1905 UNK
ZM2479C Howler Monkey Low back 80 4/3/1905 UNK
ZM2441A Tapir Top back 50 11/7/2014 South America
ZM2441B Tapir Mid back 50 11/7/2014 South America
ZM4C Bison Low back 750 2/14/1905 UNK
ZM2496A Jaguar Top back 60 9/11/2025 Brazil
ZM2496B Jaguar Mid back 60 9/11/2025 Brazil
ZM2496C Jaguar Low back 60 9/11/2025 Brazil
ZM3A Bison Top back 250 2/14/1905 UNK
ZM3B Bison Mid back 250 2/14/1905 UNK
ZM3C Bison Low back 250 2/14/1905 UNK
Z2334A Puma hide Top back 20 4/8/1905 Argentina
Z2334B Puma hide Mid back 50 4/8/1905 Argentina
Z2334C Puma hide Low back 20 4/8/1905 Argentina
Z2335A Puma hide Top back 50 4/8/1905 Argentina
Z2335B Puma hide Mid back 20 4/8/1905 Argentina
Z2335C Puma hide Low back 20 4/8/1905 Argentina
Z1933A Puma hide Top back 60 4/6/2018 Gardiner Mountain
Z1933B Puma hide Mid back 60 4/6/2018 Gardiner Mountain
Z1933C Puma hide Low back 60 4/6/2018 Gardiner Mountain
Z6699A Jaguar hide Top back 60 UNK UNK
Z6699B Jaguar hide Mid back 50 UNK UNK
Z6699C Jaguar hide Low back 20 UNK UNK
Z2717A Jaguar hide Top back 40 1/23/2026 Brazil
Z2717B Jaguar hide Mid back 80 1/23/2026 Brazil
Z2717C Jaguar hide Low back 80 ppb 1/23/2026 Brazil
73


Supplemental Table S3 Continued. Hide and animal mount samples from specimens
in the Zoology Collection with associated arsenic levels. *BDL= below detectable limit
(<4 ppb). UNK = unknown.
SamplelD Artifact Type Area Arsenic (ppb) Date acquired Origin of item
DZTM3101A Raccoon Top back BDL UNK UNK
DZTM3101B Raccoon Mid back BDL UNK UNK
DZTM3101C Raccoon Low back BDL UNK UNK
DZTM3174A Raccoon Top back BDL UNK UNK
DZTM3174B Raccoon Mid back BDL UNK UNK
DZTM3174C Raccoon Low back BDL UNK UNK
DZTM3372A Fox Top back BDL UNK UNK
DZTM3372B Fox Mid back BDL UNK UNK
DZTM3372C Fox Low back BDL UNK UNK
DZTM3360A Fox Top back BDL UNK UNK
DZTM3360B Fox Mid back BDL UNK UNK
DZTM3360C Fox Low back BDL UNK UNK
DZTM3463A Beaver Top back BDL UNK UNK
DZTM3463B Beaver Mid back BDL UNK UNK
DZTM3463C Beaver Low back BDL UNK UNK
Z6969C Bobcat hide Low back BDL 1920-1922 Gilpin, CO
Z5200A Bobcat hide Top back BDL 4/23/1943 Kingsville, TX
Z5200B Bobcat hide Mid back BDL 4/23/1943 Kingsville, TX
Z5200C Bobcat hide Low back BDL 4/23/1943 Kingsville, TX
74


Supplemental Table S4. Bird Mount samples from specimens in the Zoology Collection
with associated arsenic levels. UNK=unknown.
SamplelD Artifact Type Area Arsenic (ppb) Date acquired Origin of item
Z10510A Red Eagle Back 250 3/26/2024 Littleton, CO
Z10510B Red Eagle Belly 875 3/26/2024 Littleton, CO
Z10510C Red Eagle Cloaca 250 3/26/2024 Littleton, CO
Z10569A Red Eagle Back 150 10/3/2024 Littleton, CO
Z10569B Red Eagle Belly 125 10/3/2024 Littleton, CO
Z10569C Red Eagle Cloaca 150 10/3/2024 Littleton, CO
Z15069A Gull Back 150 9/21/1935 San Pedro, CA
Z15069B Gull Belly 80 9/21/1935 San Pedro, CA
Z15069C Gull Cloaca 70 9/21/1935 San Pedro, CA
Z21281A Gull Back 875 4/9/1940 Isabela Island, Mexico
Z21281B Gull Belly 1,000 4/9/1940 Isabela Island, Mexico
Z21281C Gull Cloaca 875 4/9/1940 Isabela Island, Mexico
Z22178A Gull Back 1,000 3/21/1940 Adams County, CO
Z22178B Gull Belly >2,000 3/21/1940 Adams County, CO
Z22178C Gull Cloaca 1,000 3/21/1940 Adams County, CO
ZB1078B Woodpecker Belly 750 10/15/2010 Jefferson County, CO
ZB763B Pygmic Nuthatch Belly 80 UNK UNK
ZB763C Pygmic Nuthatch Cloaca 50 UNK UNK
ZB762A White Breasted Nuthatch Neck 750 UNK UNK
ZB762B White Breasted Nuthatch Belly 150 UNK UNK
ZB552B Woodpecker Belly 60 2/25/2010 Yuma County, CO
ZB552C Woodpecker Cloaca 100 2/25/2010 Yuma County, CO
ZB553A Woodpecker Neck 60 2/25/2010 Yuma County, CO
ZB553B Woodpecker Belly 50 2/25/2010 Yuma County, CO
ZB553C Woodpecker Cloaca 50 2/25/2010 Yuma County, CO
ZB562A Belted Kingfisher Neck 250 5/10/2010 Adams County, CO
ZB563C Belted Kingfisher Cloaca 100 5/18/2010 Adams County, CO
ZB630A Brewers Blackbird Neck 150 9/28/2010 Weld County, CO
ZB630B Brewers Blackbird Belly 425 9/28/2010 Weld County, CO
ZB630C Brewers Blackbird Cloaca 425 9/28/2010 Weld County, CO
ZB631B Brewers Blackbird Belly 500 5/10/2012 Weld County, CO
75


Supplemental Table S4 Continued. Bird Mount samples from the Zoology Collection
with associated arsenic levels. BDL = below detectable limit (<4 ppb) UNK = unknown.
Sample Artifact Type Area Arsenic Date acquired Origin of item
ZB43144A Cooper's Hawk Back BDL 7/3/2009 Denver, CO
ZB43144B Cooper's Hawk Belly BDL 7/3/2009 Denver, CO
ZB43144C Cooper's Hawk Cloaca BDL 7/3/2009 Denver, CO
ZB43266A Cooper's Hawk Back BDL 6/9/2009 Aurora, CO
ZB43266B Cooper's Hawk Belly BDL 6/9/2009 Aurora, CO
ZB43266C Cooper's Hawk Cloaca BDL 6/9/2009 Aurora, CO
DZTB664A Dove Back BDL 6/18/2013 Golden, CO
DZTB664B Dove Belly BDL 6/18/2013 Golden, CO
DZTB664C Dove Cloaca BDL 6/18/2013 Golden, CO
DZTB837A Red Dove Back BDL 3/6/2014 Longmont, CO
DZTB837B Red Dove Belly BDL 3/6/2014 Longmont, CO
DZTB837C Red Dove Cloaca BDL 3/6/2014 Longmont, CO
DZTB1115A Bird Back BDL 9/27/2014 UNK
DZTB1115B Bird Belly BDL 9/27/2014 UNK
DZTB1115C Bird Cloaca BDL 9/27/2014 UNK
DZTB1110B Duck Belly BDL 11/14/2014 UNK
DZTB1110C Duck Cloaca BDL 11/14/2014 UNK
DZTB1008A Western Grebe Back BDL 12/31/2014 UNK
DZTB1008B Western Grebe Belly BDL 12/31/2014 UNK
DZTB1008C Western Grebe Cloaca BDL 12/31/2014 UNK
DTZB810A Red Dove Back BDL 8/3/2011 Denver, CO
DZTB810B Red Dove Belly BDL 8/3/2011 Denver, CO
DZTB114A Mud Hen Back BDL 11/28/2014 UNK
DZTB114B Mud Hen Belly BDL 11/28/2014 UNK
DZTB1111A Duck Back BDL 12/4/2014 UNK
DZTB1111B Duck Belly BDL 12/4/2014 UNK
DZTB1111C Duck Cloaca BDL 12/4/2014 UNK
76


Supplemental Table S5. Feather samples from the Anthropology
Collection with associated arsenic levels. Date acquired is not
known. BDL=below detectable limit (<4 ppb) UNK=unknown.
SamplelD Artifact Type Area Arsenic Origin of item
A552122AA Neck Ring Urubu Spot 1 BDU Brazil
A552122AB Neck Ring Urubu Spot 2 BDU Brazil
A552122AC Neck Ring Urubu Spot 3 BDU Brazil
A7252AA Feather Headdress Moro Spot 1 BDU Gran Chaco Paraguay
A7252AB Feather Headdress Moro Spot 2 BDU Gran Chaco Paraguay
A7252AC Feather Headdress Moro Spot 3 BDU Gran Chaco Paraguay
A7252BA Feather Headdress Moro Spot 1 BDU Gran Chaco Paraguay
A7252BB Feather Headdress Moro Spot 2 BDU Gran Chaco Paraguay
A7252BC Feather Headdress Moro Spot 3 BDU Gran Chaco Paraguay
AC257A Headdress Jivaro Spot 1 250 ppb Ecuador
AC257B Headdress Jivaro Spot 2 20 ppb Ecuador
AC257C Headdress Jivaro Spot 3 20 ppb Ecuador
AC5716B Dance Bustle Sioux Spot 2 BDU UNK
AC5716C Dance Bustle Sioux Spot 3 BDU UNK
AC5717A Bonet Spot 1 BDU UNK
AC5717B Bonet Spot 2 BDU UNK
AC5717C Bonet Spot 3 BDU UNK
AC5718 Bonet Spot 1 BDU UNK
77


Color Ky
Education
Zoology
Anthropology
Supplemental Figure SI. Heatmap showing relative abundance of organisms in the Education, Zoology, and Anthropology Collections. Dendrograms show hierarchical clustering (average linkage) of Bray-Curtis dissimilarity for samples on the y-axis. On the x-axis, samples are clustered according to similarity of abundance profiles.
78


REFERENCES
Alimetrics [Internet], Espoo Finland, [cited 2016 May 4], Available from: http://www.alimetrics.net/en/index.php/dna-sequence-analysis.
Allen LB, Siitonen PH, Thompson HC. 1997. Methods for the Determination of Arsenic, Cadmiun, Copper, Lead, and Tin in Sucrose, Com Syrups and High- Fructose Corn Syrups by Inductively Coupled Plasma Atomic Emission Spectrometry. J. Agric. Food Chem. 45:162-165.
Aronesty E.2011. eautils :Commandline tools for processing biological sequencing data, http: //code. googl e. com/p/eautil s
Bentley R, Chasteen TG, Bentley R, Chasteen TG. 2002. Microbial Methylation of Metalloids : Arsenic Antimony and Bismuth Microbial Methylation of Metalloids :
Arsenic Antimony and Bismuth. Microbiol. Mol. Biol. Rev. 2002, 66(2):250. DOI:
10.1128/MMBR.66.2.250-271.2002.
Biswas J, Sharma K, Harris KK, Rajput Y. 2013. Biodeterioration agents: Bacterial and fungal diversity dwelling in or on the pre-historic rock-paints of Kabra-pahad, India. Iran. J. Microbiol. 5:309-314.
Blyskal B. 2015. Fungal deterioration of a woollen textile dyed with cochineal. J. Cult. Herit. [Internet] 16:32-39. Available from: http://dx.doi.Org/10.1016/j.culher.2014.01.008
Boutakhrit K, Claus R, Bolle F, Degroodt JM, Goeyens L. 2005. Open digestion under reflux for the determination of total arsenic in seafood by inductively coupled plasma atomic emission spectrometry with hydride generation. Talanta 66:1042-1047.
Burtt. EH, Ichida JM. 2004. Gloger S Rule, Feather-Degrading Bacteria, and Color Variation Among Song Sparrows. Condor 106:681-686.
Cai L, Liu G, Rensing C, Wang G. 2009. Genes involved in arsenic transformation and resistance associated with different levels of arsenic-contaminated soils. BMC Microbiol.
9:4.
Caporaso JG, Kuczynski J, Stombaugh J, Bittinger K, Bushman FD, Costello EK, Knight R. (2010). Correspondence QIEME allows analysis of high- throughput community sequencing data Intensity normalization improves color calling in SOLiD sequencing. Nature Publishing Group, 7(5), 335-336. doi: 10.1038/nmeth0510-335
Cross PS, Odegaard N, Riley MR. 2010. Lipoic acid formulations for the removal of arsenic and mercury from museum artifact materials. J. Archaeol. Sci. 37:1922-1928.
Dash HR, Mangwani N, Das S. 2014. Characterization and potential application in mercury bioremediation of highly mercury-resistant marine bacterium Bacillus thuringiensis PW-05. Environ. Sci. Pollut. Res. Int. 21:2642-53.
79


DeSantis TZ, Hugenholtz P, Larsen N, Rojas M, Brodie EL, Keller K, Huber T, Dalevi D, Hu P, Andersen GL. 2006. Greengenes, a chimerachecked 16S rRNA gene database and workbench compatible with ARB. Appl. Environ. Microbiol. 72:5069-5072.
dos Santos WNL, Cavalcante DD, Macedo SM, Nogueira JS, da Silva EGP. 2013. Slurry Sampling and HG AFS for the Determination of Total Arsenic in Rice Samples. Food Anal. Methods 6:1128-1132.
Egland PG, Gibson J, Harwood CS. 2001. Reductive Coenzyme A-Mediated Pathway for 3-Chlorobenzoate Degradation in the Phototrophic Bacterium Rhodopseudomonas palustris. 67:1396-1399.
Faroon O, Harris OM. 2002. 7 Analytical Methods. Toxicol, profile DDT, DDE, DDD:265-277.
Flury M, Jr WTF, Jury WA. 1997. Long-term depletion of selenium from Kesterson dewatered sediments. The Science of the total environment. 198:259-270.
Fuentes S, Mendez V, AguilaP, Seeger M. 2014. Bioremediation of petroleum hydrocarbons: catabolic genes, microbial communities, and applications. Appl. Microbiol. Biotechnol. 98:4781-94. Available from:
Gauzere C, Moletta-Denat M, Blanquart H, Ferreira S, Moularat S, Godon JJ, Robine E.
2014. Stability of airborne microbes in the Louvre Museum over time. Indoor Air 24:29-40.
Gelbrich J, Mai C, Militz H. 2008. Chemical changes in wood degraded by bacteria. Int. Biodeterior. Biodegrad. 61:24-32.
Glastrup J. 2001. The effectiveness of compressed air in removal of pesticides from ethnographic objects. Collection Forum. 16:19-24.
Gunderson AR. 2008. Feather-Degrading Bacteria: a New Frontier in Avian and Host-Parasite Research? Auk [Internet] 125:972-979. Available from: http://www.bioone.org/doi/abs/10.1525/auk.2008.91008
Huang H, Wu K, Khan A, Jiang Y, Ling Z, Liu P, Chen Y, Tao X, Li X. 2016. Bioresource Technology A novel Pseudomonas gessardii strain LZ-E simultaneously degrades naphthalene and reduces hexavalent chromium. Bioresour. Technol. [Internet] 207:370-378. Available from: http://dx.doi.Org/10.1016/j.biortech.2016.02.015
Jawad A., Seifert H., Snelling A.M., Heritage, J., Hawkey P.M. 1998. Survival of Acinetobacter baumannii on dry surfaces: comparison of outbreak and sporadic isolates. J. Clin. Microbiol. Vol(36). No. 7. pp. 1938-1941.
Kim, Y.S., Singh, A.P., 2000. Micromorphological characteristics of wood biodegradation in wet environments: a review. IAWA Journal 21, 135-155.
80


Kozirog A, Otlewska A, Piotrowska M, Rajkowska K, Nowicka-Krawczyk P, Hachulka M, Wolski GJ, Gutarowska B, Kunicka-Styczynska A, Libudzisz Z, et al. 2014. Colonising organisms as a biodegradation factor affecting historical wood materials at the former concentration camp of Auschwitz II Birkenau. Int. Biodeterior. Biodegrad. 86:171-178.
Kraker D. Native American News L Archives. [Internet], RootsWeb. 2001. [cited 2014 June 1], Available from: http://archiver.rootsweb.ancestry.com/th/read/NATFVE-AMERICAN-NEWS/2001-05/0989513395
Krakova L, Chovanova K, Selim SA, Simonovicova A, Puskarova A, Makova A, Pangallo D. 2012. A multiphasic approach for investigation of the microbial diversity and its biodegradative abilities in historical paper and parchment documents. Int. Biodeterior. Biodegrad. 70:117-125.
Kretschmar El, Gelbrich J, Militz H, Lamersdorf N. 2008. Studying bacterial wood decay under low oxygen conditions-results of microcosm experiments. Int. Biodeterior. Biodegrad. 61:69-84.
Larimer FW, Chain P, Hauser L, Lamerdin J, Malfatti S, Do L, Land ML, Pelletier DA, Beatty JT, Lang AS, Tabita FR, Gibson JL, Hanson TE, Bobst C, Torres JL, Harrison FH, Gibson J, Harwood CS. 2004. Complete genome sequence of the metabolically versatile photosynthetic bacterium Rhodopseudomonaspalustris. Nat. Biotechnol. 22:55-61.
Lech, T. 2016. Evaluation of a parchment document, the 13th century incorporation charter for the city of Krakow, Poland for microbial hazards. Appl. Environ. Microbiol. 82:2620-2631.
Liu S, Zhang F, Chen J, Sun G. 2011. Arsenic removal from contaminated soil via biovolatilization by genetically engineered bacteria under laboratory conditions. J. Environ. Sci. 23:1544-1550.
Majumder A, Bhattacharyya K, Kole SC, Ghosh S. 2013. Efficacy of indigenous soil microbes in arsenic mitigation from contaminated alluvial soil of India. Environ. Sci. Pollut. Res. Int. 20:5645-53.
McGrath JE, Harfoot CG. 1997. Reductive dehalogenation of halocarboxylic acids by the phototrophic genera Rhodospirillum and Rhodpseudomonas. Appl. Environ. Microbiol. 63:3333-3335.
McMurdie PJ, Holmes S. 2015. Shiny-phyloseq: Web application for interactive microbiome analysis with provenance tracking. Bioinformatics 31:282-283.
Mergeay M, Monchy S, Vallaeys T, Auquier V, Benotmane A, Bertin P, Taghavi S, Dunn J, Van Der Lelie D, Wattiez R. 2003. Ralstonia metallidurans, a bacterium specifically adapted to toxic metals: Towards a catalogue of metal-responsive genes. FEMS Microbiol. Rev. 27:385-410.
Niel CBV. 1944. The culture, general physiology, morphology, and classification of the nonsulfur purple and brown bacteria. Microbiology and Molecular Biology Reviews. 8(1): 1
81


Nishimura T, Hamano-Nagaoka M, Sakakibara N, Abe T, Maekawa Y, Maitani T. 2010. Determination method for total arsenic and partial-digestion method with nitric acid for inorganic arsenic speciation in several varieties of rice. Food Hyg. Saf. Sci. 51:178-181.
Odegaard N, Sadongei A. Old Poisons, New Problems A museum resource for managing contaminated cultural materials. Altamira Press. 2005.
Oksanen J, Blanchet F, Kindt R, Legendre P, Ohara R, Simpson G, Solymos P, Stevens M, Wagner H. 2015. vegan: Community Ecology Package.
Oremland RS, Stolz JF. 2003. The ecology of arsenic. Science 300:939-44.
Paez-Espino D, Tamames J, de Lorenzo V, Canovas D. 2009. Microbial responses to environmental arsenic. Biometals 22:117-30.
Pantuzzo FL, Silva JCJ, Ciminelli VST. 2009. A fast and accurate microwave-assisted digestion method for arsenic determination in complex mining residues by flame atomic absorption spectrometry. J. Hazard. Mater. 168:1636-1638.
Pray LL. Taxidermy. The MacMillan Company. 1923.
Qin J, Rosen BP, Zhang Y, Wang G, Franke S, Zhangtt Y, Wangts G, Franket S, Rensingt C. 2006. Arsenic Detoxification and Evolution of Trimethylarsine Gas by a Microbial Arsenite S-Adenosylmethionine Methyltransferase. Proceedings of the National Acadmey of Sciences Vol(103). No. 7. pp. 2075-2080.
R Core Team. 2014. R: A Language and Environment for Statistical Computing. Vienna, Austria. http://www.rproject.org
Roane TM, Snelling LJ.2010. Bacterial Removal of Mercury from Museum Materials: A New Remediation Technology? Pesticide Mitigation in Museum Collections: Science in Conservation, pp. 29-34.
Rojas LA, Yanez C, Gonzalez M, Lobos S, Smalla K, Seeger M. 2011. Characterization of the metabolically modified heavy metal-resistant Cupriavidus metallidurans strain MSR33 generated for mercury bioremediation. PLoS One 6:el7555.
Shen Z, Han J, Wang Y, Sahin O, Zhang Q (2013) The Contribution of ArsB to Arsenic Resistance in Campylobacter jejuni. PLoS ONE 8(3): e58894. doi: 10.1371/joumal.pone.0058894
Sirois JP. 2001. The analysis of museum objects for the presence of arsenic and mercury: Non-destructive analysis and sample analysis. Collection Forum. 16: 65-75.
Skora J, Gutarowska B, Pielech-Przybylska K, St^pien L, Pietrzak K, Piotrowska M, Pietrowski P. 2015. Assessment of microbiological contamination in the work environments of museums, archives and libraries. Aerobiologia (Bologna). 31:389-401.
Slyemi D, Bonnefoy V. 2012. How prokaryotes deal with arsenic. Environ. Microbiol. Rep. 4:571-586.
82


Spencer, B. San Francisco State University. San Francisco (CA): c2000 [cited 2014 June 1],
Sterflinger K, Pinzari F. 2012. The revenge of time: Fungal deterioration of cultural heritage with particular reference to books, paper and parchment. Environ. Microbiol. 14:559-566.
Sterflinger K, Pinar G. 2013. Microbial deterioration of cultural heritage and works of art -Tilting at windmills? Appl. Microbiol. Biotechnol. 97:9637-9646.
The llumina HiSeq 2000 Sequencing Technology [Internet], Heidelberg Germany: German Cancer Research Center; [updated 2015 Feb 24; cited 2016 May 4], Available from: http s ://www. dkfz. de/gpcf/hi seq_technol ogy. html.
Thys RCS, Lucas FS, Riffel A, Heeb P, Brandelli A. 2004. Characterization of a protease of a feather-degrading Microbacterium species. Lett. Appl. Microbiol. 39:181-186.
Ulusoy HI, Yilmaz O, Giirkan R. 2013. A micellar improved method for trace levels selenium quantification in food samples, alcoholic and nonalcoholic beverages through CPE/FAAS. Food Chem. 139:1008-1014.
VahterM. 2002. Mechanisms of arsenic biotransformation. Toxicology. 181-182:211-7.
Vedros NA, Quinlivan J, Cranford R. 1982. Bacterial and fungal flora of wild northern fur seals (Callorhinus ursinus). J.Wildl.Dis. [Internet] 18:447-456. Available from: http://www.ncbi.nlm.nih.gov/pubmed/7154218
Wong WT, Tseng CH, Hsu SH, Lur HS, Mo CW, Huang CN, Hsu SH, Lee KT, Liu CT. 2014. Promoting effects of a single Rhodopseudomonas palustris inoculant on plant growth by Brassica rapa chinensis under low fertilizer input. Microbes and Environment. 29:303-313.
XRNA RNA Secondary Structure Files [Internet], cited 2016 May 4], Available from: http://rna.ucsc.edu/rnacenter/xrna/xrna_gallery.html.
Yergeau E, Sanschagrin S, Beaumier D, Greer CW. 2012. Metagenomic analysis of the bioremediation of diesel-contaminated Canadian high arctic soils. PLoS One
Yin X-X, Chen J, Qin J, Sun G-X, Rosen BP, Zhu Y-G. 2011. Biotransformation and volatilization of arsenic by three photosynthetic cyanobacteria. Plant Physiol. 156:1631-8.
Yuan C, Lu X, Qin J, Rosen BP, Le CX. 2008. Volatile Arsenic Species Released from Escherichia coli Expressing the AsIII S-adenosylmethionine. Environ. Sci. Technol. 42:3201-3206.
Zhao C, Zhang Y, Chan Z, Chen S, Yang S. 2015. Insights into arsenic multi-operons expression and resistance mechanisms in Rhodopseudomonas palustris CGA009. Frontiers in Microbiology. 6:1-8.
Zhou CY, Wong MK, Koh LL, Wee YC. 1997. Evaluation of high-pressure microwave digestion methods for hydride generation atomic absorption spectrometric determination of total selenium and arsenic in sediment. Mikrochim. Acta 127:77-83.
83


Full Text

PAGE 1

NOVEL METHOD OF BIOREMEDIATION AND CHARACTERIZATION OF BACTERIAL COMMUNITIES ON ARSENIC IMPACTED MUSEUM COLLECTIONS by SLADJANA SUBOTIC B.S., University of Colorado Denver, 2012 A thesis submitted to the Faculty of the Graduate School of the University of Colorado in partial fulfillment of the requirements for the degree of Master of Science Biology 2016

PAGE 2

ii 2016 SLADJANA SUBOTIC ALL RIGHTS RESERVE D

PAGE 3

iii This thesis for the Master of Science degree by Sladjana Subotic has been approved for the Biology Program by Timberley M. Roane, Chair Annika C. Mosier Alan M. Vajda July 30, 2016

PAGE 4

iv Subotic, Sladjana (M.S. Biology) Novel Method of Bioremediation and Characterization of Bacterial Communities on Arsenic Impacted Museum Collections Thesis directed by Associate Professor Timberley M. Roane ABSTRACT Metal based p esticides including toxic arsenic salts, were widely used prior to the 1970s for the prevention of rodent and insect damage to museum collections. An estimated 80% of collections hou sed in U.S. and Canadian museum s have exogenous metals present, and the use of bacteria in the removal of these metals is of interest. Research shows that Rhodopseudomonas palustris a metabolically versatile proteobacterium, is able to volatilize arsenic via methylation resulting in the convers ion of the metal into a gas which can be collected and removed. In this study, R. palustris tolerated concentrations of arsenic up to 250 ppm and showed the ability to remove up to 78% of arsenic from a starting concentration of 10 ppm soluble arsenic within 19 days. In order to optimize the potential application of bacterial volatilization of arsenic as a remediation technology for arsenic treated museum specimens the presence of in situ bacteria on the surface of museum collections needs to be addressed. High throughput sequen cing revealed diverse bacterial communities associated with the museum specimens examined at the Denver Museum of Nature and Science. Bacterial community composition seemed strongly influenced by several factors including type of materials comprising the s pecimen, the presence of arsenic, and which museum Collection the specimen was housed in. Organisms of high abundance across the items sampled included the genera Ralstonia Sediminbacterium Acinetobacter and the family Enterobacteriaceae Additionally, t esting of surface associated arsenic revealed a wide range of concentrations on arsenic impacted items at the Denver Museum of Nature and Science.

PAGE 5

v This study is the first to utilize high throughput sequencing techniques to characterize the bacterial commu nities associated with museum collections. The form and content of this abstract are approved. I recommend its publication. Approved: Timberley M. Roane

PAGE 6

vi ACKNOWLEDGMENTS There are many people I would like to thank for their role and involvement in this project. First off, thank you to my advisor Dr. Timberley Roane who allowed me the opportunity to work in her lab has encouraged me to grow as a scientist, and has been an amazing mentor. One of the best decisions I ever made was to take her microbiology course that opened the door to many amazing opportunities. Thank you to my committee members Dr. Annika Mosier and Dr. Alan Vajda without all of your contribution and exp ertise I would not be where I am today. Thank you to the amazing individuals at the Denver Museum of Nature and Science who made a huge part of this project possible. First off, thank you to Richard Busch, education collection manager, who helped us get i n touch with many of the other collection managers and supported our project from the beginning Without Rich, this project would not have been possible. I would also like to acknowledge and thank Jeff Stephenson, zoology collection manager and Melissa Bec hhoefer, anthropology collection manager and NAGPRA coordinator who allowed us to work with their collections for the completion of this project. I would like to thank several individuals for their help with sample collection: Anna Nguyen, Helen Dupree, Ke lsey Foster, and Nancy Moreno Huizar. A thank you to Bhargavi Ramanathan for support and help through the writing process. I would a lso like to thank Jeffrey Boon and the University of Colorado Shared Analytical Services laboratory for help with arsenic quantif ication and method development. Finally, there are a few people that deserve a special thank you and to whom I will always be grateful for their hel p and support. Joshua Sackett, you have been an amazing friend and colleague from the start of my program and even after you moved away I could always count on you not just for help but for advice and support. Thank you for everything

PAGE 7

vii you have done to help me from data analysis to the numerous conversations of encoura gement through out this process. gotten through the hardest part of this thesis. Thank you for investing your time and knowledge to help me navigate through the data analysis and for provi ding me with advice and support. A hu ge thank you to Dr. Miller for help with data and for allowing the use of his server and resources for data analysis. Munira Lantz and Anna Nguyen, I will always be grateful for the amazing support system you two have been. Thank you! I would finally like to thank my family my parents, brother and husband who have always supported my goals and never doubted me. I am grateful to have you on my team. La stly, I would like to thank the National Park Service National Center for Preservation Technology and T raining (Grant #P13AP00081) for funding this project.

PAGE 8

viii TABLE OF CONTENTS CHAPTER I. PROJECT OVERVIEW AND OBJECTIVES ................................ ................................ ...... 1 II. BACKGROUND: USE OF ARSENIC VO LATILIZATION AND DEVELOPMENT OF ARSENIC QUANTIFICATION METHODS ................................ ................................ .......... 3 Overview ................................ ................................ ................................ ............................... 3 Bacterial Volatilization of Arsenic ................................ ................................ ........................ 3 Quantification Methods Available for Arsenic ................................ ................................ ..... 7 Rhodopseudomonas palustris and Arsenic ................................ ................................ ........... 8 NAGPRA and Museum Pesticides ................................ ................................ ...................... 10 III. CONDUCTED STUDY: APPLICATION OF ARSENIC VOLATILIZATION BY Rhodopseudomonas palustris ................................ ................................ ................................ 14 Abstract ................................ ................................ ................................ ............................... 14 Introduction ................................ ................................ ................................ ......................... 14 Methods ................................ ................................ ................................ ............................... 16 Confirmation of Arsenic Removal by R. palustris ................................ .......................... 16 Development of an Arsenic Application and Detection Method for Simulated Museum Materials ................................ ................................ ................................ .......................... 16 Metho d 1: Quantification of Arsenic From Paper Using Microwave Digestion ......... 17 Method 2: Quantification of Arsenic From Paper Using Acid Digestion ................... 17 Method 3: Quantification of Arsenic from Paper Using Boiling Reflux Digestion .... 18 Graphite Furnace Atomic Absorption Spectrophotometry ................................ .............. 18 Results ................................ ................................ ................................ ................................ 19 Confirmation of Arsenic Removal by R. palustris ................................ .......................... 19 Development of an Arsenic Application and Detection Method for Simulated Museum Materials ................................ ................................ ................................ .......................... 22

PAGE 9

ix Method 1 Results: Quantification of Arsenic from Paper Using Microwave Digestion ................................ ................................ ................................ ................................ ...... 22 Method 2 Results: Quantification of Arsenic from Paper Using Acid Digestion ........ 23 Method 3 Results: Quantification of Arsenic from Paper Using Boiling Reflux Digestion ................................ ................................ ................................ ...................... 24 Discussion ................................ ................................ ................................ ........................... 25 IV. BACKGROUND: MICROBIOLOGY OF THE DENVER MUSEUM OF NATURE AND SCIENCE ................................ ................................ ................................ ...................... 26 Overview ................................ ................................ ................................ ............................. 26 Preservation Methods ................................ ................................ ................................ .......... 26 Microbiology of Museums ................................ ................................ ................................ .. 27 16S rDNA and Sequencing ................................ ................................ ................................ 28 Denv er Museum of Nature and Science ................................ ................................ .............. 31 V. CONDUCTED STUDY: MICROBIOLOGY OF THE DENVER MUSEUM OF NATURE AND SCIENCE ................................ ................................ ................................ ..... 34 Abstract ................................ ................................ ................................ ............................... 34 Introduction ................................ ................................ ................................ ......................... 34 Methods ................................ ................................ ................................ ............................... 35 Sampling Methodology ................................ ................................ ................................ ... 35 Bacterial DNA Collection Method ................................ ................................ .................. 36 Dust Sample Collection ................................ ................................ ............................... 36 Arsenic Detection and Quantification ................................ ................................ ............. 36 Bacterial Community Characterization ................................ ................................ ........... 37 Computational Analyses ................................ ................................ ................................ .. 39 Results ................................ ................................ ................................ ................................ 39

PAGE 10

x Presence of Arsenic on Specimens ................................ ................................ ...................... 39 Arsenic Impact on Alpha Diversity ................................ ................................ ................. 41 Collection Impact on Alpha Diversity ................................ ................................ ............. 41 Material Impact on Alpha Diversity ................................ ................................ ................ 42 Impact of Arsenic on Bacterial Communities ................................ ................................ 44 Impact of Collection on Bacterial Communities ................................ ............................. 47 Impact of Material on Bacterial Communities ................................ ................................ 51 Community Analysis ................................ ................................ ................................ ....... 58 Dust at the Denver Museum of Nature and Science ................................ ........................ 59 Discussion ................................ ................................ ................................ ........................... 64 VI. CONCLUSION AND FUTURE DIRECTIONS ................................ ............................. 69 SUPPLEMENTAL TABLES AND FIGURES ................................ ................................ ...... 71 REFERENCES ................................ ................................ ................................ ....................... 79

PAGE 11

1 CHAPTER I PROJECT OVERVIEW AND OBJECTIVES Prior to the 1970s, in order to preserve museum specimens chemicals such as naphthalene, paradichlorobenzene, carbon tetrachloride, and metal salts were applied to cultural artifacts, as well as to materials found in herbarium, educational, and zoological collections. These chemical pesticides were used to reduce damage caused by insects and rodents during storage and display More recently identified is the long term persistence of metal based preservatives and, thus, the continued toxicity of treated spe cimens. Among the metal based chemicals applied was sodium arsenite (NaAsO 2 ), which is linked to cardiovascular diseases, neurological disorders, cancers, and, in some cases, death. The use of pesticides, and specifically arsenic, in the preservation of mu seum collections became a recognized issue of public concern with the 1990 enactment of the Native American Graves Protection and Repatriation Act, NAGPRA. NAGPRA requires all federally funded agencies to return cultural artifacts to original tribal owners Early repatriation efforts found that pesticide treated artifacts could not be returned to tribal owners without health risks due to the toxicity of the used, and still remaining, pesticide concentrations. Previous work in the Roane lab looked at the us e of metal resistant bacteria in the removal of mercury from museum collections. In this work, the bacterium Cupriavidus metallidurans CH34 was capable of reducing material associated (e.g. paper and textiles ) mercury through volatilization. In working with the Arizona State Museum ( Tucson AZ) and the Smithsonian Institution (Washington D.C.) on mercury impacted specimens it was discovered that a v ariety of collection materials at the Denver Museum of Nature and Science (DMNS) are heavily impacted by arsenic. This project aimed to design the framework for the

PAGE 12

2 bacterial removal of arsenic focusing on the development of arsenic quantification methods from a simulated museum material, paper, and to explore the pr esence of bacteria associated with collection specimens housed at the DMNS The specific objectives of the work reported here were to: Objective 1: Begin method development for the removal of arsenic from designated material type s using the arsenic resis tant bacterium Rhodoposeudomonas palustris CGA009 Objective 2: Identify the bacterial communities and arsenic presence on Denver Museum of Nature and Science specimens of different materials and storage collections.

PAGE 13

3 CHAPTER II BACKGROUND: USE O F ARSENIC VOLATILIZATION AND DEVELOPMENT OF ARSENIC QUANTIFICATION METHODS Overview Microorganisms, in particular bacteria, have been studied for their unique abilities to survive in a variety of environments that can be toxic to other forms of life. Not only are they able to survive in these environments, many bacterial species are capable of breaking down toxic chemicals and are being utilized in remediation processes. Among these toxic chemicals are heavy metals, which were used in preservation process e s in museums until the 1970s. Rhodopseudomonas palustris is a bacterium capable of reducing arsenic concentrations through th e process of volatilization. The long term goal of this study is to develop a bacteria based remediation technology to remove arsen ic from impacted museum collections, especially those earmarked for repatriation under NAGPRA (Native American Graves Protection and Repatriation Act). Bacterial Volatilization of Arsenic Arsenic compounds, such as arsenite (AsO 3 3 As(III)) and arsenate ( AsO 4 3 As(V)), are well recognized as detrimental to human health. In general, ingesting > 1mg of arsenic is considered toxic and can result in chronic renal failure, paralysis, coma and death (Odegaard 2005). At the cellular level, arsenic disrupts ATP p roduction, increases hydrogen peroxide production, and disrupts the function of enzymes (Shen 2013). As(III), in particular, impairs the function of proteins by interacting with their sulfhydryl groups while As(V) inhibits oxidative phosphorylation (Majum der 2013; Shen 2013). The combination of these events leads to cell death and, on a larger scale, can lead to organ failure.

PAGE 14

4 While arsenic is toxic to all biological systems, some bacteria have developed physiological mechanisms of dealing with arsenic t oxicity. As(III) typically enters bacterial cells through aquaglyceroporins, while As(V) can enter the cell through phosphate transporters (Figure 2.1 ). Once arsenic enters the cell, bacteria can use different mechanisms im mediate toxicity. The ars operon consisting of several genes found on chromosomal or plasmid DNA code for membrane associated transport proteins that pump arsenic out of the cell (coded for by the arsB gene); for an arsenate reductase to reduce As( V) to the less toxic As(III) (coded for by the arsC gene) ; and for S adenosylmethionine methyltransferase that methylates arsenic (coded for the arsM gene) (Oremland 2003).

PAGE 15

5 The methylation of arsenic is important to this project as the volatilization of arsenic into g as can decrease arsenic concentrations associated with a material. In the volatilization process, S adenosylmethionine (SAM) serves as a methyl group donor and the methylation pathway consists of several reduction and methylation steps that produce dimethy l and trimethyl arsine gases (Figure 2.2; Bentley 2002; Paez Espino 2009; Slyemi 2012; Vahter 2002). For remediation purposes, arsine gases can be collected and disposed of effectively reducing the associated concentration of arsenic. While arsenic has l ong been considered an easily biomethylated metal/metalloid (Bentely 2002), only a few studies have looked at arsenic volatilization as a remediation t echnology. One bacterium of particular interest for arsenic removal is Rhodopseudomonas palustris CGA009. The arsenite S adenosylmethionine methyltransferase enzyme has been well characterized in R. palustris (Qin et al. 2006; Yuan et al.2008). Qin et. al (2006) showed that when E. coli cells were genetically transformed with the arsM gene from R. palu stris and incubated with As(III) for 18 hours, there was a 4,600 g decrease in the amount of arsenic in the culture. Furthermore, gaseous products were trapped on filters and analyzed showing that the amount of volatilized arsenic correlated with the decr eased concentration of arsenic in solution In another study, the arsM gene from cyanobacteria was cloned into an E. coli vector conferring arsenic resistance to the otherwise non resistant E. coli up to 100 M As(III) (Yin 2011). Another study looking at arsenic contaminated soil showed that genetically engineered bacteria expressing the arsM gene isolated from R. palustris could remove 2.2 4.5% (approximately 42 mg/kg) of arsenic during a 30 day period (Liu et al. 2011). P reliminar y studies done in the R oane lab with R. palustris c ontaining the arsM gene, showed bacterial growth in media amended with arsenic concentrations as high as 250 ppm.

PAGE 16

6 Bacteria have long been used to mitigate chemical contamination due to their ability to transform chemicals. For example, bacterial remediation has become a cost effective accepted strategy for the degradation of hydrocarbons, diesel, and naphthalene (Fuentes 2014; Chen 2015; Huang 2016). For metal based contaminants, the fo cus of the work presented here, several studies show the potential for bacteria to transform metal based chemicals. For example, a study done by Dash et al. (2014) showed that Bacillus thuringiensis PW 5, isolated from the Odisha Coast, resisted concentrat ions of mercury as high as 50 ppm HgCl 2 and volatilized 90% of the mercury. In another study Cupriavidus metallidurans MSR33 removed 100% of HgCl 2 in solution following the addition of

PAGE 17

7 thioglycolate, which provides the thiol groups needed for the methylat ion (Rojas 2011). In a field based study with selenium, the ability of in situ soil bacteria to remove selenium from agriculturally contaminated waters through methylation was examined. When treated with a protein amendment, such as gluten or casein, 68 88 % of the total soil selenium was removed via volatilization from the first 15 cm of soil over a period of 100 months (Flury 1997). Quantification Methods Available for Arsenic Various methods exist for quantifying arsenic in human samples (i.e. hair, urin e, blood) as well as environmental samples such as soil, however, a standard method does not exist for materials encountered in the museum setting (i.e. paper, feather, fur). Numerous analytical instruments can be used for sensitive and ac curate determination of arsenic, such as Graphite Furnace Atomic Absorption Spectrophotometry (GFAA), Inductively Coupled Plasma Atomic Emission Spectrometry (ICP AES), Inductively Coupled Plasma Mass Spectrometry (ICP MS) Flame Atomic Absorption Spectrop hotometry (FAAS) among others with each instrument having advantages and disadvantages (Allen 1997; Nishimura 2010; Ulusoy 2013). Required for all analytical quantification methods is the digestion of the material, which is generally done in a solution o f one or more acids. For example, microwave digestion is used when quantifying arsenic in food (i.e. rice) and sediments, wherein the sample is dissolved in acid and/or h ydrogen peroxide during heating in a microwave (Allen 1997; Nishimura 2010; Pantuzzo 2 009; Zhou 1997). The microwave is thought to provide a more even and inte nse heating than other methods, such as digesting the sample in acid during boiling on a hot plate ( Boutakhrit 2004; dos Santos 2013). Additionally, the EPA has methods for digesting biological samples using acids (i.e. nitric acid, hydrochloric acid),

PAGE 18

8 combine d with other chemicals such as chloroform or methanol (Faroon 2002). This said, there are no studies available addressing the recovery and quantificatio n of arsenic from materials likely to be encountered in the museum setting, including paper, hide, feathers, and textiles. This observation necessitated the development of an arsenic digestion method, part of objective 1 of this project. Rhodopseudomonas palustris and Arsenic Rhodopseudomonas palustris CGA009 is a non pathogenic, facultatively Proteobacteria (Larimer 2004). R. palustris was first identified in 1907 by Austrian scientist Hans Molisch and further characterized in 1944 ( van Neil 1944). The rod shaped, Gram negative, purple non sulfur bacterium is ubiquitous and can be found in abundance in swine waste lagoons, pond water, and marine sediments (Larimer 2004). R. palustris has the ability to use different types of metabolis ms including photoheterotrophy, photosynthesis, chemoheterotrophy, and chemoautotrophy. R. palustris is also capable of metabolizing aerobically and anaerobically using a variety of electron donors, e.g. H 2 (Figure 2.3 ). Under aerobic conditions, R. palust ris can utilize chemoautotrophy (use of inorganic electron donors to fix CO 2 into biomass), especially under low nutrient conditions

PAGE 19

9 R. palustris is well known for its metabolic versatility. It has the necessary genes for carbon dioxide fixation, nitrogen fixation, and under anaerobic conditions it can generate energy from light via photosynthesis (Larimer 2004). R. palustris also has the ability to degrade organic compounds (i.e. sugars, lignin monomers), major components of plant and animal wastes. R. palustris can additionally dehalogenate and degrade chlorinated benzoates and chlorinated fatty acids (Egland 2001; McGrath 1997). The optimal temper ature range for R.palustris is 25 37 C and it can survive in a pH range of 5 9 with optimum being neutral pH (6.8 7). The metabolic flexibility of R. palustris makes it a good candidate for the arsenic remediation in the museum setting where environmental conditions may be unpredictable and nutrient availability limited R. palustris is also commonly studied for its ability to survive in the presence of arsenic (Qin 20066; Yin 2011; Zhao 2015). The presence of the ars operon allows the orga nism to extrude arsenic out o f the cell, carry out oxidation /reduction reactions with arsenic and volatilize arsenic through methylation. In particular interest in this study is arsenic volatilization for the removal of arsenic.

PAGE 20

10 NAGPRA and M useum Pestic ides In the 1980s museums were introduced to a new pest management technique called Integrated Pest Management (IPM). At this time, the widespread use of toxic chemicals as pesticides applied to specimens in museum s was being fully discovered and acknowle dged. The negative hea lth impacts of these chemicals ( r anging from organics to metals) were also being recognized (Odegaard 2005). IPM based storage methods, use of insect hormones to prevent reproduction, and application of tightly controlled environmenta l conditions, e.g. temperature or oxygen availability, were adopted as non chemical alternative s to toxic pesticides. In IPM, collection areas are frequently monitored using sticky traps to identify insect presence and to determine the specific IPM respons e needed for the appropriate protection of a specimen or entire collection While IPM has been largely successful in allowing museums to control material deterioration, older chemically treated specimens which make up a majority of museum collections, pres ent a large public health issue due to the presenc e of the still toxic originally applied pesticides. Metal based pesticides, in particular, show long term persistence as metals are not degradable. Chemical forms of arsenic and mercury are found on collec tions ranging from ethnographic, zoological, herbarium, and even fine and decorative arts (Sirois 2001). Work done by Sirois, at the Analytical Research Laboratory in Ontario, Canada, showed a shocking 81% of the natural history collections from five Canad ian museums tested positive for the presence of arsenic. Additionally, cultural items tested at the Arizona State Museum showed arsenic readings ranging from 3.17 g/cm 2 to as high as 3,132 g/cm 2 (Odegaard 2005). The previous use and associated toxicity concern s with arsenic are documented in museums throughout the U.S. and Canada, and are expected to exist worldwide.

PAGE 21

11 Early methods (prior to the 1970s) of material preservation for collection and storage pu rposes frequently applied arsenic as a 9% solution of sodium arsenite. Objects were often dipped in or sprayed with the solution (Odegaard 2005). Arsenic was also applied as a powder of sodium arsenite, the presence of which can still be seen on some colle ctions today as a white powder (Figure 2.4). Unfortunately, accurate and consistent records of the use of arsenic, frequency of use, concentrations used, and method of application were not regularly kept and are therefore not readily available for the iden tification and characterization of arsenic use on museum specimens This has forced the need for non destructive quantitative methods for arsenic detection on collection materials. A frequently used, despite a number of limitations, technology for the dete ction of arsenic on museum specimens is X ray fluorescence (XRF). While XRF can accurately detect the presence of arsenic, its ability to quantify arsenic concentrations and/or identify forms of arsenic associated with materials from wood to pottery to tex tiles remains in question. XRF is additionally expensive and not readily available in most museums. Because of this, most museums have to assume any item collected prior to the 1970s is impacted by toxic chemical preservatives, including arsenic.

PAGE 22

12 One of the impacted collections of most concern, as having been frequently treated with arsen ic salts, is ethnographic (cultural) specimens In 1990, Congress enacted the Native American Graves Protection and Repatriation Act (NAGPRA), which requires the return of cultural items to lineal descendants and affiliated Indian tribes. These items include human rema ins, funerary objects, sacred objects, and objects of cultural patrimony. Sacred objects religious leaders for the practice of traditional Native American religions by their present day Mr. Leigh Kuwanwisiwma from the Hopi Cultural Preservation Office, was unaware of the toxicity of sacred objects until 1995 when he was working on the repatriation of Katsina Friends, a Hopi spiritual art ifact. Kuwanwisiwma says The Hopi, the firs t tribe to physically repatriate items, such as Katsina Friends, were initially unaware of

PAGE 23

13 the contamination and potential toxicity risk. The Hopi repatriated approximately sixty items which wer e returned to cultural use in traditional ceremonies. As several tribal leaders became ill, tribes were told of the presence of chemical pesticides and tribal leaders were confirmed as having mercury and arsenic poisoning. A moratorium was placed on the physical return of i tems to tribes across the U.S. and Canada until the items are tested C urrent approaches for the removal of pesticides such as arsenic, include washing vacuuming, UV light degradation and containment; howev er, most cleaning methods show chemicals resulted in a maximum removal of 40% of the pesticides (Glastrup 2001). Methods, such as washing, cause deterioration of fragile artifacts. Light can cause fading in dyes or changes in the materials which are often not acceptable for conserva tion. Vacuuming seemed initially promising but much like the compressed air method, pesticides remained on the objects even after vacuuming In addition, vacuuming could be destructive. Another proposed method specific to mercury and arsenic was the applic ation of alpha lipoic acid (ALA) for mercury treated and dihydrolipoic acid (DHLA) for arsenic treated materials (Cross 2010). While the use of ALA and DHLA showed potential, their use requires the immersion of objects in liquid, resulting in possible det erioration and damage. Physical containment has also been proposed for treated collections, but is not realistic for item re use under NAGPRA. Finally, specific to cultural collections is consideration of tribal beliefs, as certain methods, such as exposi ng spiritual items to heat or UV light, may not be acceptable.

PAGE 24

14 CHAPTER III CONDUCTED STUDY : APPLIC ATION OF ARSENIC VOLATILIZATION BY Rhodopseudomonas palustris Abstract Metal based pesticides, including toxic arsenic salts, were widely used prior to the 1970s for the prevention of rodent and insect damage to museum collections. An estimated 80% of collections housed in U.S. and Canadian museums have exogenous metals present, and the use of bacteria in the removal of these metals is of interest. Resea rch shows that Rhodopseudomonas palustris a metabolically versatile proteobacterium, is able to volatilize arsenic via methylation resulting in the conversion of the metal into a gas which can be collected and removed. In this study, R. palustris tolerated concentrations of arsenic up to 250 ppm and showed the ability to remove up to 78% of arsenic from a starting concentration of 10 ppm soluble arsenic within 19 days. To show that R. palustris could be as effective at reducing arsenic concentrati ons from simulated museum materials, a method for arsenic quantification from collection materials was developed. Boiling reflux digestion showed the highest recovery of arsenic at greater than 80%, while other methods, e.g. microwave digestion, showed les s than 10% recovery on average. Introduction Arsenic is a highly toxic metal/metalloid historically used as a preservative for ethnographic and zoological collections. Stil l present today on historically treated artifacts, arsenic continues to pose a hea lth risk. Commonly applied as sprays, pastes, dips and powders, arsenic is a non degradable persistent pesticide (Odegaard 2005). While arsenic effectively protects materials from microbial, insect and rodent damage, with the recognition

PAGE 25

15 of its human toxic ity in the 1970s, its use as a museum preservative has been replaced with safer integrated pest management practices. Current mitigation for arsenic on museum collections includes vacuuming and washing (Spencer 2000). These methods, however, are not widel y used because of their harsh impact on the material; their cultural insensitivity; and/or their ineffectiveness on a broad range of material types (Spencer 2000). The use of bacterial arsenic removal methods is being actively investigated as a possible re mediation technology. The use of bacteria has several advantages, including their potential for controlled growth; their ability to facilitate volatilization and physical removal of arsenic; and their cultural acceptability (Roane 2010). Present ed in this study is the early identification and characterization of the bacterium Rhodopseudomonas palustris in the removal of arsenic from arsenic amended solutions. abilit y to remove arsenic from materials such as paper and textiles. R. palustris possesses the arsM gene, providing the organism with the ability to volatilize arsenic. The volatilization of arsenic, which converts the metal into a gaseous form, has several ad vantages from a remediation standpoint. Gaseous arsenic can be collected and disposed of properly, and provides an avenue for the reduction of m aterial associated arsenic concentrations, especially important given the non degradability of metals such as arsenic. Reduction of the arsenic concentrations associated with specimens throughout the museum will provide for safer handling. Driven by the toxicity of arsenic initially identified in the repatriation of tribal artifacts to the Hopi Tribe under NAGPRA (Native American Graves Protection and Repatriation Act, enacted in 1990), active research is looking into the development of an effective, broadly accep table, method for arsenic remediation in this

PAGE 26

16 unique setting. The objectives of the work presented here were (1) to characterize the initial ability of R. palustris to volatilize and remove arsenic, and (2) to develop an analytical method for arsenic quant ification, previously non existent for materials f ound in the museum environment to provide the framework for future studies addressing the ability of R. palustris to remove arsenic from museum materials. Methods Confirmation of Arsenic Removal by R. pal ustris In order to determine the arsenic removal capabilities of Rhodopseudomonas palustris initial studies focused on testing the removal from arsenic amended tryptic soy broth Tryptic soy broth was amended with 10 ppm arsenic (17. 3 ppm sodium arsenit e, As(III)), and inoculated with >10 5 R. palustris cells. No inoculum, arsenic amended control media were used to monitor sources of abiotic loss of arsenic. I ncubation on a rotary shaker (150 RPM) at 25 C was performed. Broth samples were prepared for arsenic quantitative analysis by dilu ting to a 40 ppb concentration in a 2.5% nitric acid solution to a total volume of 10 mL, and then read on a Graphite Furnace Atomic A bsorption Spectrophotometer (GFAA, PerkinElmer, Waltham, MA; described below). Development of an Arsenic Application and Detection Method for Simulated Museum M ater ials To develop a reproducible method for arsenic recovery and quantificat ion from arsenic treated museum mimicking materials (e.g. paper) several methods of material d igestion were evaluated.

PAGE 27

17 Method 1: Quantification of Arsenic From Paper Using Microwave Digestion In order to determine the concentration of arsenic associated with a museum material (i.e. filter paper as used in this project), it was necessary to first develop an effective method of digesting the material for the GFAA. Initial work looked at a method utilizing a microwave digestion system using nitric acid, a method previously used in the Roane lab for mercury quantification from filter paper and textile s For the included arsenic work, f ilter paper was dipped in a solution of 10 ppm arsenic (17.3 ppm sodium arsenite) and allowed to air dry. The amended paper was then placed in a microwave safe Teflon container and 10 mL of concentrated nitric acid was ad ded. The microwave (Remote Microwave Systems Floyd Inc., Lake Wylie, SC ) was set to 100% power for 2 minutes and 20 seconds. Samples were diluted as needed for GFAA analysis and arsenic quantification. Because arsenic loss was detected above, p ower and t ime on the microwave were adjusted to optimize arsenic recovery Para meters included 25% power for 2 minutes 20 seconds and 25% power for 1 minute 10 seconds Method 2: Quantification of Arsenic From Paper Using Acid D igestion Instead of dipping filter paper into a 10 ppm arsenic solution, 1 mL of the solution was pipetted on to the paper and allowed to air dry. In order to digest the paper prior to GFAA analysis 10 mL of concentrated nitric acid was placed in a 50mL test tube and placed on a vortexor fo r 10 minutes at 14,000 RPM. Following vortexing, the tubes were centrifuged for 2 minutes and diluted for analysis on the GFAA. Additionally, 25 ppm and 50 ppm arsenic solutions were applied to filter paper to test whether increasing the initial arsenic co ncentration increased the amount of arsenic recovered.

PAGE 28

18 Method 3: Quantification of Arsenic f rom Paper U sing Boiling Reflux Digestion A known concentration of an arsenic standard was applied to Whatman 3 filter papers by pipetting 1 mL directly to the surface of the paper. The filter paper was transferred to a 100 mL test tube and 7.5 mL of concentrated nitric acid was added in addition to a boiling chip. The test tube was placed on a hot plate and the temperature kept at ~95 C The tes t tube was covered with a glass bulb and was allowed to boil anywhere between 3 10 hours. MiliQ water was added to the test tube to bring the volume up to 100 mL. This diluted the amount of arsenic to fit on the standard curve of the GFAA. Graphite Furnace Atomic Absorption S pectrophotometry Prior to analysis, control flasks were created with known concentrations of arsenic in order to create a standard calibration curve to be used for calculating arsenic concentrations in sample flasks. An arsenic standar d (10,000 ppm arsenic) was used and diluted to seven concentrations (0 ppb (100 mL of 2.5% nitric acid), 5 ppb, 10 ppb, 25 ppb, 50 ppb, 75 ppb, and 100 ppb) to be used for the creation of a standard curve and the calculation of arsenic concentrations in th e samples. Calculated concentrations of arsenic were added to a solution of 2.5% nitric acid and a total volume of 100 mL. GFAA uses a graphite coated tube, known as a cuvette, to vaporize the sample to allow for the absorption of light at wavelengths cha racteris tic of the element of interest ( 195 nm for arsenic ) Sample (20 L) is deposited o nto the cuvette which gets heated to 2,500 C in order to vaporize and atomize the sample. The atoms then absorb the light and the instrument provides a numerical value that can be used to create a standard curve for calibration and sample determinations.

PAGE 29

19 Results Confirmation of Arsenic Removal by R. palustris Rhodopseud o monas palustris was able to grow in tryptic soy broth amended with arsenic concentrations as high as 250 ppm. To assess removal of arsenic, f lasks were inc ubated on a rotary shaker at 25 C for 7 days, 10 days, 15 days, and 19 days to determine whether arsenic removal i ncreased with time (Table 3.1). Flask Arsenic detected ( 7 days ) Arsenic detected (10 days) Arsenic detected (14 days) Arsenic detected (15 days) Arsenic detected (19 days) R. palustris only BDL BDL BDL BDL BDL Arsenic only 4.72(1.73) 6.72(.82) 8.73(.71) 9.97(.53) 7.32(.59) R. palustris & Arsenic 2.81(.53) 1.74(1.41 ) BDL 4. 73 (1.71 ) 1.58(1.1) % Removal 40% 74% 100% 53% 78% After 7 days of incubation, arsenic concentrat ions in treatment flasks showed 40 % removal of arsenic (Table 3.1) Following 10 days of incubat ion, the removal of arsenic increased to 74 %. The 15 day flasks showed 53 % removal of arsenic while the 19 day incubation showed th e highest removal at 78% (Figure 3 .1). One study incubated for 14 days s howed 100% removal of arsenic (Table 3.1) Arsenic only c ontrol flasks contained an average of 8.73 ppm arsenic while treatment flasks showed 0 ppm arsenic present upon analysis with GFAA for this particular 14 day experiment (Table 3.1). It is important to note T able 3.1 Arsenic concentrations remaining in individual flasks with (1) R. palustris in medium only, (2) 10 ppm arsenic amended medium, (3) R. palustris with 10 ppm arsenic in medium. The standard deviation is in parentheses and the mean is in ppm of arsenic. BDL= below detectable limit (<5 ppb).

PAGE 30

20 that each experiment was done independently, rather than one set of flasks quantified at different time periods, so analytical variation is to be expected.

PAGE 31

21

PAGE 32

22 Development of an Arsenic Application and Detection Method for Simulated M useum M ater ials While R. palustris showed potential for arsenic removal from broth, the ability of R. palustris to remove arsenic from solid material such as those likely to be encountered in the museum setting needed to be demonstrated. To do this, howev er, a new arsenic quantification and digestion method needed to be developed. The percent loss for arsenic was calculated based on the expected calculation in the digestate, which included the paper digested in 10 mL of concentrated nitric acid. Method 1 Resul ts: Quantification of Arsenic f rom Paper Using Microwave Digestion When the microwave (Remote Microwave Systems) was set to 100% power for 2 minutes and 20 seconds results showed that no arsenic remained on the paper while the expected concentration was 1 ppm in the digestate solution When the power on the microwave was reduced to 50% with a digestion time of 2 minutes and 20 seconds results continued to show a 100% loss or zero recovery of the arsenic (Table 3.2). Power and time on the mic rowave w ere adjusted (25% for 2 minutes 20 seconds and 25% for 1 minute 10 seconds ) and evaluated. While both times showed higher recovery for arsenic, the detected arsenic levels were less than those predicted and arsenic loss was between 90% and 100% (Table 3.2). Consistency was also problematic with e ach of the method modifications examined as can be seen by the high standard deviation s (Table 3.2).

PAGE 33

23 M ethod 2 Results : Quantifica tion of Arsenic f rom Paper Using Acid D igestion An acid digestion method was tested in order to determine if higher amounts of arsenic could be recovered from the material. This method did not show better recovery as the ave rage amount quantified was 0 .094 ppm w ith a standard deviation of 0 .020 ppm even though 1 ppm was expected in the digestate solution and 90.6 % loss of arsenic was obse rved (Table 3.3) The applied concentration of arsenic was increased to 25 ppm while all other parameters remained the same and the expected concentration was 2.5 ppm This increased the concentration recovered from the paper, however, there was still a l arge amount of inconsistency from one experiment to the next, as seen by the large standard deviation and 50 % loss of arsenic was observed (Table 3.3). When the applied concentration was increased to 50 ppm (an expected concentration in the digestate of 5 ppm) the observed concentration did not generally increase from what was observed with 25 ppm and loss increased to 90.9 % (Table 3.3). Arsenic expected 100% power 2 min 20 sec 50% power 2 min 20 sec 25% power 2 min 20 sec 25% power 1 min 20 sec 1 ppm 0.022 ppm ( 0 .037) BDL 0 .07 ppm ( 0 .111) 0 .039 ppm ( 0 .074) % Loss 97 .8% 100% 9 3% 96.1 % Table 3.2 Summary of mean arsenic detection for paper swatches under different microwave digestion parameters examined. Values reported are mean standard deviation. Table 3.3 Summary of microwave digestion parameters tried. The numbers represent an average of samples quantified using GFAA that were dipped in water only (control ) and dippe d in a 10 ppm arsenic solution. Standard deviation is in parentheses. BDL= below detectable limit. Treatment 100% power 2 min 20 sec 50% power 2 min 20 sec 25% 2 min 20 sec 25% 1 min 20 sec Dipped in water only (control) BDL BDL BDL BDL Arsenic recovered range BDL .05 ppm BDL BDL .286ppm BDL .228ppm Arsenic recovered average .022ppm (.037) BDL .069ppm (.111) .039ppm (.074) Table 3.2 Summary of microwave digestion parameters examined. The numbers represent mean arsenic detection for paper swatches in each treatment. Values reported are mean standard deviation. BDL=below detectable limit (<5 ppb).

PAGE 34

24 Method 3 Resul ts: Quantification of Arsenic f rom Paper U sing Boiling Reflux Digestion Digestion by boiling was the final method tried that showed better consistency from one sample to the next as well as concentrations that matched those applied to the material (filter pa per). The standard deviations were smaller than those observed with the other digestion methods and percent loss of arsenic was 20% or le ss ( Table 3.4). Measured Arsenic Measured Arsenic Measured Arsenic Arsenic Expected 1 ppm 2.5 ppm 5 ppm Arsenic mean 0.094 ppm (0.020) 1.25 ppm (0.76) 0.455 ppm(0.135) % Loss 90.6% 50% 90.9% Measured Arsenic Measured Arsenic Measured Arsenic Meas ured Arsenic Arsenic Expected 0.030 ppm 0.045 ppm 0.060 ppm 0.075 ppm Arsenic Mean 0.0240 ppm (0.005 44) 0.039 ppm (0.0041) 0.055 ppm (0.00198) 0.065 ppm (0.0019) % Loss 20.4% 13.1% 8.3% 12.9% Table 3.3 Summary of mean arsenic detection for paper swatches when various concentrations of arsenic were applied and digested in acid by vortexing. Values reported are mean standard deviation. Table 3.4 Summary of arsenic recovered with digestion method when different a rsenic concentrations were applied The numbers represent an average of samples quantified using GFAA with standard deviation is in parentheses. Table 3.3 Summary of microwave digestion parameters tried. The numbers represent an average of samples quantifi ed using GFAA that were dipped in water only (control) and dippe d in a 10 ppm arsenic solution. Standard deviation is in parentheses. BDL= below detectable limit. Table 3.4 Summary of arsenic detection for paper swatches when various concentrations of arsenic were applied and digested by boiling reflux digestion. Values reported are mean standard deviation. Figure 4.1. Folding pattern of the 16S rRNA gene and the conserved (green) and variable (V) regions. Image adapted from: http://rna.ucsc.edu/rnacenter/xrna/xrna_gallery.html and http://www.alimetrics.net/en/index.php/dna sequence analysis) Table 3.4 Summary of arseni c recovered with digestion method when different arsenic concentrations were applied The numbers represent an average of samples quantified using GFAA with standard deviation is in parentheses.

PAGE 35

25 Discussion Experiments in broth showed that R. palustris could remove 100% of arsenic within 14 days at 25 C Showing that R. palustris had potential as a remediation organism, the study shifted to applying arsenic to simulated museum materials (i.e. materials that museum collections are composed of). The work here attempted to develop a standard method for the quantification of arsenic assoc iated with materials such as paper, fur, and feather. The method of acid digestion via boiling was the most effective of those tried at recovering the mathematically predicted amount of arsenic. Future work will include testing the developed metho d on othe r material types, e.g. textiles, for recovery of expected arsenic concentrations for eventual use for monitoring the ability of R. palustris to remove arsenic from museum specimens. Because of the limitations of current arsenic removal methods, and the re sulting harsh treatment of materials, arsenic removal from museum specimens is not widely performed. Additionally, for cultural items earmarked for repatriation under NAGPRA, many tribes find the currently available methods culturally unacceptable. A bacte rial based approach has been that have been poisoned by chemicals (in particular metals). While much work remains to be done, Rhodopseudomonas palustris is an organism for consideration for arsenic removal from museum specimens.

PAGE 36

26 CHAPTER IV BACKGROUND: MICROBIOLOGY OF THE DENVER MUSEUM OF NATURE AND SCIENCE Overview Studies characterizing microorganisms in the museum setting are heavily focused on organisms that contri bute to degradation of material, and, to date, a wide scale characterization of bacterial communities associated with museum materials has not been done. This study aimed to elucidate these bacterial communities associated with different materia l types housed in different Collections Departments within the museum using high throughput sequencing, a technique that has not been utilized in museum studies. Additionally, the use of arsenic in past preservation methods has caused its persistence on individual collections items, another potential variable in shaping the resulting microbial communities. Knowledge of existing microorganisms on museum specimens will contribute to our understanding of material degradation but will also help shape the use of bacteria in pesticide removal for the safer handling of collections and for the repatriation of chemically treated artifacts under NAGPRA. Preservation M ethods Taxidermy is the process of preparing, stuffing, and mounting the skins of animals for disp lay. Through the 1970s a common preservative used was arsenic, usually in the form of sodium arsenite. For taxidermy applications, the arsenic powder was dissolved by boiling in a solution of water and alcohol (added in unknown amounts). In addition, arsen ic powder was combined with a flour paste, among other chemicals, to create a papier mache mixture that was used as an adhesive throughout the preparation process. To prepare bird or mammal

PAGE 37

27 mounts, the first step included skinning the animal and washing th e skin in a solution of ammonia water, followed by alcohol and then in a solution of water and benzene (Pray 1923). Following drying, arsenic solutions were applied to the inner and outer surfaces of the skin depending on the specimen, in the form of liqu id or powder. The skin was then stuffed with wire and wood wool and mounted for display (Pray 1923). Microbiology of M useums Work done surrounding the microbiology of the museum setting is heavily focused on organisms that contribute to biodeterioration and, in particular, the deterioration associated with important historical documents. These documents are often parchment (collagen) or paper (cellulose) making them targets for microbial degradation (Lech 2016). The fungi are common degraders, including m embers of the genera Aspergillus Penicillium Cladosporium and Trichoderma (Lech 2016; Sterflinger 2012). An assessment of parchment paper documents in Krakow Poland revealed the presence of these common fungi as well as various bacterial strains includin g Bacillus spp ., Micrococcus spp ., Staphylococcus spp ., and Pseudomonas spp Other studies looking at wood, textiles, and stone (i.e. fossils) revealed the presence of various bacterial and fungal species. A study looking at wood at a museum in Auschwitz showed that fungi such as Aspergillus Penicillium and Cladosporium wer e present, as were various bacterial species including Bacillus spp (Kozirog 2014). Textiles as well as stone sampled across museums in India and Poland also de tected common fungal deterioration culprits such as Aspergillus as well as bacteria Microccous spp ., Staphyloccocus spp ., Pseudomonas spp ., Bacillus spp ., and Actinobacteria (Biswas 2013; Blyskal 2014; Sterflinger 2013).

PAGE 38

28 O ther studies did an overall characterization of microorganisms found in the museum setting by obtaining samples from the air. Interestingly, these studies found most of the same organisms as those found on the surfaces of wood, textiles, and stone with the addition of Acinetobacter spp and Lactobacillus spp (Gauzere 2014; Skora 2015). 16S rDNA and Sequencing Much of the work done characterizing the microbiology of the museum setting has relied on techniques like culturing, Denaturing Gradient Gel Electrop hore sis (DGGE), and 454 Sequencing. The current availabi lity of high throughput sequencing allows for a more comprehensive analysis of the microorganisms present on museum specimens. High throughput sequencing targets the 16S rDNA gene which is highly cons erved across bacteria and is commonly used for constructing phylogenies and identifying bacterial communities in a wide variety of environments. While the gene is found in all bacterial species, it consists of variable regions specific to genus and species allowing for identification upon sequencing (Figure 4.1; Clarridge 2004).

PAGE 39

29 Illumina DNA sequencing technology allows for the consecutive sequencing of multiple DNA molecules in a mixed sample (Caporaso et al. 2011). The hypervariable V4 region of the 16S rDNA is amplified using a fo rward primer, annealing to a conserved region of the gene, and a reverse primer containing a 12 Illumina adaptor and reverse primer pad and linker (Figure 4.2). Each DNA molecule in a sample library (a library is a sample containing multiple sequences) requires the use of a different barcode on the reverse primer, allowing for the separation and reading of multiple DNA sequences. Figure 4.1. Folding pattern of the 16S rRNA gene and the conserved (green) and variable (V) regions. Image adapted from: http://rna.ucsc.edu/rnacenter/xrna/xrna_gallery.html and http://www.alimetrics.net/en/index.php/dna sequence analysis)

PAGE 40

30 The library is prepared for loading on the MiSeq Sequencing System and a process called bridge amplifications of the DNA strands results in the formation of millions of unique clusters with thousands of copies of one amplicon (Figure 4.3). Following this step, DNA templates are copied using fluorescentl y labeled nucleotides which are then interpreted by a detector as specific nucleotides (Figure 4.4; The Illumina HiSeq 2000 Sequencing Technology 2015).

PAGE 41

31 S equencing data generated from high throughput 16S rDNA sequencing can be analyzed with QIIME (Quantitative Insights In to Microbial Ecology) and R (Caporaso et al. 2010 ; R Core Team 2014; Oksanen et al. 2015). QIIME allows for the fil tering of sequencing reads and assigning taxonomy. Additionally, it can be used to perform a variety of phylogenetic and statistical analyses including diversity determinations within samples (alpha diversity) and diversity among samples (beta diversity). Denver Museum of Nature and Science The Denver Museum of Nature and Science established in 1900, is located in Denver, Colorado, and currently houses 1.4 million items across 7 Collections/Departments. These include Anthropology, Library and Archives, Education, Geology, Health Sciences, Paleontology and Zoology. The items in each Collection are stored in rooms where humidi ty is not controlled and varies depending on the relative humidity of the outside air. Temperature of the Collections areas are kept at a standard room tem perature of approximately 25 C. In 2014, a new portion of the building was completed and many of the

PAGE 42

32 Departments moved to these new storage areas where humidity and tem perature can now be monitored. Th is study included 150 samples (individual swabs) from specimens (bird mounts, animal mounts, hides) housed in Z oology and Education Collections and 13 dust samples (Figure 4.5; Figure 4.6).

PAGE 43

33

PAGE 44

34 CHAPTER V CONDUCTED STUDY: MICROBIOLOGY OF THE DENVER MUSEUM OF NATURE AND SCIENCE Abstract High throughput sequencing revealed diverse bacterial communities associated with the museum specimens examined at the Denver Museum of Nature and Science. Bacterial community composition seemed strongly influenced by several factors including type of materials comprising the specimen, the presence of arsenic, and which museum Collection the specimen was h oused in. Organisms of high abundance across the items sampled included the genera Ralstonia Sediminbacterium Acinetobacter and the family Enterobacteriaceae Additionally, testing of surface associated arsenic revealed a wide range of concentrations on arsenic impacted items at the Denver Museum of Nature and Science. This study is the first to utilize high throughput sequencing techniques to characterize the bacterial communities associated with museum collections. Introduction Microorganisms play an important historic and current role in artifact quality and longevity. While known to participate in the degradation of materials, a comprehensive study of the microbial communities present on collection items throughout the museum setting is no t available Studies show the presence of fungal (i.e. Aspergillus Cladosporium ) and bacterial ( Bacillus Staphylococcus Pseudomonas ) species; however, few studies have looked at bacteria asso ciated with different types of specimens and hypothesized driving factors that shape these communities. The ai m of this study was to use high throughput 16S rDNA s equencing techniques in a diversity study of the bacterial communities present on

PAGE 45

35 zoological specimens at the Denver Museum of Nature and Science (Denver, CO). The De nver Museum of Nature and Science houses 1.4 million specimens across 7 Departments, with many of the specimens arsenic treated, providing a unique opportunity to evaluate how early arsenic preservation treatment has shaped the resulting bacterial communit ies. This study aimed to characterize not only the bacteria found on these specimens, but to also determine which factors have the greatest influence on surface associated bacterial diversity. Several factors were investigated, including the presence of ar senic, material type, and the museum Collection /Department the specimen belonged to. The associated museum Collection can be important because often specimens belonging to different Collections are stored in different ways and frequency of handling of item s can vary. Elucidating the communities that exi st on museum specimen collections could answer questions about effectiveness of preservation and storage methods. This study, in the long term, will contribute to the development of bacterial based remediatio n methods for the removal of early used toxic chemical preservatives on some museum specimens. Methods Sampling M ethodology In order to compare the influence of arsenic presence, material type, and Collection storage/handling on bacterial communities, individual items were chosen that were arsenic impacted and non impacted, represented different material types (hide s animal (non bird) mount s bird mount s ) and were from different Collections Departments (Zoology and Education) within the Denver Museum of Nature and Science Table 5.1 summarizes the number of items within each Collection of different material types as well as the range of arsenic detected on these items. Arsenic testing (described below) was performed for each

PAGE 46

36 bacterial sample collected. Overall, 150 individual samples for DNA extraction were collected from specimens across the two Collections and 13 dust samples throughout the museum Bacterial DNA Collection Method To collect bacterial samples from museum specimens, Whatman 42 filter paper, marked into four parts, was dipped into a solution of 0.1% glycerophosphate buffer. The damp paper was rubbed on the item in an area covering approximately 7 cm. The filter paper was immediately cut and a fourth of the paper was placed in 50 mL 0.1% glycerophosphate buffer for arsenic testing while the remaining paper was placed in a petri plate, bound with parafilm and placed in a cooler with dry ice. Each specimen was sampled in 2 3 different locations to assess bacterial diversity differences wit hin individual specimens. Dust Sample Collection Dust samples across the museum were also collected for bacterial and arsenic analyses. Thirteen dust samples were collected representing both public spaces and collection storage spaces. D ust was collecte d by wiping the top of door frames with damp Whatman 42 filter paper (as above) in e ach of thirteen spaces Arsenic Detection and Quantification As described above, a Whatman 42 filter paper was wiped along the sampling surface to collect bacterial DNA as well as surface associated arsenic residues. To wash the arsenic off the sample paper, the sample paper was then suspended in 50 mL 0.1% glycerophosphate. The one fourth filt er paper and buffer were vortexed for 5 minutes using the MO BIO Vortex adapter for 50 mL tubes (MO BIO Laboratories, Inc., Carlsbad, CA). Following vortexing, the solution was placed in a bottle and approximately 200 mL of MiliQ water was added.

PAGE 47

37 Arsenic t esting was performed on site upon collection using the LaMotte Arsenic Test Kit (LaMotte Company, Chestertown, MD). The LaMotte arsenic test kit instructions were followed to determine the arsenic concentration in the solution. The LaMotte testing kit can detect arsenic concentrations from 4 ppb to 2,000 ppb, and is used in a variety of applications for the quantification of aqueous arsenic. Bacterial Community Characterization In the laboratory, DNA was extracted from the sample paper using the MO BIO Po Laboratories, Inc., Carlsbad, CA). Following extraction, DNA was concentrated following the kit instructions with the modification s below Sodium chloride (5M, 250 L) was added to eac h sample and mixed by inverting. Ethanol (200 proof, 10.4 milliliters) was added to each sample, mixed by inverting and left on ice for one hour Samples were centrifuged for 30 minutes at 0 C and supernatant was removed. Approximately 20 mL of 70% ethanol was added and samples were centrifuged for 3 minutes at 4 C. Supernatant was removed and samples were left in a hood overnight. Following drying, 100 L of sterile molecular biology grade water (Thermo Fisher Scientific, Inc., Waltham, MA) was added to re suspend the resulting DNA pellet. Polymerase chain reaction (PCR) was performed using primers to target the V4 variable region of the 16S rRNA gene. The primers included F515NS4 (5' AAT GAT ACG GCG ACC ACC GAG ATC TAC ACT ATG GTA ATT GTG TGY CAG CMG CAA GCA GAA GAC GGC ATA CGA GAT each throughput sequencing. The revers e primer contained a 12 nucleotide Golay ba rcode, represented by X, that was specific for each

PAGE 48

38 sample allowing for multiplexing during the sequencing process (Caporaso et al. 2012). The PCR reaction mixture contained 10 L 5 PRIME HOT Master M ix (5 Prime, Gaithersburg, MD) [final reaction concentrations 0.5U Taq DNA polymerase; 22.5 mM KCl; 1.25 mM Mg 2+ ; and 100 M of each dNTP]; 200 nmol/L of F515NS4; 200 nmol/L of 806R; 200 ng of bovine serum albumen (BSA) (New Engl and BioLabs, Inc., Ipswich, MA); molec ular biology grade water (Thermo Fisher Scientific, Inc., Waltham, MA) ; and template DNA to a to tal volume of 25 L. The thermo cycler protocol included a 3 minute de naturation at 94 C; 30 cycles of a 4 5 sec ond denaturation at 94 C; a 60 second annealing step at 50 C; and a 90 s econd extension step at 72 C; followed by a ten minute final extension step at 72C; and storage at 4 C. To ensure successful amplification and the correct product size, agarose gel electrophoresis was performed using a 1% agarose gel stained with ethidium bromide. Once amplification was confirmed two replicate PCR reactions were performed for concentration and purification with t he Zymo DNA Clean and Concentrator Kit (Zymo Research, Irvine, CA), following the included protocol. Th e cleaned and concentrated DNA was quantified using the Qubit High Sensitivity dsDNA Assay Kit and Qubit 3.0 Fluorometer (Thermo Fisher Scientific, Inc., W altham, MA). The triplicate PCR reactions for each sample were pooled in equimolar ratios to achieve 20 nanograms of DNA. Pooled samples were concentrated and re quantified again to obtain a final library concentration in ng/L. Libraries were sent to the Genomics and Microarray Core at the University of Colorado Denver Anschutz Medical Campus (Aurora, CO ) for Illumina MiSeq 2x150 paired end sequencing.

PAGE 49

39 Computational Analyses Phylogenetic analyses were performed with QIIME (Quantitative Insights Into Microbial Ecology) (Caporaso et al. 2010). Paired end sequences were joined using the join_paired_ends.py command, which uses the fastq join method and a specified minimum overlap score of 30 was given (Aronesty 2011). Joined sequences were filtered (split_libraries_fastq.py command) and sequences with a Phred quality score of 29 were kept, which corresponds to a sequencing error rate of 0 .1%. Operational taxonomic units (OTU s) were picked using the pick_open_reference.py where reads were clustered together based on 97% or greater seq uence similarity. This comma nd also assigned taxonomy to OTU the August 2013 Greengenes bacterial and archaeal 16S rRNA database (DeSantis 2006). Chimeric sequences were identified using DECIPHER Find Chimeras Web tool and were removed from the da taset prior to further analysis (Wright 2012). Alpha diversity was performed in QIIME using the alpha_diversity.py command. R was used for the generation of heat maps and the phyloseq package in R was used for the generation of PCoA plots and statistical t esting (McMurdie 2015). Results Presence of Arsenic on Specimens Table 5.1 summarizes the types of specimens sampled and the corresponding arsenic concentrations. Each of the Educational animal mounts and hides tested below the detectable limit (4 ppb) for arsenic with the exception of specimen PB5758A, which was the top of the back of a red wolf hide testing at 60 ppb (Supplemental Table S1). Each of the 15 bird mount sample s collected from the Education C ollection showed arsenic concentrations ranging from 20 to 2,000 ppb (the maximum detectable concentration (Supplemental Table S2).

PAGE 50

40 From the Zoology Collection, animal mounts and hides, with and without arsenic, were examined. Arsenic ranges seen were from 20 1,000 ppb for animal mounts, while hides ranged from 20 80 ppb (Supplemental Table S3). Additionally, 5 animal mounts and 14 hides fro m the Zoology C ollection tested negative for the presence of arsenic (Supplemental Table S3). Bird mount samples in the Zoology C ollection where arsenic was detec ted ranged from 5 0 ppb to over 2,000 ppb (Supplemental Table S4 ). The bird mount samples from the Ed ucation C ollection in addition to the arsenic impacted Zoology samples showed the highest arsenic concentrations of all the items tested. Additionally, 27 b ird mount sampl es were taken from the Zoology C ollection where arseni c was not detected. Associated Collection/Department Material type sampled Arsenic Detection Education 5 animal mount 14 hide 1 hide 15 bird mount BDL BDL Detected Detected Zoology 15 animal mount 4 hide 23 animal mount 15 hide 27 bird mount 31 bird mount BDL BDL Detected Detected BDL Detected Table 5.1. Summary of specimens sampled and the corresponding arsenic ranges. BDL= b elow detectable limit (4 ppb) Figure 5.1 Heatmap showing relative abundance of organisms where arsenic was 4ppb). Dendrograms show average linkage of hierarchical clustering of Bray Curtis distances for samples on the y axis. ( A) Includes the two most abundant genera, Ralstonia and Sediminbacterium while (B) does not. Table 5.4. Summary of items sampled at the museum showing how many items of ea ch material type from the three C ollections. The arsenic range is shown in the third co lumn. BDL=Below detectable limit (4 ppb)

PAGE 51

41 Arsenic Impact on Alpha Diversity Alpha diversity t he diversity within samples, within arsenic treated and non arsenic treated bacterial communities was measured using the Chao1 estimate, Observed OTU Chao1 and Observed OTU richness measure the species richness, while Shannon Index measures species diversity. relationships with species richness. All metrics were higher for arsenic free samples when compared to samples where arsenic was detected. A t test comparing the average from each group showed a significant diversity difference exists between arsenic and no arsenic samples (p< 0 .01 ) (Table 5.2 ). Collection Impact on Alpha Diversity Alpha diversity was measured within specimens found in the two Collections (Education vs. Zoology) phylogenetic diversity and Shannon Index metrics. Di versity was higher in the Zoology Chao1 Observed OTU Richness Shannon Index Arsenic Present 542 (257) 293 (166) 15 (8) 3.3 (1.3) Arsenic Absent 805 (466) 467 (303) 24 (14) 4.3 (1.3) t test p value p<0.01 p<0.01 p<0.01 p<0.01 Table 5.2 Alpha diversity metrics for arsenic detected and non detected samples where Chao1 estimate, Observed OTU Shannon Index were performed and a t test was done to test if the mean value for arsenic detected samples was significantly different from non arsenic detected samples The data shown is the average of all samples belonging to each group with the standard deviation in parentheses. Tabl e 5.2 Alpha diversity metrics for zoology and education collection samples where Chao1 estimate, Observed OTU Shannon Index were performed and a t test was done to test if the mean value for each collection was significantly different. The data is shown as the average of all samples belonging to each group with the standard deviation in parentheses. Table 5.1 Alpha diversity metrics for arsenic detected and not detected samples where Chao1 estimate, Observ ed OTU Shannon Index were performed and a t test was done to test if the mean value for Each collection was significantly different. The data is shown as the average of all samples belonging to each group with the standard deviation in parentheses.

PAGE 52

42 C ollection (Table 5.3 ). A t test comparing the averag es from each group showed significant diversity difference s bet ween the Zoology and Education C ollections (p< 0 .01) except Material Impact on Alpha Diversity Alpha diversity comparing different source materials (hide, bird mounts, and animal mounts) was measured phylogenetic diversity and Shannon Index metrics. Diversity measures were higher for samples where arsenic was not detected compared to the same material type where arsenic was detected (Table 5.4 ). When comparing arsenic treated hide and animal mounts, a significant difference existed between the two materials for the Shannon index ( p< 0 .01 ) and OTU richness (p= 0 .048) but in the absence of arsenic none of the alpha diversity metrics were significantly different betwe en hide and animal mounts (p> 0 .01 ) (Table 5 .4 ). When comparing hide and bird mounts, all metrics were significantly diff erent between the two material types when arsenic was present (p< 0 .01 ) but were not significantly different when arsenic was not detected (p> 0 .01 ) (Table 5.4). The same pattern was observed when Chao1 Observed OTU Richness Shannon Index Zoology 703 (417 ) 399 (274) 20 (13 ) 3.94 (1.49) Education 506 (187 ) 270 (102) 16 (5 ) 3.19 (.603) t test p value Zoology/Education p< 0 .01 p< 0 .01 p= 0 .050 p< 0 .01 Table 5.3 Alpha diversity metrics for Zoology and Education Collection samples where Chao1 estimate, Observed OTU Shannon Index were performed and a t test was done to test if the mean value for each collection w as signi ficantly different. The data shown is the average of all samples belonging to each group with the standard deviation in parentheses. Table 5.3 Alpha diversity metrics for different material types with and without the presence of arsenic where Chao1 estimate, Observed OTU phylogenetic diversity, and Shannon Index were performed and a t test was done to test if the mean value for each collection was significantly different. The data is shown as the average of all samples belo nging to each group with the standard deviation in parentheses. D= arsenic detected, ND = arsenic not detected. Table 5.2 Alpha diversity metrics for zoology and education collection samples where Chao1 estimate, Observed OTU ic diversity, and Shannon Index were performed and a t test was done to test if the mean value for each collection was significantly different. The data is shown as the average of all samples belonging to each group with the standard deviation in parent heses.

PAGE 53

43 comparing bird and animal mounts a significant difference e xisted when arsenic was present (p< 0 .01 ) but not when arsenic was absent (p> 0 .01 ) (Table 5.4 ). Chao1 Observed OTU Rich ness Shannon Index Hide D 795 (284) 498 (223) 24 (10) 5.35 (1.34) Hide ND 592 (296) 365 (213) 20 (10) 3.98 (1.16) Animal Mount D 625 (265) 313 (134) 17 (6) 3.45 (.45) Animal Mount ND 1008 (549) 567 (351) 27 (14) 4.88 (1.59) Bird Mount D 413 (129) 213 (58) 12 (3) 2.60 (.480) Bird Mount ND 797 (425) 461 (293) 26 (16) 4.10 (1.09) t test p value Hide D /Animal Mount D p= 1 p= 0 .048 p= 0 .160 p< 0 .01 t test p value Hide ND /Animal Mount ND p= 0 .127 p= 0 .698 p= 1 p= 0 .974 t test p value Hide D /Bird Mount D p< 0 .01 p< 0 .01 p< 0 .01 p< 0 .01 t test p value Hide ND /Bird Mount ND p= 1 p=1 p=1 p=1 t test p value Bird Mount D /Animal Mount D p< 0 .01 p< 0 .01 p< 0 .01 p< 0 .01 t test p value Bird Mount ND /Animal Mount ND p= 1 p= 1 p= 1 p= 0 .868 Table 5.4 Alpha diveristy metrics for all samples of the same material grouping where and Shannon Index were performed and a t test was done to test if the mean value for each group was significant ly different. The data shown is the average of all samples belonging to each group with the standard deviation in parenthes es. D= arsenic dete cted (>20 ppb), ND = arsenic not detected (<4 ppb). Table 5.4. Summary of items sampled at the museum showing how many items of ea ch material type from the three C ollections. The arsenic range is shown in the third column. BDL=Below detectable limit (4 pp b) Table 5.3 Alpha diversity metrics for different material types with and without the presence of arsenic where Chao1 estimate, Observed OTU phylogenetic diversity, and Shannon Index were performed and a t test was done to test if th e mean value for each collection was significantly different. The data is shown as the average of all samples belonging to each group with the standard deviation in parentheses. D= arsenic detected, ND = arsenic not detected.

PAGE 54

44 Imp act of Arsenic on Bacterial Communities

PAGE 55

45 When comparing microbial communities among specimen samples where arsenic was Ralstonia had the highest abundance, while the genus Sediminbacterium was highest in samples without arsenic. Because Ralstonia and Sediminbacterium were highly abundant a heat map was generated without these genera to determine what other groups were highly abundant. A h igher abundance of the family Enterobacteriaceae was seen when arsenic was not detected, while Acinetobacter had a high abundance in samples where arsenic was detected (Figure 5.1A; Figure 5.1B). To evaluate the impact of arsenic presence on the specimen associated bacterial communities, the arsenic and no arsenic impacted specimens from the Z oology Collection were compared. Principal coordinate analysis ( PCoA ) plots of Bray Curtis dissimilarity sh owed that animal mount samples clustered together based on whether specimen associated arsenic was present (Figure 5.2A) Axis 1 explained 54.4% of the variation observed among samples and samples arrayed along this axis (Figure 5.2A). Samples were compare d from the same Collection (e.g. Education or Zoology) and o f the same material type (e.g. b ird mounts; animal mounts; hides) in order to detect the influence of the arsenic without the impact of the other variables being taken into considera tion. Analysis of similarity ( ANOSIM ) show ed that for animal mount samples, arsenic explained 96.3% of the variation observed with in the microbial communities (p= 0 .001) (Table 5.5). When looking at bird mount samples only, the same pattern w as observed samples with detected arsenic clustered together while those without arsenic clustered (Figure 5.2B). The first axi s explained 58.2% of the variation observed among the communities. ANOSIM showed that arsenic explained 87% of the variation observed in the bacterial communities (p<0.01) (Table 5.5).

PAGE 56

46

PAGE 57

47 Impact of Collection on Bacterial Communities When comparing bacterial communities between the E ducation and Zoology Collection specimens the genera Ralstonia and Sediminbacterium were abundant across both collections. Ralstonia was more abundant in the Zoology Collection while Sedimin bacterium was more abundant in the Education C ollection samples (Figure 5.3A) The next most abundant groups were the family Bradyrhizobiaceae amon g specimens in the Education Collection, and Enterobacteriaceae and Acinetobacter among specimens in the Zo ology C ollection (Figure 5.3 B ) ANOSIM R statistic P value Zoology bird mounts 0.8696 0.001 Zoology animal mounts 0.963 0 0.001 Tabl e 5 .5 ANOSIM tests for Bray Curtis dissimilarity comparing arsenic detected ( and non detected (<4 ppb) Zoology bird and animal mounts to d etermine whether arsenic impacted bacterial communities. Tabl e 5.5 ANOSIM tests for Bray Curtis dissimilarity comparing arsenic detected ( and non detected (<4 ppb) samples to determine whether arsenic impacts bacterial communities.

PAGE 58

48

PAGE 59

49 To evaluate the impact of Collection on the specimen associ ated bacterial communities, arsenic un impacted animal mounts, non arsenic impacted hides and arsenic impacted bird mounts from each Collection (Education vs. Zoology) were compared. PCoA plots of Bray Curtis dissimilarity showed microbial similarities amon g similar specimen types (animal mounts vs. hide) within their Collection location (Education vs. Zoology). For animal mount samples, Axis 1 explained 36.2% of the variation observed and samples arrayed along this axis (Figure 5.4A). All samples compared t ested negative for arsenic and were of the same material type in order to detect the influence of the Collection without the impact of the other variables being taken into consideration. ANOSIM statistical test ing showed that for animal mount sam ples, Coll ection explained 77.9 % of the variation observed with the bacterial communities (p= 0 .001) (Table 5.6). When looking at hide samples that tested negative for arsenic, clustering was seen by collection type among the first axis which explained 55.5% of the v ariation in bacterial communities (Figure 5.4B). ANOSIM statistical test ing showed that Collection type explained 80.7% of the variation (Table 5.6). Arsenic impacted bird mounts showed some clustering by Collection but many Education and Zoology samples a lso clu s tered together. The first axis of the PCoA explained 75.8 % of the variation in bacteri al communities, and ANOSIM showed tha t Collection type explained 19.8 % of the variation (p= 0 .029) (Figure 5.5; Table 5.6).

PAGE 60

50

PAGE 61

51 Impact of Material on Bacterial Communities When comparing bacterial communities among material types (bird mount, animal mount and hide), the genus Ralstonia was most abundant on animal mounts, while Sediminbacterium was the most abundant on bird mount samples (Figure 5.6A). When these two highly abundant genera were removed from analysis, the family Enterobacteriaceae was ANOSIM R statistic P value Animal mounts 0 .778 0 .001 Hide 0 .807 0 .001 Bird mounts 0 .198 0 .029 Table 5.6 ANOSIM tests for Bray Curtis dissimilarity comparing animal mounts, hides and bird mounts from Education and Zoology Collections to determine the influence of Collection on bacterial communities. Table 5.6 ANOSIM tests for Bray Curtis dissimilarity comparing arsenic detected and non detected (<4 pp b) samples to determine whether arsenic impacts bacterial communities.

PAGE 62

52 highest on bird mount samples while Acinetobacter was highest in animal mount samples (Figure 5.6B)

PAGE 63

53 To evaluate t he impact of material type on the specimen associated bacterial communities Zoology non arsenic impacted animal vs. bird mounts, Zoology arsenic impacted animal vs. bird mounts, Zoology arsenic impacted animal mount vs. hide, Zoology arsenic impacted bird mount vs. hide, and Education non arsenic impacted animal mount vs. hide were compared. PC oA plots of Bray Curtis dissimilarity showed microbial similarities among similar material types (hide, animal mount, bird mount) within their Collection location (Zoology). All samples compared tested negative for arsenic (<4 ppb) and w ere from the Zoology Collection in order to detect the influence of the material type without the impact of the other variables being tak en into consideration. When comparing animal and bird mounts from the Zoology collection, axis one explained 34.8% of the variation and ANOSIM showed that material explained 80.1% of the variation observed within the bacterial communities (p=0.001) (Table 5.7). Bird and animal mount, arsenic impacted (> 20 ppb) samples showed clustering by material type on a PCoA plot and the first axis explained 83.5 % of the variation in bacterial com mu nities. ANOSIM showed that material explained 97% of the variation (p= 0 .00 1) (Figure 5.7B; Table 5.7).

PAGE 64

54

PAGE 65

55 When looking at hide and animal mount samples from the Zoology Collection that tested positiv e for arsenic, the first axi s of the PCoA plot explained 62.4% of the variation in bact erial communities and ANOSIM statistical test ing showed that material explained 94.6% of this variation (p=.001) ( Figure 5.8A; Table 5.7). H ide and bird mount arsenic impacted samples from the Zoology Collection showed clustering by mat erial type and the first axi s of the PCoA plot explained 47.7 % of the vari ation in bacterial communities while ANOSIM statistical test ing showed that 74.3% of the variation was explained by material type (p= 0 .001)(Figure 5.8B; Table 5.7).

PAGE 66

56

PAGE 67

57 PCoA plot of Bray Curtis dissimilarity showed that hide and animal mount samples from the Education Collection (not impacted by arsenic), the first two axis of the PCoA plot explained 83.8 % of the variation in bacterial communities (Figure 5.9). However, ANOSIM statistical test ing showed that material type did not explain the variation seen in the bacterial communities suggesting another factor was driving the differences (p= 0 .443) (Table 5.7).

PAGE 68

58 Community Analysis Community analysis showed the consistent presence of the same genera in all arsenic impacted samples, while different genera were consistently present in all non arsenic impacted samples. W hile many groups of bacteria were found across all collected sampl es (i.e. Ralstonia Acinetobacter Enterobacteriaceae Comamonadaceae ) all samples where arsenic was not detected included Bradyrhizobiaceae Pseudomonas Staphyloccocus Sediminbacterium and Sphingomonas (Table 5.8). Ralstonia was much more abundant in samples with arsenic (44%, as opposed to 6% in non impacted samples) while Enterobacteriaceae was more abundant in non impacted samples (12%, as opposed to 4% in impacted samples). Interestingly, all specimen types (bird, animal, hide) across both Coll ections despite the presence of arsenic, had Ralstonia and families Enterobacteriaceae and ANOSIM R statistic P value Arsenic Impacted Zoology Hide & Animal Mount 0 .946 0 .001 Zoology Hide & Bird Mount 0 .743 0 .001 Zoology Animal Mount & Bird Mount 0 .970 0 .001 Arsenic Non Impacted Zoology Animal Mount & Bird Mount 0 .801 0 .001 Education Animal Mount and Hide 0 .00084 1 0 .443 Table 5.7 ANOSIM tests for Bray Curtis dissimilarity comparing material types to determine whether material impacted bacterial communities. Table 5.8 Genera and families found in all samples belonging to each group; arsenic detected samples, arsenic non detected samples, zoology collection, and education collection. Table 5.7 ANOSIM tests for Bray Curtis dissimilarity comparing material types to determi ne whether material impacts bacterial communities.

PAGE 69

59 Comamonadaceae (Table 5.8). Geobacillus and Acinetobacter were unique to specimens within the Zoology Collection. Unique to Education Collections specimens were Sphingomonas and Pseudomonas and the families Caulobacteraceae and Bradyrhizobiaceae Dust at the Denver Museum of Nature and Science Dust samples across the museum were collected in 13 spaces (Table 5.9). All samples tested below detectable limits (<4 ppb) for the presence of arsenic No Arsenic Arsenic Education Collection Zoology Collection Ralstonia Enterobacteriaceae Bradyrhizobiaceae Acinetobacter Psuedomonas Staphylococcus Sediminbacterium Comamonadaceae Sphingomonas Ralstonia Enterobacteriaceae Acinetobacter Comamonadaceae Ralstonia Comamonadaceae Sphingomonas Enterobacteriaceae Bradyrhizobiaceae Caulobacteraceae Sediminbacterium Pseudomonas Ralstonia Enterobacteriaceae Acinetobacter Geobacillus Comamonadaceae Table 5.8 Bacteria consistently present on arsenic impacted samples, arsenic absent samples and samples from the Zoology and Education Collections. All genera and f amilies are higher than 1% abundance. Table 5.9 Dust samples all collected above doors showing area (public or collection storage space ) and location sampled. BDL=below detectable limit (<4 ppb). Table 5.8 Genera and families found in all samples belonging to each group; arsenic detected samples, arsenic non detected samples, zoology collection, and education collection.

PAGE 70

60 The bacterial communities in the dust revealed the abundance of the order Streptophyta as well as the genus Ralstonia (Figure 5.10). Also abundant were the genera Staphylococcus Streptococcus and Acinetobacter (Figure 5.11). Sample ID Arsenic Area Use Door Location DUST 3 BDL Collection Storage Figgins DUST 4 BDL Collection Storage Old Zoology DUST 5 BDL Collection Storage Old Zoology DUST 6 BDL Collection Storage Fur Vault DUST 15 BDL Collection Storage Education DUST 17 BDL Collection Storage Education DUST 18 BDL Public Explore Colorado DUST 20 BDL Public Explore Colorado DUST 24 BDL Public Expedition Health DUST 25 BDL Public Discovery Zone DUST 26 BDL Public Space Odyssey DUST 27 BDL Public Gems and Minerals DUST 30 BDL Public Anschutz Gallery Table 5.9 Dust samples collected from top of door frames in different use rooms across DMNS Museum location is also indicated. BDL=below detectable limit (<4 ppb). Figure 5.10. Heatmap showing relative abundance, including the Order Streptophyta and genus Ralstonia which were highly abundant in the majority of samples. Dendrograms showed average linkage of hierarchical clustering of Bray Curtis distances for samples on the y axis. Table 5.9 Dust samples all collected above doors showing area (public or collection storage space ) and location sampled. BDL=below detectable limit (<4 ppb).

PAGE 71

61

PAGE 72

62 Bacterial communities among the dust samples were explored using PCoA plots of Bray Curtis dissimilarity and showed that the first axis of the PCoA plot explained 42.4% of the variation and samples from public areas were closer to each other than to samples fro rm collection area s (Figure 5.12). However, ANOSIM statistical testing showed that a significant difference did not exist between the collection storage area samples and the public areas (p=0.08) (Table 5.10).

PAGE 73

63

PAGE 74

64 Alpha diversity was measured using the Chao1 estimate, Observed OTU richness, public area samples while all other metrics were higher in the samples belonging to collection are samples. However, a t test comparing the average from each group showed no significant difference (Table 5.11). Discussion This study characterized the bacterial communities found on specimens housed at the Denver Museum of Nature and Science. Specimens examined included arsenic im pacted/unimpacted; animal mounts bird mounts and hide s ; and were housed in both ANOSIM R statistic P value Area 0.1495 0.08 Chao1 Estimate Observed OTU Richness Shannon Index Public 2660 (614) 1324 (295) 75 (13) 6.03 (1.04) Collection 2425 (826) 1466 (463) 80 (23) 7.37 (2.06) t test p value 0 .600 0 .552 0 .647 0 .192 Table 5.10 ANOSIM tests for Bray Curtis dissimilarity comparing collection storage and public areas to determine whether area impacts bacterial communities. Table 5.11 Alpha diversity metrics for public and collection samples where Chao1 estimate, Observed OTU diversity, and Shannon Index were performed and a t test was done to test if the mean value for each area was significantly diff erent. The data is shown as the average of all dust samples belonging to each group with the standard deviation in parentheses. Table 5.10 ANOSIM statistical tests for weighted and unweighted UniFrac distances of area as a factor driving bacterial communi ties in dust Table 5.11 Alpha diversity metrics for public and collection samples where Chao1 estimate, Observed OTU were performed and a t test was done to test if the mean value for each area was significantly different. The data is shown as the average of all dust samples belonging to each group with the standard deviation in parentheses. Supplemental Table S1 Hide a nd animal mount samples from the Education Collection. Date acquired and origin of item information is unknown. BDL = below detectable limit. Table 5.11 Alpha diversity metrics for public and collection samples where Chao1 estimate, Observed OTU richness, were performed and a t test was done to test if the mean value for each area was significantly different. The data is shown as the average of all dust samples belonging to each group with the standard deviation in parentheses.

PAGE 75

65 Education and Zoology Collections. The presence of arsenic, the material type, and the Collection likely combine d to influence community presence and contribute to the differences observed among the bacterial communities. The presence of arsenic was more influential on animal mounts versus bird mounts and hides. Alpha diversity analyses (Chao1, Phylogenetic Diversity, and Shannon Index) were higher for samples where arsenic was absent (p<0.01 ), suggesting that arsenic decreases th e bacteria l diversity Bacterial community analys e s showed that four genera/families (e.g. Ralstonia Acinetobacter C omamonadaceae and Enterobacteriaceae ) were consistently associated with arsenic presence, and nine genera/families ( e.g Ralstonia Acinetobacter Sphingomonas Comamonadaceae ) were associated with the absence of arsenic. Many bacteria do not possess the n ecessary resistance mechanisms to allow them to live in the presence of arsenic so decreased divers ity on arsenic impacted samples wa s expected. Acinetobacter and Ralstonia have genes for arsenic resistance (Cai 2009; Mergeay 2003). Non impacted samples had higher abundances of human associated organisms ( Enterobacteriaceae family, Staphylococcus and Pseudomonas ) likely due to more handling of these items. While museums do not have records, many items acquired prior to 1970 are assumed to have been tre at ed with arsenic and are therefore tre a ted with caution. When comparing microbial communities among animal mounts with arsenic presence and absence as well as bird mounts with arsenic presence and absence, a significant difference existed (p=0.001) The C oll ection Department housing individual specimens additionally influenced bacteria l community. Alpha d iversity was higher with Zoology C ollectio n samples than Education (p<0.01 netic diversity (p=0.05). Microbiome analyse s

PAGE 76

66 identifie d two genera, Geobacillus and Acinetobacter present on Zoology specimens and absent from Education specimens. Eight bacterial groups were indicative of specimens housed within the Education Collection and Sphingomonas Bradyrhizobiaceae Caulobacteraceae Sediminbacterium and Pseudomonas were all unique to these specimens Acinetobacter can be associated with humans as it can cause nosocomial infections as well as the genus Pseudomonas (Jawad 1998). Many of the other organisms found in Education and Zool ogy Collection samples are commonly associated with soil and water. When comparing bacterial communities on arsenic non impacted animal mounts, hides and arsenic impacted bird mounts, all were significantly different based on the Collection they belonged t o (p< 0 .01 ). Diversity measures showed no significant difference s between communities associate d with hides and animal mounts. When arsenic was present, hide samples had significantly higher alpha diversity than bird mounts, but when arsenic was not detected, bird mount samples had higher alph a diversity (p<0.01 ) Bird and animal mounts were not significantly dif ferent when arsenic was absent. However, in the presence of arsenic, animal mounts were significantly more diverse (p<0.01 ) Whi le the materi al type influenced bacterial diversity, the presence or absence of arsenic with the material had a larger effect. Bird and animal mount specimens from both the Zoology and Education Collections were stored on shelves. Hide samples from the Zoology Collecti on were stored in closed cabinets, while the hides from the Education Collection were store d in plastic bags and bins. These could be contributing factors to the decre ased alpha diversity observed with Education Collection samples.

PAGE 77

67 Interestingly, dust in the museum revealed that many of the same groups of bacteria as on the specimens including Ralstonia Sphingomonas Pseudomonas Acinetobacter and family Enterobacteriaceae Dust samples from public areas compared to collection stor age areas did not show a significant difference. While many of the specimens are protected from dust in bags or in storage cabinets, dust could still be a potential source of the bacterial communities observed. Human related organisms (i.e. Enterobacteriac eae ) are also observed on the specimens and some are handled more often than others, which could contribute to the surface associated bacterial communities. The presence of bacteria on museum specimens is of particular interest from a biodegradation stand point as it is well know that bacteria and fungi are common culprits in material degradation in museums. Studies looking at the decay of wood collections in museum have identified many fungal and bacterial species (e.g. Cytophaga Cellvibrio Clsotridium Bacillus Arthrobacter Flavobacterium Spirillum Aspergillus Cladosporium Trichoderma ) involved (Kim 2000; Gelbrich 2008; Kretschmar 2008). The genera Bacillus Staphylococcus Pseudomonas Virgibacillus and Micromonospora have been implicated in the deterioration of parchment paper (Krakova 2012). Studies characterizing bacteria associated with feather degradation on birds focused on bacteria isolated from feathers with keratinolytic activity (the degradation of keratin). The studies showed that bacte ria from the Firmicutes phylum (i.e. Bacillus Staphylococcus ) and Actinobacteria and Proteobacteria ( Kocuria Micrococcus Streptomyces Pseudomonas ) were common degradation culprits (Burtt 2004; Gunderson 2008; Thys 2004). Studies looking at fur seals re vealed that the rectum contained species of Staphylococcus Escherichia Proteus Acinetobacter and Salmonella (Vedros 1982). Many of these same organisms were found in this study here and

PAGE 78

68 could potentially contribute to degradation, although this study d id not characterize the viability of the material associated bacteria. While other factors could also influence the bacterial communities on specimens across the Denver Museum of Nature and Science, this study showed the influence of the presence of arsenic, Collection storage, and material type.

PAGE 79

69 CHAPTER VI CONCLUSION AND FUTURE DIRECTIONS The bacterial communities associated with museum collections have not been characterized to date and this study is the first of its kind attempting to elucidate how different factors have contributed to these communities. F uture studies will expand on different material types and Collections such as including specimens from other Collections. For this study, while not enough data was collected for comparative analysis, preliminary data were collected from the Anthropology Co llection at DMNS (Supplemental Table S5; Supplemental Figure S1). Additionally, high throughput sequencing does not reveal whether bacteria present are viable and future studies can address the viability of organisms and therefore further elucidate the pot ential for degradation of material. The presence of arsenic has been one of the influential factors and testing various objects has shown that a wide range of arsenic concentrations still persist on these collections even though the use of arsenic was dis continued decades ago. This makes the matter of remediation an even more pressing issue and future work will continue to elucidate methods for utilizing R. palustris for the removal of arsenic associated with museum collections. The bacterium will be appli ed to various materials to emulate those seen in the museum (i.e. paper, feather, fur) to determine if it will be successful as a remediation technology in the overall museum setting. In addition, future work will explore collections in other museums as w ell as determining bacterial sources on specimens. This will include expanding the dust study to see if the overall museum microbiome is similar between dust and collection or if perhaps the

PAGE 80

70 bacteria are inherent to the organisms prior to the taxidermy p rocess. This work is just a small portion in an unexplored system that hopefully opens the door for future studies.

PAGE 81

71 SUPPLEMENTAL TABLES AND FIGURES Sample Artifact Type Area Arsenic (ppb) PB4275A Red fox Top back BDL PB4275B Red fox Mid back BDL PB4275C Red fox Low back BDL PBBOBA Bobcat Top back BDL PBBOBB Bobcat Mid back BDL PB5669A Coyote hide Top back BDL PB5669B Coyote hide Mid back BDL PB5669C Coyote hide Low back BDL PB6184A Coyote hide Top back BDL PB6184B Coyote hide Mid back BDL PB6184C Coyote hide Low back BDL PB6183A Coyote hide Top back BDL PB6183B Coyote hide Mid back BDL PB5747C Arctic fox hide Low back BDL PB5758A Red wolf hide Top back 60 PB5758B Red wolf hide Mid back BDL PB5758C Red wolf hide Low back BDL PB5753A Coyote hide Top back BDL PB5753B Coyote hide Mid back BDL PB5753C Coyote hide Low back BDL Supplemental Table S1 Hide and animal mount samples from specimens in the Education Collection. Date acquired and origin of item information is unknown. BDL = below detectable limit (<4 ppb). Supplemental Table S2 Bird Mount samples from the Education Collection. Date acquired and origin of item information is unknown. Supplemental Table S1 Hide and animal mount samples from the Education Collection. Date acquired and origin of item information is unknown. BDL = below detectable limit.

PAGE 82

72 Sample Artifact Type Area Arsenic (ppb) PB1119A Hooded merganser Back >2,000 PB1119B Hooded merganser Belly > 2,000 PB1119C Hooded merganser Cloaca >2,000 PB3085A Buffle head duck Back 60 PB3085B Buffle head duck Belly 40 PB3085C Buffle head duck Cloaca 50 PB0248A Bird Back 60 PB0248B Bird Belly 500 PB0248C Bird Cloaca 1,500 PB3088A Bird Back 20 PB3088B Bird Belly 40 PB3088C Bird Cloaca 20 PB3091A Bird Back 250 PB3091B Bird Belly 425 PB3091C Bird Cloaca >2,000 Supplemental Table S2 Bird Mount samples from specimens in the Education Collection. Date acquired and origin of item information is unknown. Supplemental Table S3 Hide and animal mount samples from the Zoology C ollection with arsenic detected. UNK = unknown. Supplemental Table S2 Bird Mount samples from the Education Collection. Date acquired and origin of item information is unknown.

PAGE 83

73 SampleID Artifact Type Area Arsenic (ppb) Date Acquired Origin of item DMNSZM2B Bison Mid back 150 2/14/1905 South Park, CO DMNSZM2C Bison Low back 1000 2/14/1905 South Park, CO ZM2397A Deer Top back 100 10/11/2025 Brazil ZM2397B Deer Mid back 60 10/11/2025 Brazil ZM2397C Deer Low back 60 10/11/2025 Brazil ZM2401A Deer Top back 100 11/19/2025 Brazil ZM2401B Deer Mid back 100 11/19/2025 Brazil ZM2401C Deer Low back 100 11/19/2025 Brazil ZM2480A Howler Monkey Top back 100 4/3/1905 UNK ZM2480B Howler Monkey Mid back 50 4/3/1905 UNK ZM2480C Howler Monkey Low back 50 4/3/1905 UNK ZM2479A Howler Monkey Top back 50 4/3/1905 UNK ZM2479B Howler Monkey Mid back 50 4/3/1905 UNK ZM2479C Howler Monkey Low back 80 4/3/1905 UNK ZM2441A Tapir Top back 50 11/7/2014 South America ZM2441B Tapir Mid back 50 11/7/2014 South America ZM4C Bison Low back 750 2/14/1905 UNK ZM2496A Jaguar Top back 60 9/11/2025 Brazil ZM2496B Jaguar Mid back 60 9/11/2025 Brazil ZM2496C Jaguar Low back 60 9/11/2025 Brazil ZM3A Bison Top back 250 2/14/1905 UNK ZM3B Bison Mid back 250 2/14/1905 UNK ZM3C Bison Low back 250 2/14/1905 UNK Z2334A Puma hide Top back 20 4/8/1905 Argentina Z2334B Puma hide Mid back 50 4/8/1905 Argentina Z2334C Puma hide Low back 20 4/8/1905 Argentina Z2335A Puma hide Top back 50 4/8/1905 Argentina Z2335B Puma hide Mid back 20 4/8/1905 Argentina Z2335C Puma hide Low back 20 4/8/1905 Argentina Z1933A Puma hide Top back 60 4/6/2018 Gardiner Mountain Z1933B Puma hide Mid back 60 4/6/2018 Gardiner Mountain Z1933C Puma hide Low back 60 4/6/2018 Gardiner Mountain Z6699A Jaguar hide Top back 60 UNK UNK Z6699B Jaguar hide Mid back 50 UNK UNK Z6699C Jaguar hide Low back 20 UNK UNK Z2717A Jaguar hide Top back 40 1/23/2026 Brazil Z2717B Jaguar hide Mid back 80 1/23/2026 Brazil Z2717C Jaguar hide Low back 80 ppb 1/23/2026 Brazil Supplemental Table S3 Hide and animal mount samples from specimens in the Zoology C ollection with associated arsenic levels UNK = unknown. Supplemental Table S4 Fur and animal mount samples from the Zoology collection with no arsenic detected. *BDL= below detectable limit (<4 ppb). UNK = unknown. Supplemental Table S3 Hide and animal mount samples from the Zoology C ollection with arsenic detected. UNK = unknown.

PAGE 84

74 SampleID Artifact Type Area Arsenic (ppb) Date acquired Origin of item DZTM3101A Raccoon Top back BDL UNK UNK DZTM3101B Raccoon Mid back BDL UNK UNK DZTM3101C Raccoon Low back BDL UNK UNK DZTM3174A Raccoon Top back BDL UNK UNK DZTM3174B Raccoon Mid back BDL UNK UNK DZTM3174C Raccoon Low back BDL UNK UNK DZTM3372A Fox Top back BDL UNK UNK DZTM3372B Fox Mid back BDL UNK UNK DZTM3372C Fox Low back BDL UNK UNK DZTM3360A Fox Top back BDL UNK UNK DZTM3360B Fox Mid back BDL UNK UNK DZTM3360C Fox Low back BDL UNK UNK DZTM3463A Beaver Top back BDL UNK UNK DZTM3463B Beaver Mid back BDL UNK UNK DZTM3463C Beaver Low back BDL UNK UNK Z6969C Bobcat hide Low back BDL 1920 1922 Gilpin, CO Z5200A Bobcat hide Top back BDL 4/23/1943 Kingsville, TX Z5200B Bobcat hide Mid back BDL 4/23/1943 Kingsville, TX Z5200C Bobcat hide Low back BDL 4/23/1943 Kingsville TX Supplemental Table S3 Continued. Hide and animal mount samples from specimens in the Zoology C ollection with associated arsenic levels. *BDL= below detectable limit (<4 ppb). UNK = unknown. Supplemental Table S5 Bird Mount samples from the Zoology C olle c tion with arsenic detected. UNK=unknown. Supplemental Table S4 Fur and animal mount samples from the Zoology collection with no arsenic detected. *BDL= below detectable limit (<4 ppb). UNK = unknown.

PAGE 85

75 SampleID Artifact Type Area Arsenic (ppb) Date acquired Origin of item Z10510A Red Eagle Back 250 3/26/2024 Littleton, CO Z10510B Red Eagle Belly 875 3/26/2024 Littleton, CO Z10510C Red Eagle Cloaca 250 3/26/2024 Littleton, CO Z10569A Red Eagle Back 150 10/3/2024 Littleton, CO Z10569B Red Eagle Belly 125 10/3/2024 Littleton, CO Z10569C Red Eagle Cloaca 150 10/3/2024 Littleton, CO Z15069A Gull Back 150 9/21/1935 San Pedro, CA Z15069B Gull Belly 80 9/21/1935 San Pedro, CA Z15069C Gull Cloaca 70 9/21/1935 San Pedro, CA Z21281A Gull Back 875 4/9/1940 Isabela Island, Mexico Z21281B Gull Belly 1,000 4/9/1940 Isabela Island, Mexico Z21281C Gull Cloaca 875 4/9/1940 Isabela Island, Mexico Z22178A Gull Back 1,000 3/21/1940 Adams County, CO Z22178B Gull Belly >2,000 3/21/1940 Adams County, CO Z22178C Gull Cloaca 1,000 3/21/1940 Adams County, CO ZB1078B Woodpecker Belly 750 10/15/2010 Jefferson County, CO ZB763B Pygmic Nuthatch Belly 80 UNK UNK ZB763C Pygmic Nuthatch Cloaca 50 UNK UNK ZB762A White Breasted Nuthatch Neck 750 UNK UNK ZB762B White Breasted Nuthatch Belly 150 UNK UNK ZB552B Woodpecker Belly 60 2/25/2010 Yuma County, CO ZB552C Woodpecker Cloaca 100 2/25/2010 Yuma County, CO ZB553A Woodpecker Neck 60 2/25/2010 Yuma County, CO ZB553B Woodpecker Belly 50 2/25/2010 Yuma County, CO ZB553C Woodpecker Cloaca 50 2/25/2010 Yuma County, CO ZB562A Belted Kingfisher Neck 250 5/10/2010 Adams County, CO ZB563C Belted Kingfisher Cloaca 100 5/18/2010 Adams County, CO ZB630A Brewers Blackbird Neck 150 9/28/2010 Weld County, CO ZB630B Brewers Blackbird Belly 425 9/28/2010 Weld County, CO ZB630C Brewers Blackbird Cloaca 425 9/28/2010 Weld County, CO ZB631B Brewers Blackbird Belly 500 5/10/2012 Weld County, CO Supplemental Table S4 Bird Mount samples from specimens in the Zoology C olle c tion with associated arsenic levels UNK=unknown. Supplemental Table S6 Bird Mount samples from the Zoology C ollection with no arsenic detected. BDL = below detectable limit (<4 ppb) UNK = unknown. Supplemental Table S5 Bird Mount samples from the Zoology C olle c tion with arsenic detected. UNK=unknown.

PAGE 86

76 Sample Artifact Type Area Arsenic Date acquired Origin of item ZB43144A Cooper's Hawk Back BDL 7/3/2009 Denver, CO ZB43144B Cooper's Hawk Belly BDL 7/3/2009 Denver, CO ZB43144C Cooper's Hawk Cloaca BDL 7/3/2009 Denver, CO ZB43266A Cooper's Hawk Back BDL 6/9/2009 Aurora, CO ZB43266B Cooper's Hawk Belly BDL 6/9/2009 Aurora, CO ZB43266C Cooper's Hawk Cloaca BDL 6/9/2009 Aurora, CO DZTB664A Dove Back BDL 6/18/2013 Golden, CO DZTB664B Dove Belly BDL 6/18/2013 Golden, CO DZTB664C Dove Cloaca BDL 6/18/2013 Golden, CO DZTB837A Red Dove Back BDL 3/6/2014 Longmont, CO DZTB837B Red Dove Belly BDL 3/6/2014 Longmont, CO DZTB837C Red Dove Cloaca BDL 3/6/2014 Longmont, CO DZTB1115A Bird Back BDL 9/27/2014 UNK DZTB1115B Bird Belly BDL 9/27/2014 UNK DZTB1115C Bird Cloaca BDL 9/27/2014 UNK DZTB1110B Duck Belly BDL 11/14/2014 UNK DZTB1110C Duck Cloaca BDL 11/14/2014 UNK DZTB1008A Western Grebe Back BDL 12/31/2014 UNK DZTB1008B Western Grebe Belly BDL 12/31/2014 UNK DZTB1008C Western Grebe Cloaca BDL 12/31/2014 UNK DTZB810A Red Dove Back BDL 8/3/2011 Denver, CO DZTB810B Red Dove Belly BDL 8/3/2011 Denver, CO DZTB114A Mud Hen Back BDL 11/28/2014 UNK DZTB114B Mud Hen Belly BDL 11/28/2014 UNK DZTB1111A Duck Back BDL 12/4/2014 UNK DZTB1111B Duck Belly BDL 12/4/2014 UNK DZTB1111C Duck Cloaca BDL 12/4/2014 UNK Supplemental Table S4 Continued Bird Mount samples from the Zoology C ollection with associated arsenic levels BDL = below detectable limit (<4 ppb) UNK = unknown. Supplemental Table S7 Feather samples from the Anthropology Collection. Date acquired is not known. BDL=below detectable limit (<4 ppb) UNK=unknown. Supplemental Table S6 Bird Mount samples from the Zoology C ollection with no arsenic detected. BDL = below detectable limit (<4 ppb) UNK = unknown.

PAGE 87

77 SampleID Artifact Type Area Arsenic Origin of item A552122AA Neck Ring Urubu Spot 1 BDL Brazil A552122AB Neck Ring Urubu Spot 2 BDL Brazil A552122AC Neck Ring Urubu Spot 3 BDL Brazil A7252AA Feather Headdress Moro Spot 1 BDL Gran Chaco Paraguay A7252AB Feather Headdress Moro Spot 2 BDL Gran Chaco Paraguay A7252AC Feather Headdress Moro Spot 3 BDL Gran Chaco Paraguay A7252BA Feather Headdress Moro Spot 1 BDL Gran Chaco Paraguay A7252BB Feather Headdress Moro Spot 2 BDL Gran Chaco Paraguay A7252BC Feather Headdress Moro Spot 3 BDL Gran Chaco Paraguay AC257A Headdress Jivaro Spot 1 250 ppb Ecuador AC257B Headdress Jivaro Spot 2 20 ppb Ecuador AC257C Headdress Jivaro Spot 3 20 ppb Ecuador AC5716B Dance Bustle Sioux Spot 2 BDL UNK AC5716C Dance Bustle Sioux Spot 3 BDL UNK AC5717A Bonet Spot 1 BDL UNK AC5717B Bonet Spot 2 BDL UNK AC5717C Bonet Spot 3 BDL UNK AC5718 Bonet Spot 1 BDL UNK Supplemental Table S5 Feather samples from the Anthropology Collection with associated arsenic levels. Date acquired is not known. BDL=below detectable limit (< 4 ppb) UNK=unknown.

PAGE 88

78

PAGE 89

79 REFERENCES Alimetrics [Internet]. Espoo Finland. [cited 2016 May 4]. Available from: http://www.alimetrics.net/en/index.php/dna sequence analysis. Allen LB, Siitonen PH, Thompson HC. 1997. Methods for the Determination of Arsenic, Cadmiun, Copper, Lead, and Tin in Sucrose, Corn Syrups and High Fructose Corn Syrups by Inductively Coupled Plasma Atomic Emission Spectrometry. J. Agric. Food Chem. 45:162 165. Aronesty E.2011. ea utils :Command line tools for processing biological sequencing data. http://code.google.com/p/ea utils Bentley R, Chasteen TG, Bentley R, Chasteen TG. 2002. Microbial Methylation of Arsenic Antimony and Bismuth. Microbiol. Mol. Biol. Rev. 2002, 66(2):250. DOI: 10.1128/MMBR.66.2.250 271.2002. Biswas J, Sharma K, Harris KK, Rajput Y. 2013. Biodeterioration agents: Bacterial and fungal diversity dwelling in or on the pre historic rock paints of Kabra pahad, India. Iran. J. Microbiol. 5:309 314. [Internet] 16:32 39. Available from: http://dx.doi.org/10.1016/j.culher.2014.01.008 Boutakhrit K, Claus R, Bolle F, Degroodt JM, Goeyens L. 2005. Open digestion under reflux for the determination of total arsenic in sea food by inductively coupled plasma atomic emission spectrometry with hydride generation. Talanta 66:1042 1047. Degrading Bacteria, and Color Variation Among Song Sparrows. Condor 106:681 686. Cai L, Liu G, Rensing C, Wang G. 2009. Genes involved in arsenic transformation and resistance associated with different levels of arsenic contaminated soils. BMC Microbiol 9:4. C aporaso JG, Kuczynski J, Stombaugh J, Bittinger K, Bushman FD, Costello EK, Knight R. (2010). Correspondence QIIME allows analysis of high throughput community sequencing data Intensity normalization improves color calling in SOLiD sequencing. Nature Publishing Group, 7(5), 335 336. doi:10.1038/nmeth0510 335 Cross PS, Odegaard N, Riley MR. 2010. Lipoic acid formulations for the removal of arsenic and mercury from museum artifact materials. J. Archaeol. Sci. 37:1922 1928. Dash HR, M angwani N, Das S. 2014. Characterization and potential application in mercury bioremediation of highly mercury resistant marine bacterium Bacillus thuringiensis PW 05. Environ. Sci. Pollut. Res. Int. 21:2642 53.

PAGE 90

80 DeSantis TZ, Hugenholtz P, Larsen N, Rojas M, Brodie EL, Keller K, Huber T, Dalevi D, Hu P, Andersen GL. 2006. Greengenes, a chimera -checked 16S rRNA gene database and workbench compatible with ARB. Appl. Environ. Microbiol. 72:5069 5072. dos Santos W NL, Cavalcante DD, Macedo SM, Nogueira JS, da Silva EGP. 2013. Slurry Sampling and HG AFS for the Determination of Total Arsenic in Rice Samples. Food Anal. Methods 6:1128 1132. Egland PG, Gibson J, Harwood C S. 2001. Reductive Coenzyme A Mediated Pathway for 3 Chlorobenzoate Degradation in the Phototrophic Bacterium Rhodopseudomonas palustris. 67:1396 1399. Faroon O, Harris OM. 2002. 7 Analytical Methods. Tox icol. profile DDT, DDE, DDD:265 277. Flury M, Jr WTF, Jury WA. 1997. Long term depletion of selenium from Kesterson dewatered sediments. The Science of the total environment. 198:259 270. Fuentes S, Mndez V, Aguila P, Seeger M. 2014. Bioremediation of petroleum hydrocarbons: catabolic genes, microbial communities, and applications. Appl. Microbiol. Biotechnol. 98:4781 94. Available from: Gauzere C, Moletta Denat M, Blanquart H, Ferreira S, Moularat S, Godon JJ, Robine E. 2014. Stability of airborne microbes in the Louvre Museum over time. Indoor Air 24:29 40. Gelbrich J, Mai C, Militz H. 2008. Chemical changes in wood degraded by bacteria. Int. Biodeterior. Bio degrad. 61:24 32. Glastrup J. 2001. The effectiveness of compressed air in removal of pesticides from ethnographic objects. Collection Forum. 16:19 24. Gunderson AR. 2008. Feather Degrading Bacteria: a New Fr ontier in Avian and Host Parasite Research? Auk [Internet] 125:972 979. Available from: http://www.bioone.org/doi/abs/10.1525/auk.2008.91008 Huang H, Wu K, Khan A, Jiang Y, Ling Z, Liu P, Chen Y, Tao X, Li X. 2016. Bioresource Technology A novel Pseudomonas gessardii strain LZ E simultaneously degrades naphthalene and reduces hexavalent chromium. Bioresour. Technol. [Internet] 207:370 378. Available from: http://dx.doi.org/10.1016/j.biortech.2016.02.015 Jaw ad A., Seifert H., Snelling A.M., Heritage, J., Hawkey P.M. 1998. Survival of Acinetobacter baumannii on dry surfaces: comparison of outbreak and sporadic isolates. J. Clin. Microbiol Vol(36). No. 7. pp. 1938 1941 Kim, Y.S., Singh, A.P., 2000. Micromorphological characteristics of wood biodegradation in wet environments: a review. IAWA Journal 21, 135 155.

PAGE 91

81 Kozirg A, Otlewska A, Piotrowska M, Rajkowska K, Nowicka Wolski GJ, Gu tarowska B, Kunicka organisms as a biodegradation factor affecting historical wood materials at the former concentration camp of Auschwitz II Birkenau. Int. Biodeterior. Biodegrad. 86:171 178. Kraker D. Native American News L Archives. [Internet]. RootsWeb. 2001. [cited 2014 June 1]. Available from: http://archiver.rootsweb.ancestry.com/th/read/NATIVE A MERICAN NEWS/2001 05/0989513395 Krakova L, Chovanova K, Selim SA, Simonovicova A, Puskarova A, Makova A, Pangallo D. 2012. A multiphasic approach for investigation of the microbial diversity and its biodegradative abilities in historical paper and parchment documents. Int. Biodeterior. Biodegrad. 70:117 125. Kretschmar EI, Gelbrich J, Militz H, Lamersdorf N. 2008. Studying bacterial wood decay under low oxygen conditions results of microcosm experiments. Int. Biodeterior. Biodegrad. 61:69 84. Larimer FW, Chain P, Hauser L, Lamerdin J, Malfatti S, Do L, Land ML, Pelletier DA, Beatty JT, Lang AS, Tabita FR, Gibson JL, Hanson TE, Bobst C, Torres JL, Harrison FH, Gibson J, Harwood CS. 2004. Complete genome sequence of the metabolically versatile ph otosynthetic bacterium Rhodopseudomonas palustris Nat. Biotechnol. 22:55 61. Lech, T. 2016. Evaluation of a parchment document, the 13 th century incorporation charter for the city of Krakow, Poland for microbial hazards. Appl. Environ. Microbiol. 82:2620 2631. Liu S, Zhang F, Chen J, Sun G. 2011. Arsenic removal from contaminated soil via biovolatilization by genetically engineered bacteria under laboratory conditions. J. Environ. Sci. 23:1544 1550. Majumder A, Bhattacharyya K, Kole SC, Ghosh S. 2013. Efficacy of indigenous soil microbes in arsenic mitigation from contaminated alluvial soil of India. Environ. Sci. Pollut. Res. Int. 20:5645 53. McGrath JE, Harfoot CG. 1997. Reductive dehalogenation of halocarboxylic acids by the phototrophic genera Rhod ospirillum and Rhodpseudomonas. Appl. Environ. Microbiol. 63:3333 3335. McMurdie PJ, Holmes S. 2015. Shiny phyloseq: Web application for interactive microbiome analysis with provenance tracking. Bioinfor matics 31:282 283. Mergeay M, Monchy S, Vallaeys T, Auquier V, Benotmane A, Bertin P, Taghavi S, Dunn J, Van Der Lelie D, Wattiez R. 2003. Ralstonia metallidurans, a bacterium specifically adapted to toxic me tals: Towards a catalogue of metal responsive genes. FEMS Microbiol. Rev. 27:385 410. N iel CBV. 1944. The culture, general physiology, morphology, and classification of the non sulfur purple and brown bacteria. Microbiology and Molecular Biology Reviews. 8(1):1

PAGE 92

82 Nishimura T, Hamano Nagaoka M, Sakakibara N, Abe T, Maekawa Y, Maitani T. 2010. Determination method for total arsenic and partial digestion method with nitric acid for inorganic arsenic speciati on in several varieties of rice. Food Hyg. Saf. Sci. 51:178 181. Odegaard N, Sadongei A. Old Poisons, New Problems A museum resource for managing contaminated cultural materials. Altamira Press. 2005. Simpson G, Solymos P, Stevens M, Wagner H. 2015. vegan: Community Ecology Package. Oremland RS, Stolz JF. 2003. The ecology of arsenic. Science 300:939 44. Pez Espino D, Tamames J, de Lorenzo V, Cnovas D. 2009. Microbial responses to environmental arsenic. Biometals 22:117 30. Pantuzzo FL, Silva JCJ, Ciminelli VST. 2009. A fast and accurate microwave assisted digestion method for arsenic determination in complex mining residues by flame atomic absorption spectrometry. J. Hazard. Mater. 168:1636 1638. Pray LL. Taxidermy. The MacMillan Company. 1923. Qin J, Rosen BP, Zhang Y, Wang G, Franke S, Zhangtt Y, Wangts G, Franket S, Rensingt C. 2006. Arsenic Detoxification and Evolution of Trimethylarsine Gas b y a Microbial Arsenite S Adenosylmethionine Methyltransferase. Proceedings of the National Acadmey of Sciences Vol(103). No. 7. pp. 2075 2080. R Core Team. 2014. R: A Language and Environment for Statistical Computing.Vienna, Austria. http://www.r -proje ct.org Roane TM, Snelling LJ.2010. Bacterial Removal of Mercury from Museum Materials: A New Remediation Technology? Pesticide Mitigation in Museum Collections: Science in Conservation. pp. 29 34. Rojas LA, Yez C, Gonzlez M, Lobos S, Smalla K, Seeger M 2011. Characterization of the metabolically modified heavy metal resistant Cupriavidus metallidurans strain MSR33 generated for mercury bioremediation. PLoS One 6:e17555. Shen Z, Han J, Wang Y, Sahin O, Zhang Q (2013) The Contribution of ArsB to Arsenic Resistance in Campylobacter jejuni PLoS ONE 8(3): e58894. doi:10.1371/journal.pone.0058894 Sirois JP. 2001. The analysis of museum objects for the presence of arsenic and mercury: Non destructive analysis and sample analysis. Collection Forum. 16: 65 75 Skra J, Gutarowska B, Pielech Pietrowski P. 2015. Assessment of microbiological contamination in the work environments of museums, archives and libraries. Ae robiologia (Bologna). 31:389 401. Slyemi D, Bonnefoy V. 2012. How prokaryotes deal with arsenic. Environ. Microbiol. Rep. 4:571 586.

PAGE 93

83 Spencer, B. San Francisco State University. San Francisco (CA): c2000 [cited 2014 June 1]. Sterflinger K, Pinzari F. 2012. The revenge of time: Fungal deterioration of cultural heritage with particular reference to books, paper and parchment. Environ. Microbiol. 14:559 566. Sterfl inger K, Piar G. 2013. Microbial deterioration of cultural heritage and works of art Tilting at windmills? Appl. Microbiol. Biotechnol. 97:9637 9646. The llumina HiSeq 2000 Sequencing Technology [Internet]. Heidelberg Germany: German Cancer Research Center; [updated 2015 Feb 24; cited 2016 May 4]. Available from: https://www.dkfz.de/gpcf/hiseq_technology.html. Thys RCS, Lucas FS, Riffel A, Heeb P, Brandelli A. 200 4. Characterization of a protease of a feather degrading Microbacterium species. Lett. Appl. Microbiol. 39:181 186. Ulusoy HI, Yilmaz Grkan R. 2013. A micellar improved method for trace levels selenium qu antification in food samples, alcoholic and nonalcoholic beverages through CPE/FAAS. Food Chem. 139:1008 1014. Vahter M. 2002. Mechanisms of arsenic biotransformation. Toxicology. 181 182:211 7. Vedros NA, Qu inlivan J, Cranford R. 1982. Bacterial and fungal flora of wild northern fur seals (Callorhinus ursinus). J.Wildl.Dis. [Internet] 18:447 456. Available from: http://www.ncbi.nlm.nih.gov/pubmed/7154218 Wong WT, Tseng CH, Hsu SH, Lur HS, Mo CW, Huang CN, Hsu SH, Lee KT, Liu CT. 2014. Promoting effects of a single Rhodopseudomonas palustris inoculant on plant growth by Brassica rapa chinensis under low fertilizer input. Microbes and Environment. 29:303 313. XRNA RNA Secondary Structure Files [Internet]. ci ted 2016 May 4]. Available from: http://rna.ucsc.edu/rnacenter/xrna/xrna_gallery.html. Yergeau E, Sanschagrin S, Beaumier D, Greer CW. 2012. Metagenomic analysis of the bioremediation of diesel contaminated Canadian high arctic soils. PLoS One Yin X X, Ch en J, Qin J, Sun G X, Rosen BP, Zhu Y G. 2011. Biotransformation and volatilization of arsenic by three photosynthetic cyanobacteria. Plant Physiol. 156:1631 8. Yuan C, Lu X, Qin J, Rosen BP, Le CX. 2008. Volatile Arsenic Species Released from Escherichia coli Expressing the AsIII S adenosylmethionine. Environ. Sci. Technol. 42:3201 3206. Zhao C, Zhang Y, Chan Z, Chen S, Yang S. 2015. Insights into arseni c multi operons expression and resistance mechanisms in Rhodopseudomonas palustris CGA009. Frontiers in Microbiology. 6:1 8. Zhou CY, Wong MK, Koh LL, Wee YC. 1997. Evaluation of high pressure microwave diges tion methods for hydride generation atomic absorption spectrometric determination of total selenium and arsenic in sediment. Mikrochim. Acta 127:77 83.