(Title of the Thesis)*

ICE ASSOCIATION IN MICROBES

Sandra Louise Wilson

A thesis submitted to the Department of Biology

In conformity with the requirements for

the degree of Doctor of Philosophy

Queen’s University

Kingston, Ontario, Canada

(September, 2010)

Microbes have a remarkable ability to adapt to a host of environmental stressors, including low temperature, high pressure and osmotic stresses. The adaptations of resistant microbes to low temperatures are varied, and may include the accumulation of solutes to maintain osmotic balance, the production of antifreeze proteins (AFPs) or ice nucleation proteins (INPs) to manipulate ice growth or formation. AFPs depress the freezing point, inhibit ice recrystallization, and have been reported to inhibit or delay the growth of gas hydrates. Conversely, INPs precipitate ice formation at relatively high subzero temperatures. Collectively, these activities can be described as ‘ice-association’ activities. Here, ice-affinity and/or freeze-thaw cycling were used to either select for isolates with ice association properties or to assess the low temperature resistance of microbial consortia derived from various environments. Ice-affinity successfully selected psychrotolerant microbes from cultured temperate and boreal soils, some of which had been previously reported in glaciers and Arctic/Antarctic sites. Many of the recovered microbes demonstrated ice-association activities. Freeze-thaw selection also greatly decreased the abundance and diversity of consortia from distinct sites, and allowed the recovery of individual isolates, many of which demonstrated ice-association. Freeze- thaw selection was also used to assess the role of cross-tolerance between osmotic and freeze-thaw stresses, based on the common challenge of desiccation. Microbial consortia from lakes with varying degrees of salinity were subjected to freeze-thaw stress, and the consortia from more saline lakes tended to show greater low temperature resistance.

While few of the recovered microbes demonstrated ice-association activities, those from

ii the more saline lakes tended to contain a higher intracellular solute concentration and were more likely to form biofilms. This underscores the diversity of resistance strategies and supports the notion of cross-tolerance. To determine if these selective regimes would have applications for hydrate growth inhibition, microbes derived from an oil well sample were subjected to freeze-thaw stress. Selection reduced microbial abundance, shifted the diversity, and resulted in the recovery of microbes with some ice-association activity. Taken together, this thesis demonstrates that the application of low temperature stress can be used to successfully investigate stress resistance mechanisms within microbial communities from distinct environments.

iii Co-Authorship

The following researchers contributed to the success of this thesis, and are therefore

greatly appreciated and are acknowledged with co-authorship.

Chapter 2

Deborah Kelley assisted with the ice-finger experiments, identification and screening of

the recovered microbes for ice-association activity as part of her undergraduate honours thesis.

Chapter 3

Dr. Paul Grogan kindly provided the Daring Lake soil samples as well as lab space.

iv Acknowledgements

First and foremost, huge thanks to Virginia, who is a wonderful supervisor and mentor! Without her constant support, guidance, encouragement and ‘outside of the box’ thinking, this thesis would not exist. I look forward to working as your peer in the future.

Of course, enormous thanks are happily due to my family. Mom and dad, thanks for fostering my love of science, encouraging and supporting me through this process.

Jen and Ryan, thanks for all your help, support and the constant adventures – you’re wonderful siblings, and I love you all.

Thanks to Bev for your friendly advice and constant thoughts, you’re a fabulous cousin. And dearest Gramps, where would I be without all the love and good times while you were with us?

To my closest friends, Sarah Bean, Jennifer May-Focht and Sabrina Mueller, thanks for the constant smiles and laughter, which kept me going. And to my favourite roomie and fantastic friend, Chantelle Peter, your support was more helpful then you could realize – thanks for the walks and talks. Lastly, to those at FBCK who supported me, thanks.

Finally, thanks to the bug-groupers: J. Affleck, A. Brown, M. Chalifoux, J. Choi,

H. Esposito, R. Gordienko, E. Huva, N. Kumar, K. Lauerson, S. Nowickyj, W. Qin, T.

Vanderveer, S. Wu, and the AFP-groupers, past and present for the suggestions and advice in lab meetings, and the good times in the lab.

v Table of Contents

Abstract ...... ii Co-Authorship ...... iv Acknowledgements ...... v Table of Contents ...... vi List of Figures...... ix List of Table...... xi List of Abbreviations: ...... xi Chapter 1 : General Introduction and Literature Review ...... 1 Literature Cited: ...... 3 Chapter 1.1: Selection of Low Temperature Resistance in Bacteria and Potential Applications .... 4 Abstract: ...... 4 Introduction: ...... 6 Challenges of Low Temperature and Osmotic Stresses: ...... 6 Adaptations to Hyperosmotic Stress: ...... 8 Adaptations to Low Temperature Stress: ...... 9 Antifreeze Proteins: ...... 10 Ice Nucleation Proteins: ...... 12 Freeze-Thaw and Ice-Affinity Selective Methods and Ice Assays: ...... 14 Equipment for Selection: ...... 14 Microbe Ice-Association Activities: ...... 16 Collection Sites and Culture Conditions: ...... 17 Isolate Identification: ...... 20 Experimental Selection and Resistance: ...... 21 Soil Communities and Isolates:...... 21 Lake Communities and Isolates: ...... 30 Biotechnological Implications: ...... 33 Acknowledgements: ...... 35 Literature Cited: ...... 36 Chapter 1.2: Gas Hydrates and Microbes ...... 45 Literature Cited: ...... 49 Chapter 1.3: Research Objectives ...... 52

vi Chapter 2 : Ice-Active Characteristics of Soil Bacteria Selected by Ice-Affinity...... 55 Abstract: ...... 55 Introduction: ...... 56 Materials and Methods: ...... 58 Results: ...... 63 Discussion: ...... 73 Acknowledgements: ...... 77 Literature Cited: ...... 79 Chapter 3 : Frequency of Freeze-Thaw Resistance in Microbial Communities Derived from Latitudinally-Distant Soils ...... 83 Abstract: ...... 83 Introduction: ...... 84 Materials and Methods: ...... 86 Results: ...... 94 Discussion: ...... 103 Acknowledgements: ...... 108 Literature Cited: ...... 110 Chapter 4 : Community Level Cross-Tolerance between Osmotic and Freeze-Thaw Stresses in Temperate Lakes ...... 116 Abstract: ...... 116 Introduction: ...... 117 Materials and Methods: ...... 118 Results: ...... 124 Discussion: ...... 134 Acknowledgements: ...... 138 Literature Cited: ...... 140 Chapter 5 : Freeze-Thaw Selection of a Cultured Oil Well-Derived Community ...... 145 Abstract: ...... 145 Introduction: ...... 146 Materials and Methods: ...... 147 Results: ...... 151 Discussion: ...... 156 Acknowledgements: ...... 158 vii Literature Cited: ...... 159 Chapter 6: General Discussion ...... 162 Low Temperature and Osmotic Stresses - Challenges and Resistance: ...... 162 Selecting for Microbes from Terrestrial Environments – the Role of Ice-Association Activities: ...... 163 Selecting for Microbes from Lake Environments – the Role of Cross-Tolerance: ...... 167 Selecting for Microbes with Hydrate-Association Activity: ...... 168 Recovered Isolates: ...... 169 Conclusions: ...... 170 Future Directions: ...... 171 Literature Cited: ...... 173 Appendix A : Preliminary Assessment for Cross-Tolerance Between Osmotic and Freeze-Thaw Stresses...... 179 Appendix B : The Use of Phage Display for Isolating and Cloning Putative-AFPs from Chryseobacterium sp. C14 and Pseudomonas borealis ...... 184

viii List of Figures

Figure 1.1. Schematics of the cryocycler and the ice-finger...... 15 Figure 1.2. Daring and Gould Lake consortia richness prior to and following selection. . 25 Figure 1.3. Abundance of the freeze-thaw resistant genera within the unselected Daring and Gould Lake consortia...... 27 Figure 1.4. Ice-association assays: ice recrystallization inhibition, ice-shaping, and ice nucleation assays...... 29 Figure 2.1. Colony diversity before and after ice-affinity selection...... 66 Figure 2.2. Ice-nucleation assessment of P. borealis...... 68 Figure 2.3. Ice recrystallization inhibition assays of bacterial isolates...... 69 Figure 2.4. Ice shaping assays of bacterial isolates...... 72 Figure 3.1. Phylogenetic tree of soil isolates prior to and following freeze-thaw...... 97 Figure 3.2. Soil consortia abundance vs. freeze-thaw cycles...... 100 Figure 3.3. Ice recrystallization inhibition and ice shaping assays of soil freeze-thaw resistant microbes...... 102 Figure 3.4. Ice nucleation activity of soil freeze-thaw resistant isolates...... 104 Figure 4.1. Lake water and sediment consortia abundance post freeze-thaw...... 125 Figure 4.2. Phylogenetic tree of lake isolates recovered post freeze-thaw...... 130 Figure 5.1. Produced water consortia abundance post freeze-thaw...... 152 Figure 5.2. RFLP analysis of gyrB clone libraries...... 153 Figure 5.3. DGGE analysis of the produced water consortia...... 155 Appendix A.1. Lake sediment consortia abundance post freeze-thaw ...... 181 Appendix A.2. Phylogenetic tree of lake isolates post freeze-thaw...... 183 Appendix B.1. PCR screen of LpAFP phage display library...... 186

ix List of Tables

Table 1.1. Summary of sample collection sites...... 18 Table 1.2. Microbial abundance following freeze-thaw selection...... 22 Table 1.3. Soil-derived enrichment culture richness following selection...... 24 Table 1.4. Lake sediment and water derived enrichment culture richness following freeze-thaw selection...... 32 Table 2.1. Ice-affinity selection of control bacteria...... 64 Table 2.2. Isolates recovered following ice-affinity...... 67 Table 2.3. Ice-association activities of the recovered isolates...... 70 Table 3.1. QPCR primer sequences and threshold cycles...... 91 Table 3.2. Consortia richness before and after freeze-thaw selection...... 95-96 Table 3.3. Coverage represented by the clone libraries...... 98 Table 4.1. Lake Sample Location and Lake Water Chemistry...... 119 Table 4.2. Lake consortia richness following freeze-thaw...... 127-129 Table 4.3. Lake isolate characterization assays...... 132-133 Table 6.1...... 164-165 Appendix A.1...... 182

x List of Abbreviations:

ANME-1/ANME-2.....anaerobic methane oxidizers AFP(s)...... antifreeze protein(s) CfAFP...... Choristoneura fumiferana antifreeze protein CFB...... Cytophaga-Flavobacterium-Bacteriodes CFU...... colony forming units CSP(s)...... cold shock protein(s) DGGE...... denaturing gradient gel electrophoresis DNA...... deoxyribonucleic acid E83...... Lake East of 83 medium EDTA...... ethylenediaminetetra acetic acid EMA...... ethidium monoazide bromide gyrB...... gyrase subunit B kb...... kilobase INA...... ice nucleation activity INP(s)...... ice nucleation protein(s) IR...... ice recrystallization IRI...... ice recrystallization inhibition LM...... Liberty Lake medium LpAFP...... Lolium perenne AFP LSD...... least significant difference MB...... marine broth mOsm...... milliosmoles MPN...... most probable number nm...... nanometer OD...... optical density PCR...... polymerase chain reaction PVP...... polyvinylpyrrolidone PYA...... peptone, yeast extract, alkaline buffer medium QPCR...... quantitative polymerase chain reaction RDP II...... ribosomal database project II xi RFLP...... restriction fragment length polymorphism rDNA...... ribosomal deoxyribonucleic acid rRNA...... ribosomal ribonucleic acid RT-QPCR...... reverse transcriptase, quantitative polymerase chain reaction sPGC...... saline Postgate C medium SRB...... sulphate reducing bacteria TH...... thermal hysteresis THF...... tetrahydrofuran Tris...... tris(hydroxymethyl) aminomethane TSB...... tryptic soy broth WfAFP...... winter flounder antifreeze protein

xii

Chapter 1: General Introduction and Literature Review

Microbes have shown a remarkable range of adaptations to extreme environments, ranging from acidophiles to alkalophiles and thermophiles to psychrophiles. Amongst the multiple challenges posed by low temperature environments is the presence of ice, and potentially, the related challenge of ice recrystallization (IR).

IR, is an energetically favourable process whereby the total surface area to volume ratio of ice is decreased resulting in fewer but larger, potentially damaging ice crystals (Mazur,

1966). Therefore, microbial adaptations which impede IR may be one of the strategies that confer a survival advantage in these environments. To date, two classes of microbial proteins have been identified that interact with ice: antifreeze proteins (AFPs), and ice nucleation proteins (INPs).

Microbes may produce AFPs, which adsorb to embryonic ice crystals, thereby inhibiting IR. A second key feature of AFPs, which also results from their adsorption onto ice crystals, is thermal hysteresis, or a decrease in the freezing temperature relative to the melting temperature (Raymond & DeVries, 1977; Knight & Duman, 1986).

Alternatively, some bacteria can initiate extracellular freezing, which may help prevent intracellular freezing, by the production of INPs (Zachariassen & Hammel, 1976; Xu et al., 1998). INPs form protein aggregates which are able to promote ice nucleation at high, subzero temperatures (Govindarajan & Lindow, 1988).

Gas hydrates are gas molecules enclathrated by a cage of stabilized water molecules, which can spontaneously form in high pressure and low temperature

1 conditions (Sloan, 1998; Kvenvolden, 1999). AFPs have recently been shown to hinder the growth of hydrates (Zeng et al., 2006; Gordienko et al., 2010), and may therefore be useful in mitigating the negative effects hydrates may exert when present in gas and oil pipelines.

The following sections more fully review the literature and introduce the thesis.

This consists of a commentary (Ch. 1.1), a review of gas hydrates and the role of AFPs as potential gas hydrate growth inhibitors (Ch. 1.2) and the thesis objectives (Ch. 1.3).

Literature Cited:

Gordienko, R., Ohno, H., Singh, V.K., Jia, Z., Ripmeester, J.A., and Walker, V.K. (2010). Towards a green hydrate inhibitor: imaging antifreeze proteins on clathrates. PLoS ONE 5: e8953.

Govindarajan, A.G., and Lindow, S.E. (1988). Size of bacterial ice-nucleation sites measure in situ by radiation inactivation analysis. Proc. Natl. Acad. Sci. USA 85: 1334- 1338.

Knight, C.A., and Duman, J.G. (1986). Inhibition of recrystallization of ice by insect thermal hysteresis proteins: a possible cryoprotective role. Cryobiology 23: 256-262.

Kvenvolden, K.A. (1999). Potential effects of gas hydrate on human welfare. Proc. Natl. Acad. Sci. USA 96: 3420-3426.

Mazur, P. (1966). Theoretical and experimental effects of cooling and warming velocity on the survival of frozen and thawed cells. Cryobiology 2: 181-192.

Raymond, J.A., and DeVries, A.L. (1977). Adsorption inhibition as a mechanism of freezing resistance in polar fishes. Proc. Natl. Acad. Sci. USA 74: 2589-2593.

Sloan, E.D. (1998). Clathrate Hydrates of Natural Gas (2nd Edition), Dekker, New York.

Xu, H., Griffith, M., Patten, C.L., and Glick, B.R. (1998). Isolation and characterization of an antifreeze protein with ice nucleation activity from the plant growth promoting rhizobacterium Pseudomonas putida GR12-2. Can. J. Microb. 44: 64-73.

Zachariassen, K.E., and Hammel, H.T. (1976). Nucleating agents in the haemolymph of insects tolerant to freezing. Nature 262: 285-287.

Zeng, H., Wilson, L.D., Walker, V.K., and Ripmeester, J.A. (2006). Effect of antifreeze proteins on the nucleation, growth, and the memory effect during tetrahydrofuran clathrate hydrate formation. J. Amer. Chem. Soc. 128: 2844-2850.

Chapter 1.1: Selection of Low Temperature Resistance in Bacteria and Potential Applications

Abstract:

Microbial consortia may harbour an array of resistance mechanisms that facilitate survival under harsh conditions, including antifreeze and ice nucleation proteins.

Antifreeze proteins lower freezing points as well as inhibit the growth of large,

potentially damaging ice crystals from small ice embryos. In contrast, ice nucleation

proteins prevent supercooling, and allow ice formation at high, subzero temperatures.

Psychrophiles and psychrotolerant microbes are typically sought in extremely cold

environments. However, given that geography is unlikely to present an insurmountable

barrier to microbial dispersal, we reasoned that species with low temperature adaptations

should also be present, although rare, in more temperate environments. In consequence,

the challenge then becomes one of selecting for rare microbes present in a larger

community. Following the commentary, we demonstrate that freeze-thaw survival and

ice-affinity selection have both been used to identify microbes that demonstrate low-

temperature resistance from enrichments derived from temperate environments.

Selection resulted in a drastic decrease in cell abundance and diversity, allowing the

isolation of a subset of resistant microbes. Depending on the origin of the consortia,

these resistant microbes demonstrated cross-tolerance to osmotic stress, or a high proportion of antifreeze and/or ice-nucleation protein activities. Both types of ice- association proteins presumably facilitate microbial survival at low temperatures. These

4 proteins, as well as molecules that maintain osmotic balance, are also of commercial interest, with applications in the food, energy and medical industries. In addition, the resistant phenotypes described here provide a glimpse into the breadth of strategies microbes use to survive and thrive at low temperatures.

Introduction:

It has been estimated that 80% of the biosphere is 15°C or lower (Kawahara et al.,

2001; Yamashita et al., 2002) with a great proportion at temperatures ideal for

psychrophilic microbes, and quite tolerable for psychrotrophs (Jacobs, 1957; Morita,

1975; Thieringer, 1998). In addition, 26% of terrestrial habitats are permafrost, soil with

temperatures of 0°C or colder for at least 2 years (Williams & Smith, 1989; Steven et al.,

2007). Even temperate climates experience subzero temperatures for at least a portion of

the year, and these regions as well as high latitudes and polar environments also are

subjected to seasonal freeze-thaw cycles. For instance, the active layer of tundra soil

microbes must contend with freezing conditions, followed by freeze-thaw cycling during

the spring thaw. Although this thaw may be accompanied by a peak in cell abundance,

this is ultimately followed by a rapid decline, possibly because of the effects of thawing

and/or subsequent temperature fluctuations (eg. Nelson and Parkinson, 1978). Freeze-

thaw resistance therefore is presumably crucial for survival. Likewise, the freezing and

thawing of soil consortia (ex situ) decreases microbial abundance and diversity (Walker

et al., 2006).

Challenges of Low Temperature and Osmotic Stresses:

Microbes in freezing environments must contend with an array of challenges

including decreases in protein synthesis, stability and function (eg. Beaufils et al., 2007;

Bennett et al., 1981) as well as decreased membrane fluidity (Chattopadhyay &

Jagannadham, 2001). Freeze-thaw stress can further challenge microbes by inducing

membrane, DNA and oxidative damage, fluctuating osmotic gradients, and intra- and extracellular ice formation (Mazur, 1966a; Alur & Grecz, 1975; Calcott & MacLeod,

1975b; Stead & Park, 2000). In addition, growing ice excludes solutes (Kuiper et al.,

2003), including the particles with which bacterial cells are often associated (Junge et al.,

2004). This solute exclusion results in an increased solute concentration in the liquid portion. Given that the majority of cells in frozen environments are present in brine veins

(Mader et al., 2006; Amato et al., 2009), freezing may result in cell desiccation or select for halophilic or halotolerant microbes.

The cumulative effects of these metabolic changes is a reduction in viability, however, the main target of freeze-thaw damage has been suggested to be compromised membrane integrity (Gomez Zavaglia et al., 2000). Nutrient availability, growth rates/phases, the duration frozen, and rates of freezing and thawing (Calcott & MacLeod,

1974a, 1974b; Gao et al., 2006) are all important determinants of survival. Indeed, internal ice formation at high cooling rates causes membrane damage, since there is limited removal of intracellular water (Mazur, 1966a). A theoretical equation illustrating the loss of water versus the rate of freezing and other parameters can be found in Mazur

(1963). Low cooling rates can result in increased solute concentrations and desiccation-

associated damage (Mazur, 1966a). Additionally, a larger number of smaller crystals are formed at high cooling rates compared to fewer and larger ice crystals seen at slower rates of cooling (Deal, 1970). Similarly, gradual increases in temperature or prolonged periods at near 0°C, are correlated with ice recrystallization (IR), resulting in large,

growing ice crystals. Faster thaw rates decrease IR since the crystals melt too quickly for

substantial IR to occur (Mazur, 1966a).

As previously mentioned, one challenge associated with subzero environments is desiccation, which can induce a number of similar physiological challenges as well as inhibit carbohydrate transport and DNA replication (Roth et al., 1985; Meury, 1988). As a result, low temperature and osmotically challenging environments present common physiological stresses (Ko et al., 1994).

Adaptations to Hyperosmotic Stress:

As indicated, low temperatures can result in osmotic stress due to the exclusion of solutes as water freezes. Adaptations to counter hyperosmotic stress thus benefit not only microbes found in saline environments, but also at low temperatures. Coping mechanisms can include alterations in membrane composition (Allakhverdiev et al.,

1999; Sakamoto & Murata, 2002). Genes involved in protein synthesis, folding, trafficking, fatty acid metabolism and transporters can also be affected by osmotic stress and these are important features of bacteria adapted to high saline environments (Li et al.,

2006). As with low temperature stress response, organic osmolytes such as amino acids are protective during osmotic stress (Shahjee et al., 2002; see below). Although inorganic salts can also help maintain osmotic balance, significant accumulation can result in cellular injury, which is avoided by the production of compatible solutes (solutes such as sugars, that are compatible to cellular metabolic activities; Yancey et al., 1982;

Ko et al., 1994).

Adaptations to Low Temperature Stress:

Microbes have developed a number of cryotolerant or cryoprotective strategies to counter the stresses posed by low temperature environments. Cryotolerance, an adaptation to freezing, can be induced in certain microbes by a shift to suboptimal, but not lethal, temperatures (Panoff et al., 2000), with surviving microbes upregulating cold shock protein (CSP) production (Jones et al., 1987). CSPs are present in the majority of

microbes, and may act as chaperones and during transcription and translation (Jones et

al., 1987; La Teana et al., 1991; Beaufils et al., 2007). Maintenance of membrane

integrity includes alterations in membrane components such as an increased proportion of

branched and unsaturated fatty acids (Klein et al., 1999; Gomez Zavaglia et al., 2000).

Cryoprotection on the other hand, is the result of cryoprotectants or

cryopreservatives that preserve cell viability at low temperatures (Panoff et al., 2000). In

general, these cryoprotectants function at low temperatures and freeze-generated high salt

concentrations to decrease the amount of ice formation, increase cell volume, stabilize

proteins and membranes and decrease permeability barrier damage (Mazur, 1966b;

Calcott & MacLeod, 1975a, 1975b; Izawa et al., 2004). Penetrating cryoprotectants such as glycerol, move across cell membranes and achieve osmotic balance. In contrast, non-

penetrating cryopreservatives such as trehalose, sucrose and other peptides or polymers

do not move across membranes. Additional cryoprotectants include lactose, sorbitol, and

also amines such as betaine, glycine betaine, arginine and proline (Shahjee et al., 2002;

Hubalek, 2003). Osmoprotectants, or compatible solutes, are low molecular mass organic molecules; the accumulation of which restores cellular growth rates by

countering water loss, maintaining the native state of proteins (Yancey et al., 1982;

Shahjee et al., 2002). One such osmoprotectant, glycine betaine, can accumulate to

molar concentrations. Osmotic or cold stress activates glycine betaine transporters in many organisms, but in others, including Listeria, the transporter may be constitutively expressed (Ko et al., 1994).

The aforementioned adaptations are used by microbes to tolerate freezing.

Microbes at subzero, though near zero temperatures can avoid freezing by decreasing ice nucleation temperatures. Alternatively, freezing can be regulated by controlling the freezing temperature and consequently the morphology of the ice crystals (Yamashita et

al., 2002). By decreasing cellular water content, via export and or osmotic potential

increase, the likelihood of internal ice formation is decreased, along with the potential for

injury and possibly death (Mazur, 1963). In addition, certain microbes can synthesize

proteins that interact with ice such as antifreeze proteins (AFPs) and ice nucleation

proteins (INPs). Each of these protein groups can be induced by low temperatures and

presumably are important for survival. They are also of commercial value.

Antifreeze Proteins:

Antifreeze proteins (AFPs), adsorb to embryonic ice crystals, likely because of

their structural complementary to the ice lattice (Knight et al., 1991; Jia et al., 1996).

While still debated, it has been argued that AFP adsorption to ice involves a combination

of one or more of the following: van der Waals forces (Sonnichsen et al., 1998),

hydrogen bonding (DeVries et al., 1977; Knight et al., 1991), or hydrophilic and

hydrophobic interactions (eg. Cheng & DeVries, 1991; summarized in Barrett, 2001). 10

Since AFPs adsorb to ice, AFPs demonstrate an affinity for ice, thus, ice-affinity

purification was developed in order to purify recombinant AFPs (Kuiper et al., 2003;

discussed following this commentary). Ice recrystallization (IR) naturally occurs when

ice is held at relatively high subzero temperatures, and results in the formation of larger

ice crystals at the expense of smaller crystals. This decreases the volume to surface area

ratio, resulting in more thermodynamically stable ice crystals (Mazur, 1966a). AFP

adsorption to ice crystals at temperatures near the equilibrium freezing point results in

local curvature of the ice face, due to the Kelvin effect. This makes ice growth

energetically unfavourable until a critical temperature is reached. Below that critical

temperature, ice growth is uninhibited (Raymond & DeVries, 1977). Because of the

Kelvin effect, the activation energy for the addition of water molecules to ice is increased, thus AFP adsorption results in ice recrystallization inhibition (IRI), without significantly affecting the melting point. The temperature difference between the freezing and melting points, or thermal hysteresis (TH), can help AFP-producing organisms delay or even avoid freezing.

Certain microorganisms, including bacteria, show modest levels of TH, but demonstrate significant levels of IRI. This indicates that the role of AFPs in microbes may be to decrease cellular damage rather than to prevent freezing altogether (Gilbert et al., 2004, 2005). AFPs have been cloned and characterized from a Flavobacteriaceae

(Raymond et al., 2008), Colwellia (Raymond et al., 2007), Marinomonas primoryensis

(Gilbert et al., 2005; Garnham et al., 2008), Moraxella sp. (Yamashita et al., 2002), and

Pseudomonas putida (Xu et al., 1998). AFPs have also been associated with

Chryseobacterium (Walker et al., 2006), Flavobacterium xanthum (Kawahara et al.,

2001), Micrococcus cryophilus and Rhodococcus erythropolis (Duman & Olsen, 1993).

Additionally IRI activity, possibly due to AFPs, has been associated with a number of genera including Acinetobacter, Bacillus, Buttiauxella, Chryseobacterium,

Enterobacter, Halomonas, Idiomarina, Marinomonas, Pseudoalteromonas,

Pseudomonas, Psychrobacter, Sphingomonas and Stenotrophonomas (Gilbert et al.,

2004; Walker et al., 2006; Wilson et al., 2006). However, IRI can also be mediated by polysaccharides and bacterial gums, which have application in the frozen food industry

(Goff, 1995). Thus some genera listed above likely do not produce AFPs.

Ice Nucleation Proteins:

Biological ice nucleation was first identified by Schnell and Vali (1972), while

Maki et al., (1974) showed that ice nucleation can be mediated by bacteria. Ice- nucleation proteins (INPs), proteins that induce nucleation, are found in a limited number of bacteria, typically Gram negative bacteria, often belonging to the genus Pseudomonas

(Cochet & Widehem, 2000). The presence of INPs may increase survival of their hosts in subzero conditions by nucleating ice externally, possibly resulting in an osmotic gradient and concomitant cellular dehydration and thereby reducing the likelihood of internal ice formation (Zachariassen & Hammel, 1976; Xu et al., 1998). Since membrane damage is correlated with rates of freezing, the production of INPs could allow microbes to control the rate within an optimal range (Lundheim, 2002).

INPs are outer membrane bound proteins that form aggregates, and are thought to provide a template for ice growth by structurally mimicking the ice lattice (Warren et al., 12

1986; Cochet & Widehem, 2000). INPs aggregate in order to promote ice nucleation at high, subzero temperatures (Govindarajan & Lindow, 1988), which then decreases the activation energy needed for the water to ice phase transition to occur (Margaritis &

Bassi, 1991). INPs have been categorized with respect to their activity; type I ice nucleation activity (INA), allows nucleation between -2 and -5°C, while types II and III nucleate between -5 and -7°C or -7 to -10°C, respectively (Yankofsky et al., 1981). The activity of INPs depends on a number of factors, including the size of INP aggregates

(Govindarajan & Lindow, 1988). Glycerol incorporation into the growth media also increases INP activity (Lindow et al., 1982), possibly because glycerol may increase membrane fluidity, thereby allowing for larger protein aggregates. In comparison, non- biological nucleators such as silver iodide are less efficient, nucleating at or below -8°C

(Kozloff et al., 1984). By convention, the lower limit of nucleation activity is -10°C

(Lundheim, 2002).

INPs have been characterized from Erwinia ananas (Abe et al., 1989), Erwinia herbicola (Lindow et al., 1978), Pseudomonas antarctica (Obata et al., 1999; Muryoi et al., 2003), Pseudomonas borealis (Wu et al., 2009), Pseudomonas fluorescens (Corotto et al., 1986), Pseudomonas putida (Xu et al., 1998), Pseudomonas syringae (Green &

Warren, 1985), Pseudomonas viridiflava (Obata et al., 1989; Hasegawa et al., 1990), and

Xanthomonas campestris (Zhao & Orser, 1990). Additionally, INA has been associated with some other genera including: Bacillus, Clavibacter, Corynebacterium,

Curtobacterium, Exiguobacterium, Flavobacterium, Frigoribacterium, Pedobacter,

Sphingobacterium, and Sphingomonas (Nejad et al., 2004; Ponder et al., 2005; Nejad et al., 2006; Wilson et al., 2006).

Freeze-Thaw and Ice-Affinity Selective Methods and Ice Assays:

Equipment for Selection:

Even though certain microbes possess adaptations to resist or tolerate low temperature damage, the presence of ice and its propensity to cause desiccation challenges viability. Freeze-thaw or ice-affinity selection was therefore used as a tool, allowing for the isolation of members of microbial consortia that were resistant to cold and/or osmotic stresses. Freeze-thaw selection used multiple sequential freeze-thaw cycles with a fabricated cryocycler (Fig. 1.1A; Walker et al., 2006). The cryocycler consists of two baths set at different temperatures and containing ~50% ethylene glycol, which was directed to a sample chamber using 3-way valves and controlled by a timer.

Each freeze-thaw cycle was programmed for one hour of freezing (-18°C), and a subsequent hour of thawing (+5°C), for a total of two hours per cycle. In practice, samples (2 mL) were below 0°C for approximately 95 min of each cycle (see Walker et al., 2006 for detailed parameters). Samples were subjected to one or two sets (with recovery via re-culturing between sets) of 48 tandem freeze-thaw cycles.

Ice-affinity selection employs a polycrystalline ice finger formed over a hollow brass finger which was chilled with circulating 50% ethylene glycol (Fig. 1.1B; Kuiper et al., 2003). The temperature was slowly decreased and since impurities are excluded during ice formation, only molecules that interacted with the growing ice should be

Figure 1.1. Schematics of the cryocycler (A; Walker et al., 2006) and the ice-finger (B; Kuiper et al, 2003), used to isolate freeze-thaw resistant microbes (cryocycler) and those with an affinity for ice (ice-finger).

incorporated. Ice-affinity was developed to purify properly folded AFPs from crude cell

lysates (Kuiper et al., 2003), and was later adapted to select for ice-association microbes from 50 mL of dilute culture (see Wilson et al., 2006 for details). For both selection

regimes, community abundance was determined by colony forming units (CFU) per mL,

and diversity using morphology and sequence analysis (see below).

Microbe Ice-Association Activities:

Microbes isolated from various environmental sites following freeze-thaw and/or

ice-affinity selection were assayed for AFP activity (IRI and TH assays) and INP activity

(ice nucleation activity; INA assay). IRI assays assessed the ability of the microbial

isolates to prevent ice crystal growth and were conducted by loading whole cell broth

cultures (~108 CFU/mL) into capillary tubes (10 µL), which were snap frozen (< -25°C) and annealed for at least 16 h at a high subzero temperature (-6°C). Digital images were captured at the beginning and end of incubation through a cross polarizing filter, which allowed for the comparison of pre- and post-annealed ice crystal morphology. Microbes that visually prevented ice crystal growth were deemed to demonstrate IRI and thus, possibly AFP activity (Tomczak et al., 2003). TH assays were performed by placing selected cultures (~108 CFU/mL) onto the grid of a nanolitre osmometer. After snap freezing the temperature was slowly increased until a single visible ice crystal remained.

The temperature was then slowly decreased until the ice crystal began to grow (described by Chakrabartty & Hew, 1991). TH activity was assessed by the difference between the freezing and melting points of individual crystals in milliosmoles, mOsm, and converted

to °C. Ice crystal shaping was assessed by noting an altered (non-circular) crystal 16

morphology as previously described (Chakrabartty and Hew, 1991). It is important to

note that with whole cell cultures, the concentration of AFPs, if present, on the outer

membrane may be insufficient for TH measurements, but sufficient for the assessment of

altered crystal morphology. Thus, with whole cell culture, ice-shaping serves as another

indicator of putative AFP activity. Purified recombinant AFPs, prepared as described in

(Gordienko et al., 2010), were used as positive controls for IRI, TH and ice-shaping assays.

Isolates were also assayed for INA, wherein droplets of whole cell culture were placed on a polarized film, which was then put in a chamber containing circulating ethylene glycol, which was slowly cooled using a programmable bath. The temperature of the sample chamber was continuously monitored and cross-polarized digital images were captured over the course of the assay. The temperature at which 90% of the droplets froze was deemed to be the ice-nucleation temperature. Any sample that increased the temperature at which ice nucleation occurred, relative to the controls, was classified as a nucleator. An INP-positive Pseudomonas syringae preparation (Wards

Natural Science Establishment, Rochester, NY, USA) was used as the positive control.

Plots of ice nuclei vs. temperature were plotted for INA+ isolates of interest (modified

from Vali, 1971; Maki et al., 1974; Lindow et al., 1982).

Collection Sites and Culture Conditions:

Collections were made at a number of North American sites, in the late summer, including soil samples from temperate and low arctic sites, temperate lake sediments, and water samples (see Table 1.1 for sample description and location). Collections were 17

Table 1.1. Summary of sample collection sites.

Collection Sample and Name: GPS Location: Comments: A. Soil Collections: 1. Bank, YT, Canada 63º26'N, 138º48'W Boreal Forest 2. Calgary, AB, Canada 51º17'N, 115º5'W Temperate 3. Camp, YT, Canada 63º23'N, 139º03'W Boreal Forest 4. Daring Lake, NT, Canada 64º52'N, 111º35'W Exposed Esker, Low Arctic 5. Fort Nelson, BC, Canada 58º49'N, 123º30'W Boreal Forest 6. Gould Lake, ON, Canada 44º26'N, 76º35'W Temperate

B. Lake Sediment: 1. Fly Lake, BC, Canada 51°54'N, 121°19'W Temperate, Freshwater 2. Frozen Creek, YT, Canada 63º31'N, 138º45'W Boreal Forest, Freshwater 3. Lake East of 83, BC, Canada 51°28'N, 121°23'W Temperate, Brine 4. Leeches Lake, BC, Canada 52°2'N, 122° 19'W Temperate, Freshwater 5. Liberty Lake, BC, Canada 51°17'N, 121°43'W Temperate, Brine 6. Pyramid Lake, NV, U.S.A. 39º59'N, 119º37'W Temperate, Low Salinity 7. Ranch Lake, BC, Canada 51°34'N, 121°39'W Temperate, Alkali 8. Spotted Lake, BC, Canada 49°5'N, 119°34'W Temperate, Alkali

C. Lake Water: 1. Bank, YT, Canada 63º26'N, 138º48'W Boreal Forest, Freshwater 2. Fly Lake, BC, Canada 51°54'N, 121°19'W Temperate, Freshwater 3. Frozen Creek, YT, Canada 63º31'N, 138º45'W Boreal Forest, Freshwater 4. Lake East of 83, BC, Canada 51°28'N, 121°23'W Temperate, Brine 5. Leeches Lake, BC, Canada 52°2'N, 122°19'W Temperate, Freshwater 6. Liberty Lake, BC, Canada 51°17'N, 121°43'W Temperate, Brine 7. Pyramid Lake, NV, U.S.A. 39º59'N, 119º37'W Temperate, Low Salinity 8. Spotted Lake, BC, Canada 49°5'N, 119°34'W Temperate, Alkali

made as aseptic as logistically possible. In all cases, soils were collected in plastic from

the top 2-3 cm and from at least three adjacent locations and composited. Lake samples

were collected in plastic, as per Whitman et al., (2006). Briefly, sediments were from the top 2-3 cm below the water column and water samples were obtained away from the shore. Sites were designated according to map names, watershed catchments or local features. Daring Lake soil samples were stored at -20°C, while all others were stored at

4°C until used. Sample site locations can be found in Table 1.1.

Soil samples were enriched in modified 10% tryptic soy broth (TSB; 3 g tryptic

soy broth (Bacto, Dickinson and Company, Sparks, MD, USA), 0.1 g KNO3, 0.1 g

(NH4)2SO4 and 0.1 g K2HPO4 per liter of deionized water), for ~48 h at ambient

temperature. These enrichment cultures were incubated at 4°C overnight or for ~48 h

prior to selection or assay, respectively. Following one set of 48 freeze-thaw cycles, the

consortia were recovered in fresh media and cultured as above. The recovered cultures

were routinely treated with a second set of 48 freeze-thaw cycles with the aforementioned

cryocycler. Surviving microbes were isolated as monocultures. The goal of this

selection regime was to isolate the truly resistant microbes, thus the conditions are

admittedly rather extreme. Enrichment cultures were cold-acclimated to 4°C prior to

selection to allow for the induction of cold adaptive proteins, however, no cryoprotectants

were added to the media. While no optimization was done with respect to cycling or

cold-acclimation conditions, experiments with alternative cycling conditions are currently

under way.

Freshwater lake samples were cultured in 10% TSB. Pyramid Lake (major ions

+ + 2+ 2+ - 2- 2- 2- include Na , K , Mg , Ca , Cl , HCO3 , CO3 , SO2 ; www.ilec.or.jp/database

/nam/nam-23.html) samples were also enriched in 10% TSB, supplemented with minimal

salts. Lake East of 83 (major ion is Na+) and Liberty Lake (major ions include Na+ and

Cl-) were cultured in media designed to reflect the ion composition of the respective lake

water (as determined by the Analytical Services Unit; Queen’s University, Kingston, ON,

2+ 2- Canada). Ranch Lake (major ions include Mg and SO4 ) and Spotted Lake (major ions

+ + 2+ - 2- 2- include Na , K , Mg , Cl , SO4 and S ) were cultured in 50% Marine Broth (Difco,

Becton, Dickinson and Company, Sparks, MD, USA). As with the soil-derived consortia, lake enrichments were cultured at 22°C for ~48 h, with subsequent incubation at 4°C

overnight or for ~48 h prior to selection or assay, respectively. Resistant lake-derived

microbes were isolated as monocultures in their respective media.

Isolate Identification:

Soil and lake isolates were putatively identified on the basis of 16S rRNA gene sequence analysis, using universal bacterial primers (8F and r1406; Lane et al., 1985;

Hicks et al., 1992; Sigma-Genosys, Oakville, ON, Canada) and the amplification

conditions of Telang et al., (1997). The ~1.4 kbp product was purified (Qiagen,

Mississauga, ON, Canada), and sequenced on both strands (University Core DNA

Services, University of Calgary, AB, Canada or Plateforme de Génomique at the Centre

de recherché du CHUL, Quebec, QC, Canada). The putative identity of each isolate was

assigned based on the nearest phylogenetic relative in the BLASTn

(http://www.ncbi.nlm.nih.gov/blast; Altschul et al., 1997) or Ribosomal Database Project

II (http://rdp.cme.msu.edu; Cole et al., 2005, 2009) databases.

Experimental Selection and Resistance:

Soil Communities and Isolates:

Ice-affinity and/or freeze-thaw selection (see above) were used to isolate resistant microbes from soil consortia derived from a number of locations (Table 1.1). Initially, soil samples were taken from the boreal region of Canada (between 58º and 63º N). After cryocycler-mediated freeze-thaw selection, community diversity of the samples decreased as determined by visible colony morphology differences. Cultivable cell viability also decreased, from an initial culture density normalized to 108 CFU/mL, by

~3, 4 and 4 orders of magnitude, respectively, in the Bank, Camp and Fort Nelson

samples after 48 freeze-thaw cycles (Table 1.2). When survivors were used to initiate

new cultures, viability was again monitored after a subsequent 48 cycles, with the

profiles showing similar degrees of resistance. Comparably, the positive and negative

controls; Chryseobacterium sp. C14 (Walker et al., 2006) and E. coli TG-2, showed near

complete resistance and susceptibility, respectively (Table 1.2). Since all of these soils

came from a similar geographic region, such minor variance in freeze-thaw resistance is

unlikely to be significant. Subsequently, soils were collected from a low arctic site

located further north (Daring Lake, 64ºN) and from further south (Calgary, 51ºN and

Gould Lake, 44ºN). Again, there was a marked decrease in cell viability after freeze-

thaw selection, irrespective of geographical origin. Following two sets of 48 freeze-thaw

Table 1.2. Microbial abundance following freeze-thaw selection.

Soil Samples: Abundance1 Following Abundance Following First 48 Cycles* Second 48 Cycles* Bank (YT, Canada) 1.51 x 105 + 3.72 x 104 1.63 x 104 + 4.18 x 103 Calgary2 (AB, Canada) 3.31 x 103 + 1.31 x 103 5.73 x 106 + 6.52 x 106 Camp (YT, Canada) 3.15 x 104 + 1.19 x 104 2.60 x 106 + 3.29 x 106 Daring Lake3 (NT, Canada) 8.99 x 103 + 1.09 x 104 4.65 x 104 + 3.18 x 104 Fort Nelson (YT, Canada) 3.46 x 104 + 9.32 x 103 9.25 x 104 + 3.20 x 104 Gould Lake3 (ON, Canada) 7.80 x 104 + 1.25 x 104 3.14 x 105 + 4.84 x 105

Lake Samples4: Abundance: Water Abundance: Sediment Freshwater Lakes: Bank Creek (YT, Canada) 5.44 x 100 + 1.07 x 102 NA Fly Lake (BC, Canada) 2.34 x 103 + 2.09 x 103 9.76 x 101 + 2.95 x 101 Frozen Creek (YT, Canada) 0 NA Leeches Lake (BC, Canada) 3.08 x 103 + 1.77 x 103 7.87 x 102 + 2.74 x 101

Brine Lakes: Lake East of 83 (BC, 5.45 x 104 + 1.95 x 104 9.73 x 103+ 1.99 x 103 Canada) Liberty Lake (BC, Canada) 5.17 x 106 + 6.99 x 106 2.70 x 104 + 2.01 x 104 Pyramid Lake (NV, U.S.A.) 2.18 x 101 + 1.31 x 101 2.09 x 104 + 3.00 x 104

Alkali Lakes: Ranch Lake (BC, Canada) NA 1.89 x 106 + 2.44 x 106 Spotted Lake (BC, Canada) 4.69 x 106 + 9.23 x 105 4.34 x 106 + 5.30 x 106

Controls: Abundance: Chryseobacterium sp. C145 2.55 x 107 + 8.48 x 106 E. coli TG-26 0 1 All data normalized to 1x108 CFU/mL as the initial (pre-selection) enrichment culture abundance, N~3. 2 Walker et al., 2006. 3 Chapter 3 4 Following 1 set of freeze-thaw cycles. 5 Positive control; isolated following freeze-thaw selection, Walker et al., 2006. 6 Negative control; no known ice-association (Walker et al., 2006; Wilson et al., 2006). * One-way analysis of variance (ANOVA) tests indicated significant differences amongst the soil and lake samples (p < 0.05). Subsequent Tukey-Kramer HSD tests were used to assess pair wise differences amongst the soil and lake samples (p < 0.05); (http://udel.edu/~mcdonald/anova.xls) addressed in text.

cycles, the Calgary consortium was the most resistant, more so than following the first set

of freeze-thaw cycles, or any other soil sample (p< 0.05; Table 1.2), losing approximately

one logarithm of viability. The other consortia lost approximately three to four

logarithms of viability, showing no increased resistance relative to the first set of freeze-

thaw cycles, or the other soil samples (p< 0.05; Table 1.2). It is possible that natural

freeze-thaw conditions in the Calgary region could have pre-selected for a greater recovery of freeze-thaw resistance after the second 48 cycles (Walker et al., 2006).

Overall, soil-derived consortia from these geographically distant locations, spanning more than 2800 km, displayed rather similar resistance profiles in response to the same selective regime.

After selection, consortia diversity decreased, either assessed by phenotype or

quantified using DNA sequence analysis, such that at the conclusion of the experiment

with each of the soil communities, only a limited number of genera were isolated (Table

1.3). Limited rRNA gene sequencing was conducted on the boreal microbial samples. A

subset of the isolates from the Camp consortia yielded Rhanella, while Enterococcus and

Exiguobacterium were identified from the Fort Nelson consortia. Members of the Bank

consortia survived, but sequencing and subsequent putative identification, have not been

done. Freeze-thaw selection of enrichment cultures from the geographically distant soils,

including Calgary, Gould and Daring Lake soils, resulted in a decreased diversity and

abundance. Figure 1.2 represents phylum-level groupings of richness prior to and

following freeze-thaw selection of the Daring- and Gould Lake-derived enrichments. As

Table 1.3. Soil-derived enrichment culture richness following selection.

Sample: Genera Present Following Selection: Calgary Acinetobacter1, Bacillus2, Buttiauxella1, Carnobacterium1, Chryseobacterium1, Enterococcus1, Paenibacillus2 Camp Rhanella1 Daring Lake Acinetobacter2, Bacillus2, Chryseobacterium1, Paenibacillus1, Pseudomonas1,2, Sphingomonas2, Streptomyces2 Fort Nelson Enterococcus1, Exiguobacterium1 Gould Lake Acinetobacter2, Arthrobacter2, Bacillus1, Buttiauxella1, Chryseobacterium2, Flavobacterium2, Microbacterium2, Paenibacillus2, Pantoea (Erwinia)1, Pseudomonas2, Stenotrophomonas2 1 Isolated with freeze-thaw selection (Daring and Gould Lake, Chapter 3; Calgary, Walker et al., 2006). 2 Isolated with ice-affinity (Chapter 2; Wilson et al., 2006). Note, ‘richness’ is used in this table, and throughout this thesis to refer to the organisms that are present (as opposed to diversity, which accounts for the organisms present, as well as their relative proportions).

Figure 1.2. The richness of the consortia derived from Daring Lake (A) before (left) and after (right) as well as Gould Lake (B) before (left) and after (right) freeze-thaw selection, shown as pie graphs. Proportions represent groupings of unique genera at the phylum level, as determined by 16S rDNA clone libraries (before) or 16S rDNA sequencing of the recovered isolates (after). This decrease in richness corresponds to approximately a 10-fold decrease in diversity and a 103-104 fold decrease in bacterial abundance.

indicated, the number of phyla represented in each decreased, thus reflecting decreased microbial richness following selection. Ultimately, the decrease in abundance (Table 1.2) and richness (Fig. 1.2) resulted in the isolation of 6 different genera from Gould and/or

Daring Lake (Fig. 1.2, Table 1.3; Chapter 3), and 5 different genera from Calgary soil- derived samples. Two of the Calgary isolates were also identified from the Gould and/or

Daring Lake consortia (Table 1.3; Walker et al., 2006). It is noteworthy that many of these genera have previously been isolated from low temperature environments, including the Arctic and Antarctic (eg. Nelson & Parkinson, 1978; Christner et al., 2001;

Gilbert et al., 2004).

Since the cryocycler was fabricated to select for microbes that might be rare in the original consortia, it was important to try to determine the relative abundance of the survivors in the unselected consortia. Semi-quantitative, real time PCR (QPCR) with universal and genus-specific primers was used with the Daring and Gould Lake enrichment cultures, to assess the relative abundance of the genera to which a subset of the resistant isolates belong. Figure 1.3 represents the abundance of the assessed genera in the original enrichment cultures. Our particular isolated strains would represent a small proportion of the genera (Chapter 3). The recovered survivors, following freeze- thaw selection, were indeed rare in the respective original cultured consortium. Thus, freeze-thaw cycles select for those rare bacteria with a phenotype that confers a high level of resistance, which could be mediated by intra- or extra-cellular properties. In contrast, the second selective regime, ice-affinity, should enrich for those isolates with extracellular molecules, allowing microbes to adsorb to ice.

Figure 1.3. The percentage of the initial Daring and Gould Lake enrichment cultures represented by a subset of resistant genera (Table 1.3), based on semi-quantitative, real time PCR (QPCR). The abundance of the resistant strains would be considerably less than that represented by the genera to which the strains belong.

Ice-affinity selection of enrichment cultures derived from the Calgary, Daring and

Gould Lake soil samples resulted in the isolation of 11 different genera. Strikingly, 5 of these (Acinetobacter, Bacillus, Chryseobacterium, Paenibacillus, and Pseudomonas) were also isolated from these soils by freeze-thaw selection (Table 1.3). Again, the selective regime resulted in a decrease in consortia abundance and diversity, the latter indicated by a decrease in distinct colony morphologies, as well as reflected in the limited number of sequence identities. Presumably, ice-affinity selection also results in the recovery of rare microbes. Semi-quantitative QPCR studies, however, have not yet been undertaken with these samples.

Ice-association assays were used to determine if any of the survivors from either regime had ice-shaping or IRI characteristics and/or INA (Fig. 1.4). Of the genera recovered and subsequently assayed, Acinetobacter (Daring Lake; ice-affinity),

Buttiauxella (Gould Lake; freeze-thaw), and a different Chryseobacterium from each

Calgary (freeze-thaw), Gould Lake (ice-affinity) and Daring Lake (freeze-thaw) demonstrated ice-shaping or IRI activity. A Pseudomonas (Daring Lake, freeze-thaw and ice-affinity), Flavobacterium (Gould Lake; ice-affinity) and a Bacillus (Gould Lake; freeze-thaw) demonstrated both ice-shaping or IRI, as well as type I INA activities

(Walker et al., 2006; Wilson et al., 2006), and Exiguobacterium (Fort Nelson; freeze- thaw) showed type I INA. Remarkably, more than 60% of the assayed isolates from Fort

Nelson, Calgary, Daring and Gould Lakes demonstrated one or more ice-association activities (Wilson et al., 2006). Comparably, when randomly chosen isolates from the

Figure 1.4. Representative ice-association assays: A. Ice recrystallization inhibition assays before (A I) and after (A II) 16 h annealing at -6°C. The five samples, from left to right (in duplicate capillaries, capillaries are 1mm in diameter) are; Paenibacillus and Pseudomonas (both from Daring Lake; demonstrating ice recrystallization and ice recrystallization inhibition, respectively), E. coli and 10% TSB (both are negative controls) and Type 1 AFPs (positive control). B. Ice-shaping assay with E. coli TG-2 (B I; negative control) and Pseudomonas borealis from Daring Lake (B II; showing ice shaping). BI and BII are ~76 and 120 µm, respectively, along the widest dimension; Figure modified from Wilson et al., (2006). C. Ice nucleation assay, samples include, from top to bottom (with 10, 1µL droplets per row); Bacillus (from Gould Lake; positive for INA), E. coli TG-2 and 10% TSB (both are negative controls) and Ward’s INPs (positive control). Pictures were captured at 2.7°C (C I), -2.4°C (C II), and -11.3°C (C III).

original, unselected, Daring and Gould Lake soil consortia were assayed for these three

ice-association characteristics, none showed such activity.

Lake Communities and Isolates:

Enriched consortia derived from 9 different lakes (see Table 1.1 for locations)

were also subjected to freeze-thaw selection in the cryocycler. Lake water and sediment

samples from two freshwater sources in the boreal region (63ºN), Bank and Frozen

Creeks, as well as two from temperate regions (51-52ºN), Fly and Leeches Lakes were

obtained. Three temperate climate brine lakes were sampled: Lake East of 83 and

Liberty Lake (51ºN), as well as Pyramid Lake (39ºN). Finally, two temperate climate

alkali lakes were included in the sample collection: Ranch Lake and Spotted Lake (49-

51ºN).

Enrichment cultures derived from the boreal freshwater Bank and Frozen Creeks

showed little resistance to freeze-thaw selection, retaining little (100) or no viability,

respectively, after 48 cycles. As a result, no sequencing data are available for these

samples. Preliminary results indicated that cultures derived from the more temperate

freshwater samples (Fly and Leeches Lakes) also showed little resistance, retaining

approximately 3 or 2 logarithms of viability for the water and sediment-derived samples,

respectively. Interestingly, the more southerly, slightly saline Pyramid Lake-derived consortia were similarly resistant, retaining approximately 101 and 104 CFU/mL for the

water and sediment-derived samples, respectively (Table 1.2). Overall, in comparison to

the soil-derived consortia, these water samples were more freeze-thaw susceptible.

Since freeze stress shares characteristics with osmotic challenge, we considered

that the low-salinity Pyramid Lake consortia might have preadapted to freeze stress, even

though this southerly lake has never been observed to freeze

(www.ilec.or.jp/database/nam/nam-23.html). This hypothesis was tested by collecting

samples from two temperate climate brine lakes: Lake East of 83 and Liberty Lake

(51°N). They showed similar resistance, with viability decreasing by ~2 and 4

logarithms for water-derived and ~4 logarithms for the sediment-derived samples, with

the Liberty Lake water-derived consortium being significantly more resistant then the

previously mentioned samples (p< 0.05). Two temperate climate alkali lakes: Ranch and

Spotted Lakes (49-51°N), showed the highest freeze-thaw resistance, losing only ~2

logarithms of viability for both water and sediment samples. The Spotted Lake water-

derived consortium was more resistant than all other samples, while the Spotted Lake

sediment-derived consortium was more resistant than all but the Liberty Lake water-

derived and Ranch Lake sediment-derived consortia (p<0.05). Overall, the brine and

alkali lake-derived consortia tended to be more resistant than their freshwater

counterparts. Thus, although these experiments were conducted on a limited number of

lakes, it appears that resistance to freeze-thaw stress may increase with increasing lake salinity.

As with the resistant soil microbes, the isolates derived from the lake enrichments

were phylogenetically diverse, but again, demonstrated considerable overlap amongst

lake types (Table 1.4). Resistant microbes include a number of spore formers, as well as

those associated with psychrophily and/or halophily. A subset of the isolates were

Table 1.4. Lake sediment and water derived enrichment culture richness following freeze-thaw selection.

Lake: Richness Following Selection Freshwater: Fly Lake Acinetobacter, Arthrobacter, Bacillus, Brevibacillus, Cryobacterium, Herbaspirillium, Leifsonia, Lysinibacillus, Paenibacillus, Pseudomonas, Rhodococcus, Sporocarcina, Trichococcus Leeches Lake Arthrobacter, Bacillus, Marinibacillus, Paenibacillus, Pseudomonas, Sphingomonas, Trichococcus

Brine: Lake East of 83 Bacillus, Halomonas, Nesterenkonia, Pseudomonas Liberty Lake Alkalibacterium, Bacillus, Halomonas, Idiomarina, Nesterenkonia, Roseinatronobacter

Alkali: Ranch Lake Bacillus, Gillisia, Marinobacter, Pseudidiomarina, Pseudomonas Spotted Lake Bacillus, Halomonas, Idiomarina, Lysinibacillus, Nesterenkonia, Salegentibacter

assayed for AFP and INP activity, and preliminary results indicate, in striking contrast to

the results obtained from the soil-derived isolates, that none of these isolates demonstrated AFP and/or INP activity. However, many of the isolated resistant bacteria

are known to be associated with stress resistance, and some with the production of

osmoprotectants and/or cryoprotectants. This, coupled with the trends in freeze-thaw

resistance depending on lake type, supports our contention that the microbes were able to resist freeze-thaw stress due to previous adaptations to osmotic stress. Since freeze-thaw

cycling did not appear to result in the recovery of any bacteria with AFP or INP activity

in these samples, it suggests that these ice-association activities either do not function in

saline solutions or that microbes with AFPs and INPs do not flourish in brine or alkali

lakes.

Biotechnological Implications:

Antifreeze proteins can be used in the food and medical industries to improve

frozen food texture or tissue and gamete preservation, respectively. It is thought that the

uncontrolled growth of large ice crystals or spicules are potentially damaging to

biological membranes, while small ice crystals and slowly growing, rounded ice crystals, in comparison, are relatively benign (Mazur, 1966a). High levels of Type I fish AFPs, when added to suspended prostatic adenocarcinoma cells during freezing, promoted ice spicule formation at temperatures below the equilibrium freezing point and therefore the death of the carcinoma cells. While this is a function of AFP type and concentration,

33 certain AFPs have the potential to increase the efficacy of cryosurgery (Koushafar &

Rubinsky, 1997). Under conditions of snap freezing, however, as previously mentioned,

IRI lessens the probability of membrane damage and helps maintain the integrity of frozen cells. Therefore the addition of AFPs to frozen foods can increase the shelf-life and quality of these products, which may otherwise be decreased by ice recrystallization

(Feeney & Yeh, 1998; Zhang et al., 2007). Similarly, the use of AFPs for tissue cryopreservation is under study (Arav et al., 1993; Bagis et al., 2008). Lastly, AFPs are being studied as potential ‘green’ gas hydrate growth inhibitors (Zeng et al., 2006;

Gordienko et al., 2010).

INPs have found applications in cloud seeding, recreation, food and energy industries. Microbial INPs have replaced inorganic nucleators, such as silver iodide, for some applications since they are considered more environmentally safe and efficient, mediating nucleation at warmer temperatures (eg. see the separate works of Ward &

DeMott ,1989, as cited in Cochet & Widehem, 2000). The amount of energy required for the initial freezing phase of some refrigeration and air conditioning systems can be reduced by the use of INPs to increase crystallization temperature (Stewert & Bear, 1998;

Tsuchiya et al. 2004). Applications also include artificial snow production (SnowmaxTM;

Woerpel, 1980; Cochet & Widehem, 2000). Lastly, the preparation of many beverages, including juices containing volatile flavours or aromas, involves a freeze-concentration step. The ice crystals formed in the presence of INPs are larger thus making the process more efficient (summarized in Cochet & Widehem, 2000).

Proteins resistant to osmotic and cold stresses are also of value since preserved

tissues and cells must contend with osmotic stress (Wang & Ben, 2004). Sakamoto et al.

(1998) produced cold and osmotic stress resistant transgenic rice plants by the transfer of

a bacterial gene responsible for the production of glycine betaine. Proteins that confer

increased osmotolerance are also of use in the food industry (eg. Hernandez-Lopez et al.,

2003).

In conclusion, we have shown that freeze-thaw and ice-affinity selection can be used to isolate bacteria from geographically-distinct soils and lakes, which demonstrate resistance to freeze stress and have a high frequency of ice-association protein activities and molecules important for osmotic stress. We anticipate the continued analysis of these isolates will results in products for diverse biotechnological applications.

Acknowledgements:

This work was financially supported by NSERC (Canada), an International Polar Year

(CiCAT) grant and a Queen’s Research Chair award to VKW and NSERC and OGS

(Canada) scholarships to SLW. Drs. B. Cumming, P. Grogan and G. Voordouw are thanked for the initial lake sediment samples, the Daring Lake soil sample, and the E. coli

TG-2, respectively. Dr. G. Palmer is thanked for his technical assistance. V. Nesbitt and

T. Vanderveer are acknowledged for assistance with the data collection, and A.

Kanawaty, S. Franchuk, B. Momciu, R. Murray and A. Stanczak for some preliminary analysis.

Literature Cited:

Abe, K., Watabe, S., Emori, Y., Watanabe, M., and Arai, S. (1989). An ice nucleation active gene of Erwinia ananas. Sequence similarity to those of Pseudomonas species and regions required for ice nucleation activity. FEBS Letters 258: 297-300.

Allakhverdiev, S.I., Nishiyama, Y., Suzuki, I., Tasaka, Y., and Murata, N. (1999). Genetic engineering of the unsaturation of fatty acids in membrane lipids alters tolerance of Synechocystis to salt stress. Proc. Natl. Acad. Sci. USA 96: 5862-5867.

Altschul, S.F., Madden, T.L., Schaffer, A.A., Zhang, J., Zhang, Z., Miller, W., and Lipman, D.J. (1997). Gapped BLAST and PSI-BLAST: a new generation of protein database search programs. Nucleic Acids Res. 25: 3389-3402.

Alur, M.D., and Grecz, N. (1975). Mechanism of injury of Escherichia coli by freezing and thawing. Biochem. Biophys. Res. Commun. 62: 308-312.

Amato, P., Doyle, S., and Christner, B.C. (2009). Macromolecular synthesis by yeasts under frozen conditions. Environ. Microb. 11: 589-596.

Arav, A., Rubinsky, B., Fletcher, G., and Seren, E. (1993). Cryogenic protection of oocytes with antifreeze proteins. Mol. Reprod. Dev. 36: 488-493.

Bagis, H., Akkoc, T., Tas, A., and Aktoprakligil, D. (2008). Cryogenic effect of antifreeze protein on transgenic mouse ovaries and the production of live offspring by orthotopic transplantation of cryopreserved mouse ovaries. Mol. Reprod. Dev. 75: 608- 613.

Barrett, J. (2001). Thermal hysteresis proteins. Int. J. Biochem. Cell Biol. 33: 105-117.

Beaufils, S., Sauvageot, N., Mazé, A., Laplace, J.-M., Auffray, Y., Deutscher, J., and Hartke, A. (2007). The cold shock response of Lactobacillus casei: relation between HPr phosphorylation and resistance to freeze/thaw cycles. J. Mol. Microb. Biotechnol. 13: 65- 75.

Bennett, G.M., Seaver, A., and Calcott, P.H. (1981). Effect of defined lipopolysaccharide core defects on resistance of Salmonella typhimurium to freezing and thawing and other stresses. Appl. Environ. Microb. 42: 843-849.

Calcott, P.H., and MacLeod, R.A. (1974a). Survival of Escherichia coli from freeze-thaw damage: A theoretical and practical study. Can. J. Microb. 20: 671-681.

Calcott, P.H., and MacLeod, R.A. (1974b). Survival of Escherichia coli from freeze-thaw damage: Influence of nutritional status and growth rate. Can. J. Microb. 20: 683-689.

Calcott, P.H., and MacLeod, R.A. (1975a). The survival of Escherichia coli from freeze- thaw damage: permeability barrier damage and viability. Can. J. Microb. 21: 1724-1732.

Calcott, P.H., and MacLeod, R.A. (1975b). The survival of Escherichia coli from freeze- thaw damage: the relative importance of wall and membrane damage. Can. J. Microb. 21: 1960-1968.

Chakrabartty, A., and Hew, C.L. (1991). The effect of enhanced α-helicity on the activity of a winter flounder antifreeze polypeptide. Eur. J. Biochem. 202: 1057-1063.

Chattopadhyay, M.K., and Jagannadham, M.V. (2001). Maintenance of membrane fluidity in Antarctic bacteria. Polar Biol. 24: 386-388.

Cheng, C.C., and DeVries, A.L. (1991). The role of antifreeze glycopeptides and peptides in the freezing avoidance of cold water fishes. In Life Under Extreme Conditions. di Prisco, G., ed., Springer Verlag, Berlin, pp. 1–14.

Christner, B.C., Mosley-Thompson, E., Thompson, L.G., and Reeve, J.N. (2001). Isolation of bacteria and 16S rDNAs from Lake Vostok accretion ice. Environ. Microb. 3: 570-577.

Cochet, N., and Widehem, P. (2000). Ice crystallization by Pseudomonas syringae. Appl. Microbiol. Biotechnol. 54: 153-161.

Cole, J.R., Chai, B., Farris, R.J., Wang, Q., Kulam, S.A., McGarrell, D.M., Garrity, G.M., and Tiedje, J.M. (2005). The ribosomal database project (RDP-II): sequences and tools for high-throughput rRNA analysis. Nucleic Acids Res. 33: D294-D296.

Corotto, L.V., Wolber, P.K., and Warren, G.J. (1986). Ice nucleation activity of Pseudomonas fluorescens: Mutagenesis, complementation analysis and identification of a gene product. EMBO J. 5: 231-236.

Deal, P.H. (1970). Freeze-thaw behavior of a moderately halophilic bacterium as a function of salt concentration. Cryobiology 7: 107-112.

DeVries, A.L., and Lin, Y. (1977). The role of glycopeptide antifreezes in the survival of Antarctic fishes. In Adaptations within Antarctic Ecosystems. Llano, G. A., ed., Gulf Publishing Co., Houston, pp. 439-458.

Duman, J.G., and Olsen, T.M. (1993). Thermal hysteresis protein activity in bacteria, fungi, and phylogenetically diverse plants. Cryobiology 30: 322-328.

Feeney, R.E., and Yeh, Y. (1998). Antifreeze proteins: current status and possible food uses. Trends Food Sci. Technol. 9: 102-106.

Gao, W., Smith, D.W., and Li, Y. (2006). Natural freezing as a wastewater treatment method: E. coli inactivation capacity. Water Res. 40: 2321-2326.

Garnham, C.P., Gilbert, J.A., Hartman, C.P., Campbell, R.L., Laybourn-Parry, J., and Davies, P.L. (2008). A Ca2+-dependent bacterial antifreeze protein domain has a novel ß- helical ice-binding fold. Biochem. J. 411: 171-180.

Gilbert, J.A., Davies, P.L., Laybourn-Parry, J. (2005). A hyperactive, Ca2+-dependent antifreeze protein in an Antarctic bacterium. FEMS Microb. Letters 245: 67-72.

Gilbert, J.A., Hill, P.J., Dodd, C.E.R., and Laybourn-Parry, J. (2004). Demonstration of antifreeze protein activity in Antarctic lake bacteria. Microbiology 150: 171-180.

Goff, H.D. (1995). The use of thermal analysis in the development of a better understanding of frozen food stability. Pure Appl. Chem. 67: 1801-1808.

Gomez Zavaglia, A., Disalvo, E.A., and De Antoni, G.L. (2000). Fatty acid composition and freeze-thaw resistance in lactobacilli. J. Dairy Res. 67: 241-247.

Gordienko, R., Ohno, H., Singh, V.K., Jia, Z., Ripmeester, J.A., and Walker, V.K. (2010). Towards a green hydrate inhibitor: imaging antifreeze proteins on clathrates. PLoS ONE 5: e8953.

Govindarajan, A.G., and Lindow, S.E. (1988). Size of bacterial ice-nucleation sites measure in situ by radiation inactivation analysis. Proc. Natl. Acad. Sci. USA 85: 1334- 1338.

Green, R.L., and Warren, G.J. (1985). Physical and functional repetition in a bacterial ice nucleation gene. Nature 317: 645-648.

Hasegawa, Y., Sakai, N., Yoshitome, H., Kawate, S., Obata, H., and Tokuyama, T. (1990). Cloning of bacterial ice nucleation genes from Pseudomonas viridiflava in Escherichia coli. J. Ferment. Bioeng. 70: 143-146.

Hernandez-Lopez, M.J., Prieto, J.A., and Randez-Gil, F. (2003). Osmotolerance and leavening ability in sweet and frozen sweet dough. Comparative analysis between Torulaspora delbrueckii and Saccharomyces cerevisiae baker’s yeast strains. A. van Lee. 84: 125-134.

Hicks, R.E., Amann, R.I., and Stahl, D.A. (1992). Dual staining of natural bacterioplankton with 4',6-diamidino-2-phenylindole and fluorescent oligonucleotide probes targeting kingdom-level 16S rRNA sequences. Appl. Environ. Microb. 58: 2158- 2163.

Hubalek, Z. (2003). Protectants used in the cryopreservation of microorganisms. Cryobiology 46: 205-229.

Izawa, S., Sato, M., Yokoigawa, K., and Inoue, Y. (2004). Intracellular glycerol influences resistance to freeze stress in Saccharomyces cerevisiae: analysis of a quadruple mutant in glycerol dehydrogenase genes and glycerol-enriched cells. Appl. Microbiol. Biotech. 66: 108-114.

Jacobs, M.B., Gerstein, M.J., and Walter, W.G. (1957). Dictionary of Microbiology, Van Nostrand, New York.

Jia, Z., DeLuca, C.I., Chao, H., Davies, P.L. (1996). Structural basis for the binding of a globular antifreeze protein to ice. Nature 384: 285-288.

Jones, P.G., VanBogelen, R.A., and Neidhardt, F.C. (1987). Induction of proteins in response to low temperature in Escherichia coli. J. Bacteriol. 169: 2092-2095.

Junge, K., Eicken, H., and Deming, J.W. (2004). Bacterial activity at -2 to -20°C in Arctic wintertime sea ice. Appl. Environ. Microb. 70: 550-557.

Kawahara, H., Iwanaka, Y., Higa, S., Muryoi, N., Sato, M., Honda, M., Omura, H., and Obata, H. (2007). A novel, intracellular antifreeze protein in an antarctic bacterium, Flavobacterium xanthum. Cryo Letters 28: 39-49.

Kawahara, H., Li, J., Griffith, M. and Glick, B.R. (2001). Relationship between antifreeze protein and freezing resistance in Pseudomonas putida GR12-2. Current Microbiol. 43: 365-370.

Klein, W., Weber, M.H.W., and Marahiel, M.A. (1999). Cold shock response of Bacillus subtilis: Isoleucine-dependent switch in the fatty acid branching pattern for membrane adaptation to low temperatures. J. Bacteriol. 181: 5341-5349.

Knight, C.A., Cheng, C.C., and DeVries, A.L. (1991). Adsorption of α-helical antifreeze peptides on specific ice crystal surface planes. Biophys. J. 59: 409-418.

Ko, R., Smith, L.T., and Smith, G.M. (1994). Glycine betaine confers enhanced osmotolerance and cryotolerance on Listeria monocytogenes. J. Bacteriol. 176: 426-431.

Koushafar, H., and Rubinsky, B. (1997). Effect of antifreeze proteins on frozen primary prostatic adenocarcinoma cells. Urology 49: 421-425.

Kozloff, L.M., Lute, M., and Westaway, D. (1984). Phosphatidylinositol as a component of the ice nucleating site of Pseudomonas syringae and Erwinia herbicola. Science 226: 845-846.

Kuiper, M.J., Lankin, C., Gauthier, S.Y., Walker, V.K., and Davies, P.L. (2003). Purification of antifreeze proteins by adsorption to ice. Biochem. Biophys. Res. Commun. 300: 645-648.

La Teana, A., Brandi, A., Falconi, M., Spurio, R., Pon, C.L., and Gualerzi, C.O. (1991). Identification of a cold shock transcriptional enhancer of the Escherichia coli gene encoding nucleoid protein H-NS. Proc. Natl. Acad. Sci. USA 88: 10907-10911.

Lane, D.J., Pace, B., Olsen, G.J., Stahl, D.A., Sogin, M.L., and Pace, N.R. (1985). Rapid determination of 16S ribosomal RNA sequences from phylogenetic analyses. Proc. Natl. Acad. Sci. USA 82: 6955-6959.

Li, S., Xiao, X., Li, J., Luo, J., and Wang, F. (2006). Identification of genes regulated by changing salinity in the deep-sea bacterium Shewanella sp. WP3 using RNA arbitrarily primed PCR. Extremophiles 10: 97-104.

Lindow, S.E., Arny, D.C., and Upper, C.D. (1978). Erwinia herbicola: a bacterial ice nucleus active in increasing frost injury to corn. Phytopathology 68: 523-527.

Lindow, S.E., Hirano, S.S., Barchet, W.R., Arny, D.C., and Upper, C.D. (1982). Relationship between ice nucleation frequency of bacteria and frost injury. Plant Physio. 70: 1090-1093.

Lundheim, R. (2002) . Physiological and ecological significance of biological ice nucleators. Philos. Trans. R. Soc. Lond. B. 357: 937-943.

Mader, H.M., Pettitt, M.E., Wadham, J.L., Wolff, E.W., and Parkes, R.J. (2006). Subsurface ice as a microbial habitat. Geology 34: 169-172.

Maki, L.R., Galyan, E.L., Chang-Chien, M.-M., and Caldwell, R.D. (1974). Ice nucleation induced by Pseudomonas syringae. Appl. Microbiol. 28: 456-459.

Margaritis, A., and Bassi, A.S. (1991). Principles and biotechnological applications of bacterial ice nucleation. Crit. Rev. Biotech. 11: 277-295.

Mazur, P. (1963). Kinetics of water loss from cells at subzero temperatures and the likelihood of intracellular freezing. J. Gen. Physiol. 47: 347-369.

Mazur, P. (1966a). Theoretical and experimental effects of cooling and warming velocity on the survival of frozen and thawed cells. Cryobiology 2: 181-192.

Mazur, P. (1966b). Physical and chemical basis of injury in single-celled microorganisms subjected to freezing and thawing. In Cryobiology. Merman, H.T., ed., Academic Press, New York, pp. 214-315.

Meury, J. (1988). Glycine betaine reverses the effects of osmotic stress on DNA replication and cellular division in Escherichia coli. Arch. Microb. 149: 232-239.

Morita, R.Y. (1975). Psychrophilic bacteria. Bacteriol. Rev. 39: 144-167.

Muryoi, N., Kawahara, H., and Obata, H. (2003). Properties of a novel extracellular cell- free ice nuclei from ice-nucleating Pseudomonas antarctica IN-74. Biosci. Biotechnol. Biochem. 67: 1950-1958.

Nejad, P., and Ramstedt, M. (2006). Presence of quorum-sensing-mediated gene regulation in pathogenic ice-nucleation-active (INA) bacteria. World J. Microbiol. Biotechnol. 22: 1373-1375.

Nejad, P., Ramstedt, M., and Granhall, U. (2004). Pathogenic ice-nucleation active bacteria in willows for short rotation forestry. For. Pathol. 34: 369-381.

Nelson, L.M., and Parkinson, D. (1978). Effect of freezing and thawing on survival of three bacterial isolates from an arctic soil. Can. J. Microb. 24: 1468-1474.

Obata, H., Muryoi, N., Kawahara, H., Yamade, K., and Nishikawa, J. (1999). Identification of a novel ice-nucleating bacterium of Antarctic origin and its ice nucleation properties. Cryobiology 38: 131-139.

Obata, H., Nakai, T., Tanishita, J., and Tokuyama, T. (1989). Identification of an ice- nucleating bacterium and its ice nucleation properties. J. Ferment. Bioeng. 67: 143-147. 41

Panoff, J.-M., Thammavongs, B., and Guéguen, M. (2000). Cryoprotectants lead to phenotypic adaptation to freeze-thaw stress in Lactobacillus delbrueckii ssp. bulgaricus CIP 101027T. Cryobiology 40: 264-269.

Ponder, M.A., Gilmour, S.J., Bergholz, P.W., Mindock, C.A., Hollingsworth, R., Thomashow, M.F., and Tiedje, J.M. (2005). Characterization of potential stress responses in ancient Siberian permafrost psychroactive bacteria. FEMS Microb. Ecol. 53: 103-115.

Raymond, J.A., Christner, B.C., and Schuster, S.C. (2008). A bacterial ice-binding protein from the Vostok ice core. Extremophiles 12: 713-717.

Raymond, J.A., and DeVries, A.L. (1977). Adsorption inhibition as a mechanism of freezing resistance in polar fishes. Proc. Natl. Acad. Sci. USA 74: 2589-2593.

Raymond, J.A., Fritsen, C., and Shen, K. (2007). An ice-binding protein from and Antarctic sea ice bacterium. FEMS Microb. Ecol. 61: 214-221.

Roth, W.G., Leckie, M.P., and Dietzler, D.N. (1985). Osmotic stress drastically inhibits active transport of carbohydrates by Escherichia coli. Biochem. Biophys. Res. Commun. 126: 434-441.

Sakamoto, T., and Murata, N. (2002). Regulation of the desaturation of fatty acids and its role in tolerance to cold and salt stress. Curr. Opin. Microbiol. 5: 206-210.

Sakamoto, A., Murata, A., and Murata, N. (1998). Metabolic engineering of rice leading to biosynthesis of glycinebetaine and tolerance to salt and cold. Plant Mol. Biol. 38: 1011-1019.

Schnell, R.C., and Vali, G. (1972). Atmospheric ice nuclei from decomposing vegetation. Nature 236: 163-165.

Shahjee, H.M., Banerjee, K., and Ahmad, F. (2002). Comparative analysis of naturally occurring L-amino acid osmolytes and their D-isomers on protection of Escherichia coli against environmental stresses. J. Biosci. 27: 515-520.

Sonnichsen, F.D., Davies, P.L., and Sykes, B.D. (1998). NMR structural studies on antifreeze proteins. Biochem. Cell Biol.76: 284–293.

Stead, D. and Park, S.F. (2000). Roles of Fe superoxide dismutase and catalase in resistance of Campylobacter coli to freeze-thaw stress. Appl. Environ. Microb. 66: 3110- 3112.

Steven, B., Briggs, G., McKay, C.P., Pollard, W.H., Greer, C.W., and Whyte, L.G. (2007). Characterization of the microbial diversity in a permafrost sample from the Canadian high Arctic using culture-dependent and culture-independent methods. FEMS Microb. Ecol. 59: 513-523.

Stewert, W.E. Jr., and Bear, L.L. (1998). Concentration Effects of Ice Nucleation Active Bacteria on the Water Nucleation Temperature. In Proc. 23rd Intersoc. Energy. Convers. Eng. Conf. Vol 2. Goswami, D.Y., ed., American Society of Mechanical Engineers, New York, pp. 147.

Telang, A.J., Ebert, S., Foght, J.M., Westlake, D.W.S., Jenneman, G.E., Gevertz, D., and Voordouw, G. (1997). Effect of nitrate injection on the microbial community in an oil field as monitored by reverse sample genome probing. Appl. Environ. Microb. 63: 1785- 1793.

Thieringer, H.A., Jones, P.G., and Inouye, M. (1998). Cold shock and adaptation. BioEssays 20: 49-57.

Tomczak, M.M., Marshall, C.B., Gilbert, J.A., and Davies, P.L. (2003). A facile method for determining ice recrystallization inhibition by antifreeze proteins. Biochem. Biophys. Res. Commun. 311: 1041-1046.

Tsuchiya, Y., Sasaki, K., and Hasegawa, H. (2004). Effect of freeze-thaw repetitions upon the supercooling release ability of ice-nucleating bacteria. J. Biosci. Bioeng. 97: 71- 74.

Vali, G. (1971). Quantitative evaluation of experimental results on the heterogeneous freezing nucleation of supercooled liquids. J. Atmos. Sci. 28: 402-409.

Walker, V.K., Palmer, G.R., and Voordouw, G. (2006). Freeze-thaw tolerance and clues to the winter survival of a soil community. Appl. Environ. Microb. 72: 1784-1792.

Wang, W., and Ben, R.N. (2004). Upregulation and protein trafficking of Aquaporin-2 attenuate cold-induced osmotic damage during cryopreservation. In Vitro Cell. Dev. Biol.- An. 40: 67-70.

Ward, P.J., and DeMott, P.J. (1989). Preliminary experimental evaluation of Snomax snow inducer, Pseudomonas syringae, as an artificial ice nucleus for weather modification. J. Weather Modif. 21: 9-13.

Warren, G., Corotto, L., and Wolber, P. (1986). Conserved repeats in diverged ice nucleation structural genes from two species of Pseudomonas. Nucleic Acids Res. 14: 8047-8060.

Whitman, R.L., Nevers, M.B., and Byappanahalli, M.N. (2006). Examination of the watershed-wide distribution of Escherichia coli along southern Lake Michigan: an integrated approach. Appl. Environ. Microb. 72: 7301-7310.

Williams, P.J. and Smith, M.W. (1989). The Frozen Earth: Fundamentals of Geocryology. Cambridge University Press, Cambridge.

Wilson, S.L., Kelley, D.L., and Walker, V.K. (2006). Ice-active characteristics of soil bacteria selected by ice-affinity. Environ. Microb. 8: 1816-1824.

Woerpel, M.D. (1980). Snow Making. US Patent 4,200,228.

Wu, Z., Qin, L., and Walker, V.K. (2009). Characterization and recombinant expression of a divergent ice nucleation protein from ‘Pseudomonas borealis’. Microbiology 155: 1164-1169.

Yamashita, Y., Nakamura, N., Omiya, K., Nishikawa, J., Kawahara, H., and Obata, H. (2002). Identification of an antifreeze lipoprotein from Moraxella sp. of Antarctic origin. Biosci. Biotechnol. Biochem. 66: 239-247.

Yancey, P.H., Clark, M.E., Hand, S.C., Bowlus, R.D., and Somero, G.N. (1982). Living with water stress: Evolution of osmolyte systems. Science 217: 1214-1222.

Yankofsky, S.A., Levin, Z., Bertold, T., and Sandlerman, N. (1981). Some basic characteristics of bacterial freezing nuclei. J Appl. Meteor. 20: 1013-1019.

Zachariassen, K.E., and Hammel, H.T. (1976). Nucleating agents in the haemolymph of insects tolerant to freezing. Nature 262: 285-287.

Zhang, C., Zhang, H., Wang, L., Gao, H., Guo, X.N., and Yao, H.Y. (2007). Improvement of texture properties and flavor of frozen dough by carrot (Daucus carota) antifreeze protein supplementation. J. Agri. Food Chem. 55: 9620-9626.

Zhao, J.-L., and Orser, C.S. (1990). Conserved repetition in the ice nucleation gene inaX from Xanthomonas camperstris pv. translucens. Mol. Gen. Genet. 223: 163-166.

Chapter 1.2: Gas Hydrates and Microbes

Gas hydrates are ice-like “cages” of stabilized water molecules which can host a

variety of small gas molecules. Methane hydrates are the most common, and at least at

some sites the incorporated methane often results from microbial methanogensis

(Kvenvolden, 1995). Hydrates form naturally at low temperatures and high pressures,

and are therefore abundant in permafrost and continental shelves (Kvenvolden, 1988).

Gas hydrates are a challenge for oil companies since formation within a pipeline can, and

has, had terrible results, including their role in the Piper Alpha and Deepwater Horizon

disasters. The former killed 167 men aboard the oil rig in 1988, while the latter was

responsible for upwards of an estimated 60,000 barrels of oil per day from April 20th to

July 15th, 2010, of pollution into the Gulf of Mexico (Hays & Culvin, Reuters, 2010;

note, estimates are highly variable). Clearly, these incidents demonstrate the impact that

hydrate formation within pipelines can have on human life and the ecosystem

surrounding a resultant oil spill.

Oil companies have been aware of hydrates since the early 20th century

(Hammerschmidt, 1934), and a current hydrate growth inhibition strategy entails adding methanol, a thermodynamic inhibitor to the pipelines as needed (Koh et al., 2002).

Kinetic inhibitors, such as polyvinylpyrrolidone (PVP; Sloan, 1995; Lederhos et al.,

1996), and anti-agglomerants (Behar et al., 1991; Huo et al., 2001) can delay hydrate

formation. However, AFPs (see Chapter 1.1) have recently been investigated as pot ential

45 gas hydrate inhibitors, with promising results. AFPs from a number of organisms have been shown to adsorb to and alter the growth of hydrates. Type I AFP from winter flounder (WfAFP), an insect AFP (CfAFP) from spruce budworm (Zeng et al., 2003),

Type III AFP from eel pout, and a plant AFP from perennial rye grass (LpAFP;

Gordienko et al., 2010) have all been shown to adsorb to tetrahydrofuran (THF) hydrate, a model hydrate. WfAFP, Type III, and LpAFP also delayed or prevented THF hydrate formation as well as propane or natural mixed gas hydrate formation (Zeng et al., 2003;

Gordienko et al., 2010; Ohno et al., 2010). Moreover, they did so more efficiently than

PVP. Since a number of AFPs with different structures and activities show inhibition and they are non-toxic, more study appears to be justified for their use as “green” hydrate inhibitors.

The microbial community structure, and in some cases function, has been studied in hydrate-bearing sediments. As one would anticipate, methanogenic microbes are present within these sediments (Colwell et al., 2008; Kormas et al., 2008). For example, it has been suggested that the metabolic basis of the consortia associated with the

Cascadia Hydrate Ridge sediments, and presumably other hydrates, is methane and sulphide oxidation (Sassen et al., 1994; Boetius & Suess, 2004). In keeping with this,

Bidle et al., (1999) found a predominance of sulphide oxidizers, sulphate reducers and sulphur disproportionators as well as a methanotroph amongst the consortia members within sediment samples from Cascadia. Anaerobic methane oxidation by archaea also occurs, and is believed to involve symbiosis with the sulphate reducing bacteria (SRB).

This entails a reversal of methanogenesis, with SRB as electron acceptors (Hoehler et al.,

1994; Hinrichs et al., 1999; Boetius et al., 2000).

Methane oxidation and sulphate reduction may be concurrent. This is not

surprising since the majority of methane oxidizing archaea (related to

Methanosarcinales) and SRB (related to Desulfosarcina/Desulfococcus) were present in

aggregates in Cascadia hydrate-bearing sediments (Boetius et al., 2000). As well,

Bagwell et al., (2009) found that SRB were more abundant in hydrate-containing

sediments than similar sediments lacking hydrates. One peak of SRB abundance

corresponded with an abundance of the anaerobic methane oxidizers (ANME; within

sediment cores). Such methane oxidizers have been found at other hydrate-rich sites

including the Eel River basin. These included archaea related to known ANME-1 as well

as Methanomicrobiales/Methanosarcinales (Hinrichs et al., 1999). Thus, there seems to

be a clear link between methane oxidation/sulphate reduction and gas hydrates. While

microbes involved in sulphur and methane metabolism are often found, a number of other

microbes are also present in “hydrate sediments”. These included α-, γ-, δ-, ε-

Proteobacteria and Firmicutes (Mills et al., 2003). Members of the α-, β-, γ-, δ-

Proteobacteria, as well as Planctomycetes (Inagaki et al., 2006), Ralstonia (β-

Proteobacteria), and Pseudomonas (γ-Proteobacteria), and some related to the

Cytophaga-Flavobacterium-Bacteriodes (CFB; Marchesi et al., 2001) have also been identified.

While the aforementioned studies and others have investigated microbial communities in the sediments containing hydrates, only Lanoil et al., (2001) has

investigated communities directly associated with hydrates. This group’s report on the abundance and diversity within a physical hydrate structure from the Gulf of Mexico was quite impressive. Microbial abundance was three logarithms lower in the dissociated

‘melted’ hydrate (106 cells per mL) than in the hydrate-bearing sediments. A number of bacteria were identified belonging to the Actinobacteria, Firmicutes, α-, β-, γ-, δ-

Proteobacteria (including SRB), CFB and Thermus. Archaea relating to ANME-1 and

ANME-2 as well as Methanosaeta were also identified. These results indicate that hydrates are a rich source of microbial diversity and some microbial types are found in hydrates as well as the associated sediments, although some are distinct.

Given the promising results of hydrate inhibition by AFPs, and the diverse microbial populations associated with hydrates, there is the potential for the discovery of microbes capable of inhibiting hydrate growth. As well, it is possible that microbes with properties that can adsorb to or even inhibit hydrates, or metabolize the methane, may be found. Such microbes could express proteins similar to AFPs.

Literature Cited:

Bagwell, C.E., Formolo, M., Ye, Q., Yeager, C.M., Lyons, T.W., and Zhang, C.L. (2009). Direct analysis of sulfate reducing bacterial communities in gas hydrate-impacted marine sediments by PCR-DGGE. J. Basic Microb. 49: S87-S92.

Behar, P., Kessel, D., Sugier, A., and Thomas, A. (1991). Advances in hydrate control. Proceedings of the 70th gas processors association conference, San Antonio, TX.

Bidle, K.A., Kastner, M., and Bartlett, D.H. (1999). A phylogenetic analysis of microbial communities associated with methane hydrate containing marine fluids and sediments in the Cascadia margin (ODP site 892B). FEMS Microb. Letters 177: 101-108.

Boetius, A., Ravenschlag, K., Schubert, C.J., Rickert, D., Widdel, F., Gieseke, A., Amann, R., Jørgensen, B.B., Wittle, U., and Pfannkuche, O. (2000). A marine microbial consortium apparently mediating anaerobic oxidation of methane. Nature 407: 623-626.

Boetius, A., and Suess, E. (2004). Hydrate Ridge: a natural laboratory for the study of microbial life fueled by methane from near-surface gas hydrates. Chem. Geol. 205: 291- 310.

Colwell, F.S., Boyd, S., Delwiche, M.E., Reed, D.W., Phelps T.J., and Newby, D.T. (2008). Estimates of biogenic methane production rates in deep marine sediments at Hydrate Ridge, Cascadia Margin. Appl. Environ. Microb. 74: 3444-3452.

Gordienko, R., Ohno, H., Singh, V.K., Jia, Z., Ripmeester, J.A., and Walker, V.K. (2010). Towards a green hydrate inhibitor: imaging antifreeze proteins on clathrates. PLoS ONE 5: e8953.

Hammerschmidt, E.G. (1934). Formation of gas hydrates in natural gas transmission lines. Ind. Eng. Chem. 26: 851-855.

Hays, K., and Colvin, R. (2010). New BP cap stops Gulf oil leak. IFpress.com http://www.lfpress.com/news/world/2010/07/15/14725026.html; accessed July 27, 2010.

Hinrichs, K.-U., Hayes, J.M., Sylva, S.P., Brewer, P.G., and DeLong, E.F. (1999). Methane-consuming archaebacteria in marine sediments. Nature 398: 802-805.

Hoehler, T.M., Alperin, M.J., Albert, D.B., and Martens, C.S. (1994). Field and laboratory studies of methane oxidation in an anoxic marine sediment: evidence for a methanogenic-sulfate reducer consortium. Glob. Biogeochem.Cycles 8: 451-463.

Huo, Z., Freer, E., Lamar, M., Sannigrahi, B., Knauss, D.M., and Sloan, E.D. Jr. (2001). Hydrate plug prevention by anti-agglomeration. Chem. Eng. Science 56: 4979-4991.

Inagaki, F., Nunoura, T., Nakagawa, S., Teske, A., Lever, M., Lauer, A., Suzuki, M., Takai, K., Delwiche, M., Colwell, F.S., Nealson, K.H., Horikoshi, K., D’Hondt, and Jørgensen, B.B. (2006). Biogeographical distribution and diversity of microbes in methane hydrate-bearing deep marine sediments on the Pacific Ocean Margin. Proc. Natl. Acad. Sci. USA 103: 2815-2820.

Koh, C.A., Westacott, R.E., Zhang, W., Hirachand, K., Creek, J.L. and Soper, A.K. (2002). Mechanisms of gas hydrate formation and inhibition. Fluid Phase Equilibr. 194- 197: 143-151.

Kormas, K.A., Meziti, A., Dählmann, A., De Lange, G.J., and Lykousis, V. (2008). Characterization of methanogenic and prokaryotic assemblages based on mcrA and 16S rRNA gene diversity in sediments of the Kazan mud volcano (Mediterranean Sea). Geobiology 6: 450-460.

Kvenvolden, K. (1988). Methane hydrates and global climate. Glob. Biogeochem.Cycles 2: 221-229.

Kvenvolden, K. (1995). A review of the geochemistry of methane in natural gas hydrate. Org. Geochem. 23: 997-1008.

Lanoil, B.D., Sassen, R., La Duc, M.T., Sweet, S.T., and Nealson, K.H. (2001). Bacteria and archaea physically associated with Gulf of Mexico gas hydrates. Appl. Environ. Microb. 67: 5143-5153.

Lederhos, J.P., Long, J.P., Sum, A., Christiansen, R.L., and Sloan, E.D. Jr. (1996). Effective kinetic inhibitors for natural gas hydrates. Chem. Eng. Science 51: 1221-1229.

Marchesi, J.R., Weightman, A.J., Cragg, B.A., Parkes, R.J., and Fry, J.C. (2001). Methanogen and bacterial diversity and distribution in deep gas hydrate sediments from the Cascadia margin as revealed by 16S rRNA molecular analysis. FEMS Microb. Ecol. 34: 221-228.

Mills, H.J., Hodges, C., Wilson, K., MacDonald, I.R., and Sobecky, P.A. (2003). Microbial diversity in sediments associated with surface-breaching gas hydrate mounds in the Gulf of Mexico. FEMS Microb. Ecol. 46: 39-52.

Ohno, H., Susilo, R., Gordienko, R., Ripmeester, J., and Walker, V.K. (2010). Interaction of antifreeze proteins with hydrocarbon hydrates. Chem. Eur. J. DOI: 10.1002/chem.200903201.

Sassen, R., MacDonald, I.R., Requejo, A.G., Guinasso, N.L. Jr., Kennicutt, M.C. II., Sweet, S.T., and Brooks, J.M. (1994). Organic geochemistry of sediments from chemosythetic communities, Gulf of Mexico slope. Geo-Marine Letters 14: 110-119.

Sloan, E.D. (1995). A method for controlling clathrate hydrates in fluid systems. US Patent 5,420,370 (for PVP). International patents also filed.

Zeng, H., Wilson, L.D., Walker, V.K., and Ripmeester, J.A. (2003). The inhibition of tetrahydrofuran clathrate-hydrate formation with antifreeze protein. Can. J. Phys. 81: 17- 24.

Chapter 1.3: Research Objectives

It is clear that microbes have adapted to a number of harsh environments, including those at low temperatures, high pressures or high osmotic potential. Increased osmolyte content, biofilm formation and ice-association activities (antifreeze and ice nucleation protein; AFP and INP activities) all appear to be important determinants for microbial survival (see Chapter 1.1 for references), and are further investigated here. We wondered if there was cross-tolerance between some of these stresses, such as osmotic and freeze-thaw stresses and oil well pressure and low temperature survival, and these were also investigated. It was also important to assess the overall utility of using low temperature selective pressure to identify microbes with a host of survival mechanisms.

There are four overarching questions that are the key objectives of this thesis:

1. Can ice-affinity be used to select for low temperature resistant microbes from soil- derived communities? (Chapter 2)

Can psychrophilic and psychrotolerant microbes from different environments be selectively incorporated into growing ice? If so, what microbes are incorporated and do ice-association activities play a role in the affinity to ice?

2. Are microbial consortia derived from geographically distant soils similarly resistant to freeze-thaw stress? (Chapter 3)

Are the effects of freeze-thaw stress on the abundance and diversity of consortia derived

from temperate and boreal soils comparable? What microbes survive, what is the

abundance of the survivors in the original consortia and what ice-association activities do

the survivors demonstrate?

3. Is there cross-tolerance between freeze-thaw and osmotic stresses? (Chapter 4)

Is the degree of consortia resistance to freeze-thaw stress dependent upon the salinity of

the lake from which the consortia were derived? What microbes survive freeze-thaw

selection and do osmolyte content, biofilm formation, and ice-association activities

contribute to resistance?

4. Can freeze-thaw be used to select for microbes from “produced water” (a water and

crude oil mixture) that demonstrate ice-association activities, or gas hydrate growth

inhibition? (Chapter 5)

What microbes are present in enriched produced water prior to and following selection?

What effect does freeze-thaw selection have on the consortia and can this regime be used

to select for microbes with ice- or hydrate-association activities?

This thesis attempted to answer all these questions and several others (see Appendices).

It is my belief that should the objectives be met, insight has been gained into the ways by which microbes resist these environmental stresses. In addition, this thesis aims to

53 demonstrate the means to select for microbes with certain, industrially-relevant, characteristics. It is also hoped that this will have utility for the future.

Chapter 2: Ice-Active Characteristics of Soil Bacteria Selected by Ice-Affinity

Abstract:

As an initial screen for microorganisms that produce ice-active macromolecules, ice-affinity was used to select microorganisms from soil consortia originating from three temperate regions. Once selected and subsequently purified to single colonies, these microbes were putatively identified by 16S ribosomal DNA sequencing and assayed for various ice-active properties. Ice-affinity selection appeared to select for bacteria with ice-association activities: inhibition of ice recrystallization; ice nucleation; ice-shaping.

Although none of these activities were observed in Paenibacillus amyloliticus C8, others such as Chryseobacterium sp. GL8, demonstrated both ice recrystallization inhibition and ice-shaping activities. Pseudomonas borealis DL7 was classified as a type I ice nucleator, Flavobacterium sp. GL7, was identified as a type III ice nucleator and

Acinetobacter radioresistens DL5 demonstrated ice recrystallization inhibition. In all, 19 different culturable bacteria were selected from the thousands of microbes in late-summer collected soil samples. Many of the selected microbes have been previously reported in glacial ice cores or polar sea ice, and of five isolates that were further characterized, four showed ice-association activities. These results indicate the significant potential of ice- affinity selection even with temperate climate soils, suggesting that sampling in more extreme and remote areas is not required for the isolation of ice-active bacteria.

Introduction:

Microbial survival at subzero temperatures is challenging as ice crystal growth

may damage microbial membranes and increase the cell volume beyond the critical level.

In addition, cells must counter the osmotic gradient produced by the increasing

extracellular solute concentration resulting from freezing (Mindock et al., 2001). Some

microbes lower the ice-nucleation temperature and stabilize cellular fluids and

membranes by producing cryoprotectants, by elevating the proportion of shorter acyl

chains (Mindock et al., 2001) or unsaturated fatty acids (Broadbent & Lin, 1999) in

membranes, or by producing antifreeze proteins (AFPs) (Xu et al., 1998; Kawahara,

2002). Cryoprotectants provide limited protection against ice crystal growth by lowering

the freezing and melting points (Mindock et al., 2001). In contrast, by adsorbing to

embryonic ice crystals, AFPs lower the equilibrium freezing point relative to the melting

point, with the resulting temperature difference termed thermal hysteresis. AFPs also

impede water mobility at ice crystal surfaces resulting in ice recrystallization inhibition

(IRI). Ice recrystallization (IR) is the growth of larger ice crystals at the expense of

smaller crystals at temperatures close to the melting point (Mazur, 1984; Knight &

Duman, 1986; Barrett, 2001). To date, AFPs have been described in a number of bacteria

including Moraxella sp. (Yamashita et al., 2002), Pseudomonas putida (Xu et al., 1998;

Kawahara et al., 2001; Muryoi et al., 2004), Micrococcus cryophilus and Rhodococcus erythropolis (Duman & Olsen, 1993), cyanobacterial mats (Raymond & Fritsen, 2001), and in certain Proteobacteria isolated from Antarctic lakes (Gilbert et al., 2004; 2005). 56

Alternatively, a few Gram-negative bacteria actively promote extracellular freezing at high subzero temperatures by the production of ice-nucleation proteins (INPs)

(Ruggles et al., 1993; Xu et al., 1998). INPs appear to be homologous (Denninger et al.,

1988) and highly repetitive outer membrane proteins that form aggregates (Kozloff et al.,

1983; 1984; Xu et al., 1998), and provide templates for ice nucleation (Ruggles et al.,

1993; Yankofsky et al., 1997). Ice-nucleating activity (INA) has been subdivided into three types, based on the temperature at which the INPs initiate ice formation. Type I nucleators, which nucleate water at or above -5ºC, are the most active natural nucleators with the exception of ice itself (Yankofsky et al., 1997). Type II nuclei are active between -5ºC and -7ºC, and type III nuclei are active below -7ºC (Ruggles et al., 1993;

Cochet & Widehem, 2000). INPs are encoded by a single gene (Tegos et al., 2000) and cells producing each of these types of ice nuclei can be derived from a single colony.

This suggests that INA type is a reflection of culture conditions, cell cycle phase or INP glycosylation and anchoring (Pooley & Brown, 1991; Ruggles et al., 1993; Guiran-

Sherman & Lindow, 1995; Cochet & Widehem, 2000). INA has been described in several plant pathogens and epiphytic bacteria (Tegos et al., 2000) including several species of Pseudomonas (P. syringae, P. fluorescens, P. viridiflava, P. antarctica and P. putida), Xanthomonas campestris, X. ananas, Erwinia herbicola and E. uredovora

(Pantoea ananatis) (Yankofsky et al., 1997; Xu et al., 1998; Kawahara, 2002; Yamashita et al., 2002; Muryoi et al., 2003; 2004).

Microbes with adaptations for subzero survival, including AFPs and INPs, have

been sought in glacial cores, cryoconite holes, accretion ice (Christner, 2000; Christner et

al., 2003; 2005), and Antarctic lakes (Gilbert et al., 2004). Perhaps a more feasible

alternative to sampling from extreme locations, where the proportion of microbes

encoding proteins and other molecules for freeze adaptation is likely high, would be to

devise a method to select for these possibly rare microbes in soil communities from more

accessible temperate locations. A subset of the selected bacteria may have an affinity for

ice via surface AFPs or INPs.

Ice-affinity has recently been developed for the purification of AFPs from crude

protein extracts (Kuiper et al., 2003; Marshal et al., 2004). As water is frozen, molecules that do not adsorb to the ice surface are excluded, whereas AFPs, with an affinity for ice become incorporated into the slowly grown multi-crystalline ‘ice finger’. Here, an

adaptation of this method to select bacteria from natural soil enrichments has resulted in

the identification of bacterial species with ice-affinity.

Materials and Methods:

Microbes and culture conditions: Soil samples were obtained in August and September

prior to the first autumn frost from the upper 1-2 cm of soil from two temperate climate

Canadian locations: Calgary, Alberta (52°6.4’ N, 115°4.78’ W) and the Gould Lake

region, Ontario (44°26’ N, 76°35’ W), and a temperate, near north location, the Daring

Lake region, Northwest Territories (64°52’ N, 111°35’W). Soil samples were collected

just prior to the initiation of the experiments (Daring Lake and Gould Lake samples) or

stored for 1 yr at 4°C (Calgary samples). Most samples were collected in triplicate and 58

composited prior to use to increase homogenity. Soil enrichment cultures were obtained by removing large roots and inoculating 3 mL sterile 10% Tryptic Soy Broth (TSB) [3 g tryptic soy broth (Bacto, Becton, Dickinson and Company, Sparks, MD), 0.1 g KNO3, 0.1

g (NH4)2SO4 and 0.1 g K2HPO4 in 1L deionized water] with 1 g of soil. The mixture was

then shaken at 100 rpm (Gyrotory shaker G2, New Brunswick Scientific Co. Inc. Edison,

NJ) for ~ 48 h at 22°C when the cultures appeared to be in the stationary phase. The

enriched broth (used in subsequent experimentation) was separated from the sediment

with a clinical centrifuge. Isolates derived from the consortia and control bacteria,

including E. coli TG-2 (G. Voordouw, University of Calgary), P. syringae (D. Guttman,

University of Toronto) and Chryseobacterium sp. C14 (V.K. Walker, Queen’s

University) were cultured in 10% TSB as above. Subsequently, all cultures were placed

at 4°C for ~ 48 h prior to experimentation.

Ice-affinity selection: The incorporation of bacteria into slowly grown ice was based on a

modification of a method to purify AFPs (Kuiper et al., 2003). A brass tube or ‘finger’

was sterilized with 70% ethanol followed by sterile distilled water, chilled by circulating

ethylene glycol (slightly below 0ºC) and seeded with a thin layer (~ 1 mm) of

polycrystalline ice derived from sterile water. The finger was submerged in an insulated

beaker containing 50 mL (4°C) sterile 10% minimal salts (6.0 g Na2HPO4•7H2O, 3.0 g

KH2PO4, 0.5 g NH4Cl, in 1.0 L deionized water, pH 7.0) and the bacterial culture. Initial

bacterial counts ranged from 104 to 107 colony forming units per mL (CFU/mL). The

beaker contents were stirred with a magnetic stir bar and monitored to ensure slow

growth of the ice over a ~ 48 h period. The volume of the ice grown was not consistent,

but typically was 7 mL. The ice was rinsed with cold (~4°C) sterile water to remove

traces of the unfrozen fraction (including liquid and cells) and then melted at 22°C. The

water used to rinse the ice was collected independently of the water fraction and was

discarded. The final volume of the melted ice was recorded and the CFU/mL was

determined (in triplicate) for the ice and unfrozen fractions. The percent of bacteria in

the ice fraction was determined as the CFU/mL bound to ice divided by the sum of the

CFU/mL in the ice and the liquid fraction. A partitioning index (PI; Marshal et al., 2004)

was determined as the proportion of culturable bacteria in ice divided by the proportion

of liquid in ice. This value, however, does not take into account the possible variation in

cell viability during the assay.

Identification of bacteria from soil enrichments: Three DNA extraction protocols were

used successively until DNA was obtained. The first protocol entailed centrifuging the

bacterial culture (107 – 108 CFU/mL) for 5 min at 10,000 xg and freezing the pellet at -

20ºC. After resuspension in 200 μL of TE buffer (10 mM Tris-Cl pH 8.0, 1 mM EDTA), the pellet was placed at -20ºC for > 16 h, heated at 95ºC for 3 min and plunged into ice.

Lysate (10 µL) was added to polymerase chain reaction (PCR) tubes and heated at 95ºC

for 3 min in a thermocycler (Perking Elmer Gene Amp PCR system 2400, Woodbridge,

ON). In the second method, the pellet obtained after centrifugation was resuspended in

180 μL of enzymatic lysis buffer (5 mM Tris-HCl, 5 mM Na2EDTA, pH 8, 1 mg/mL

lysozyme), and incubated at 38°C for 1 h, after which 25 μL of proteinase K mix was

added. The remainder of the extraction procedure was as per the DNeasy kit (Qiagen,

Mississauga, ON). Failing these protocols, DNA was extracted as per the SoilMaster™

DNA extraction kit (Epicenter, Madison, WI).

In all cases, the extracted DNA was PCR amplified with universal bacterial 16S

rRNA gene primers (8F: 5´AGAGTTTGATCCTGGCTCAG and r1406

5´ACGGGCGGTGTGTAC) Taq polymerase (1 unit, Fermentas, Burlington, ON), and

the cycling conditions of Telang et al. (1997). The expected 1.4 kb amplified rRNA gene product was confirmed by agarose electrophoresis and was purified with a PCR

pur ification kit (Qiagen).

The purified 16S rRNA gene fragment was sequenced on both strands

(Plateforme de Génomique at the Centre de recherché du CHUL, Quebec, QC) using the

8F and r1406 primers and sequence quality was assessed with CodonCode Aligner

(http://www.codoncode.com/aligner/trial.htm). ‘Manipulate and Display a DNA

sequence’ (http://arbl.cvmbs.colostate.edu/molkit/manip) from Molecular Toolkit was

used to obtain the inverse compliment of the reverse sequence and the sequence was

entered into BLAST from NCBI (http://www.ncbi.nlm.nih.gov/blast; Altschul et al.,

1997), or the Ribosomal Database Project II (http://rdp.cme.msu.edu; Cole et al., 2005) to

determine the nearest phylogenetic relative. Sequences with > 97% similarity to known

species were designated as such followed by a sample identifier. Redundant isolates

were compared using Clustal W (http://www.ebi.ac.uk/clustalw; Thompson et al., 1994)

and isolates with sequence identity of > 98% were deemed to be the same bacteria.

Ice recrystallization inhibition (IRI) and ice nucleation activity (INA) assays: Both the

IRI and INA assays consist of sampling bacterial cultures (107 – 108 CFU/mL) into 10 µL

capillary tubes (Drummond Scientific, Broomall, PA), sealed with silicone stopcock

grease. The prepared capillaries were stored at 4°C for > 1 hr. The samples and control

solutions were microscopically observed (7 – 40x) under polarized light. For the IRI

assay, the capillaries were snap frozen in 50% ethylene glycol chilled to ~ -30°C, and an

image captured. Capillaries were incubated in 50% ethylene glycol at -6°C for > 16

hours and a second image captured (Tomczak et al., 2003). The INA assay used was

modified from previously described assays (Vali, 1971; Maki et al., 1974; Kozloff et al.,

1983). Capillaries containing bacterial cultures (107 – 108 CFU/mL) or controls were

submerged in a 50% ethylene glycol bath at -1°C and the temperature of the circulating

bath was decreased at ~ 0.2ºC/min from -1ºC to -10ºC, at which time the temperature was held constant for ~ 15 min. The temperature at which 90% of the capillaries had frozen

(T90) was deemed the ice nucleation temperature of the most active subpopulation and the

isolates were classified according to INA type. Bacteria with a T90 < -9ºC were not

considered to have INA. For both the IRI and INA assays, the controls were sterile 10%

TSB and a culture of E. coli. Commercial INPs (Wards Natural Science Establishment

Inc., Rochester, NY) or Chryseobacterium sp. C14 were used as additional controls for

the INA and IRI assays respectively.

Ice shaping assay: A nanolitre osmometer (Clifton Technical Physics, Hartford, NY)

was used to determine the effect of the isolates on ice crystal morphology. Bacterial

62 cultures in 10% TSB were suspended in the mineral oil-filled well of the sample holder.

Samples were snap frozen in situ by decreasing the temperature to -40ºC, and the temperature increased until a single ice crystal was obtained and examined under a microscope (40x) mounted above the osmometer (Chakrabartty & Hew, 1991).

Results:

Ice-Affinity Selection: In initial experiments, cultures of P. syringae with a known INP, and E. coli with no known ice-association activities were subjected to ice-affinity selection. Chryseobacterium sp. C14, which had been originally isolated from the

Calgary soil after freeze-thaw selection (Walker et al., 2006), was also used to assess the utility of this selection regime since Chryseobacterium sp. C14 has been shown to have

IRI. The percent of bacteria in the ice fraction was higher for both Chryseobacterium sp.

C14 and P. syringae than for E. coli (Table 2.1). Furthermore, when a mixed culture of

Chryseobacterium sp. C14 (105 CFU/mL; yellow colonies) and E. coli (107 CFU/mL; beige colonies) was used, there was a disproportionate recovery of the Chryseobacterium sp. C14 (103 CFU/mL), relative to E. coli (103 CFU/mL), in the ice fraction relative to the initial CFU/mL (ie. a relative difference of 100-fold). The percent of cells recovered for

Chryseobacterium sp. C14 compared to E. coli was 2.4% and 0.1%, respectively.

Partitioning index (PI) values showed a similar trend. Overall, Chryseobacterium sp.

C14 with a PI of 0.3 and P. syringae with a PI of 5 had a greater affinity for ice than did

E. coli with a PI < 0.01 (Table 2.1), suggesting that ice-affinity selection may be useful for microbial populations. 63

Table 2.1. Ice-affinity selection of control bacteria

% CFU Recovered in Ice Partitioning Index PI range2 Bacteria Average1 Range Average PI1 PI range

Chryseobacterium sp. C14 3 ± 2 0.8 – 5 0.3 ± 0.2 0.1 – 0.6 E. coli 0.15 ± 0.1 0.0 – 0.5 0.008 ± 0.01 0.0 – 0.03 P. syringae 70 ± 20 50 – 97 5 ± 3 2.4 – 9.8 1Standard errors are shown (n = 5-6). 2Proportion of culturable bacteria in ice divided by the proportion of liquid in ice.

When culturable communities from the three geographic locations were

independently subjected to ice-affinity selection, there was a reduction in consortia diversity and abundance, compared to the unselected samples (Fig. 2.1). The closest

phylogenetic relatives of the isolates recovered from the ice fraction were putatively

determined by 16S rRNA gene partial sequencing (Table 2.2), and included multiple

isolates of the same bacterial genera. Particularly notable are the recovery of bacteria in

the Pseudomonas (6 isolates) and Stenotrophomonas (4 isolates),

Chryseobacterium/Flavobacterium (5 isolates) and Acinetobacter (3 isolates) genera

from the Daring and Gould Lake regions, while Paenibacillus (7 isolates) dominated the

selected Calgary consortia. Flavobacterium and Chryseobacterium are listed together as

certain Flavobacterium species have been grouped in the genus Chryseobacterium (Xu et al., 1994). Preliminary studies to assess microbial diversity using 16S rDNA clone

library methods followed by restriction fragment length polymorphism (RFLP) analysis

and sequencing indicated that the original enriched consortia from these regions were not

dominated by the genera selected in these experiments (Wilson et al., 2005). As well, at

least five different isolates were identified in the Calgary samples that had not been

subjected to ice-affinity selection (Walker et al., 2006).

Ice Recrystallization Inhibition and Ice-Nucleation Activity: A portion of isolates recovered by ice-affinity were used for INA (Fig. 2.2) and IRI (Fig. 2.3) assays along with the appropriate controls (Table 2.3). P. syringae and commercial INPs had T90s (the

temperature at which 90% of the samples froze) consistent with type I activity.

Figure 2.1. Colony diversity as indicated by colony morphology differences following culture on 10cm Petri dishes. Samples are from before (left) and after (right) ice-affinity selection of cultured soils from Calgary (A), the Daring Lake region (B) and the Gould Lake region (C). Typically, the initial density was ~104 CFU/mL but ~10 CFU/mL after ice affinity selection, (recovered from the liquid fraction).

Table 2.2. Isolates recovered following ice-affinity.

Sample Nt2 % Identity Nearest phylogenetic relative Accession identifier1 Number

C-1 905-1315 (7) 98-99 Paenibacillus amylolyticus3 AAB115960 C-6 785-1243 (2) 98-100 Bacillus sp. AY822614

DL-1 1285-1302 (2) 99 Bacillus sp. AJ316308 DL-2 1265 (1) 99 Streptomyces tubercidicus AJ621612 DL-3 669 (1) 100 Streptomyces sp. AY754715 DL-5 1261 (1) 99 Acinetobacter radioresistens4 ASZ93445 DL-6 1231-1296 (4) 99-100 Pseudomonas fluorescens D11188 DL-7 981-1295 (2) 100 Pseudomonas borealis5 AJ012712 DL-12 1295 (1) 99 Acinetobacter baumannii AY847284 DL-13 993 (1) 99 Sphingomonas sp. AF395031

GL-1 1302-1308 (3) 99-100 Stenotrophomonas maltophilia AF100734 GL-2 1305 (1) 99 Pseudomonas sp. AY263482 GL-4 1285 (1) 99 Stenotrophomonas maltophilia AY689084 GL-5 1280 (1) 100 Microbacterium foliorum AJ249780 GL-6 1291 (1) 100 Sphingobacteriaceae8 GL-7 1273 (1) 97 Flavobacterium sp.6 AF395549 GL-8 1273 (1) 98 Chryseobacterium sp.7 AY468478 GL-10 1262 (1) 99 Arthrobacter sp. AY512630 GL-11 1290 (1) 99 Acinetobacter baumannii AY847284 GL-12 1264 (1) 100 Microbacteriaceae8 GL-13 1257 (1) 99 Chryseobacterium sp. AY468462 GL-14 1027 (1) 99 Paenibacillus glucanolyticus AB073189 1Sample identifier given to the isolates, and used with the name of the closest phylogenetic relative in the text descriptions. Isolates recovered from Calgary, Daring Lake and Gould Lake soil are denoted as C, DL and GL respectively. 2Number of nucleotides with the number of independent isolates in parentheses. 3Designated Paenibacillus amylolyticus C8 for the subsequent characterization. 4Designated Acinetobacter radioresistens DL5 for the subsequent characterization. 5Designated Pseudomonas borealis DL7 for the subsequent characterization. 6Designated Flavobacterium sp. GL7 for the subsequent characterization. 7Designated Chryseobacterium sp. GL8 for the subsequent characterization. 8Family name determined using the Ribosomal Database Project II, as NCBI BLASTn searches yielded multiple genera with 99% similarity.

Figure 2.2. Ice-nucleation assessment as demonstrated by assays of P. borealis cultures. Bacteria were suspended in 10% TSB at 107 CFU/mL. Frozen samples are indicated by light refraction (white) while unfrozen capillaries appear dark blue or grey under polarized light. Photographs were taken at (A) 0 min (−1°C), (B) 5 min (−2°C), (C) 10 min (−3°C) and (D) 15 min (−4°C). Capillaries are 1mm in diameter.

Figure 2.3. Ice recrystallization inhibition assays on representative bacterial isolates. Bacteria were suspended in 10% TSB at ~ 107 CFU/mL and snap frozen at ~ -30°C in capillaries, which are 1mm in diameter. Images were captured immediately after freezing (top) and after > 16 h at -6°C (bottom). IRI is indicated by an unaltered crystal size, as indicated by the observation of the same ice crystals. From left to right samples include E. coli (1,2), 10% TSB (3,4), P. syringae (5,6), P. borealis DL7 (7,8), Flavobacterium sp. GL7 (9,10), A. radioresistens DL5 (11,12), P. amyloliticus C8 (13,14), Chryseobacterium sp. GL8 (15,16) and Chryseobacterium sp. C14 (17,18).

Table 2.3. Ice-association activities of the recovered isolates.

1 1 1 Sample IRI INA (T90) INA Type Ice Shaping Chryseobacterium sp. GL8 +/- -10.0 + 0.4 °C 0 Minimal P. borealis DL7 +2 -3.7 + 0.5 °C I + Flavobacterium GL7 - -8.3 + 0.4 °C III + A. radioresistens DL5 +3 -9.7 + 0.9 °C 0 Minimal P. amylolyticus C8 - -9.8 + 0.7 °C 0 - Chryseobacterium sp. C14 + NA NA Minimal4 P. syringae +2 -3.0 + 0.3 °C I - E. coli - -10.2 + 1.0 °C 0 - 10% TSB - -9.6 + 1.1 °C 0 - INPs +2 -3.4 + 0.7 °C I - 1Assays were conducted to determine if the samples have IRI (ice recrystallization inhibition); INA (ice nucleation activity) or ice shaping (the shape of microscopic ice crystals did not appear as thin, circular disks). 2Although initially large, the ice crystals did not recrystallize and therefore these bacteria are tentatively classified as exhibiting IRI. 3IR was occasionally observed. 4 Ice shaping activity only after storage at 4°C for a prolonged period of time. NA: samples were not assayed

P. borealis DL7 and Flavobacterium sp. GL7 were classified as type I and type III,

respectively. Paenibacillus sp. C8, Chryseobacterium sp. GL8 and A. radioresistens

DL5 did not exhibit INA (Table 2.3).

In IRI assays, P. borealis DL7, P. syringae (Fig. 2.3) and commercial INPs

initially showed large ice crystals, likely as a result of their INA and consequently

warmer freezing temperatures. According to the Kelvin (Gibbs-Thomson) equation, the warmer the freezing temperature, the larger the resultant ice crystal (Inada & Modak,

2006). Because the ice crystals did not recrystallize further, these bacteria were tentatively classified as exhibiting IRI. In Chryseobacterium sp. C14, the ice crystals

remained small, consistent with IRI activity and there was some evidence of similar

activity in Chryseobacterium sp. GL8 and A. radioresistens DL5. However, there was

variability in its appearance in the latter two isolates, perhaps due to non-optimal culture conditions. Paenibacillus sp. C8 did not demonstrate any IRI activity. It should be noted that although IRI is usually associated with AFP activity, other macromolecules have been reported to prevent the recrystallization of ice (Knight et al., 1995).

Ice Shaping: When ice crystals were formed in the presence of several isolates, the morphology of the resulting crystals were distinct from those found in the controls

(culture media and E. coli). In the presence of P. borealis DL7, ice crystals were hexagonal, rectangular or oval (Fig. 2.4A). The former two shapes are likely similar shapes viewed from different angles. A. radioresistens DL5, P. syringae, Paenibacillus sp. C8, commercial INPs, culture media and E. coli did not exhibit ice-shaping activity

Figure 2.4. Ice shaping assays on representative bacterial isolates. Ice crystal shapes were examined with a microscope and produced in the presence of (A) P. borealis DL7, (B) Flavobacterium sp. GL7 (which is similar to Chryseobacterium sp. C14 after several days at 4ºC), (C) Chryseobacterium sp. GL8, and (D) A. radioresistens DL5 (indistinguishable from P. syringae, Paenibacillus sp. C8, E. coli, INPs and 10% TSB; the latter three are negative controls). Approximate ice crystal sizes are (left to right); top row: 120 μm, 108 μm, 108 μm; bottom row: 41 μm, 58 μm, 103 μm, respectively along the dimension.

(Fig. 2.4D). Oval crystals were formed in the presence of Chryseobacterium sp. GL8 in

50% of the trials (Fig. 2.4C), while Flavobacterium sp. GL7 typically produced an oblong-shaped ice crystal (Fig. 2.4B), similar to that produced by Chryseobacterium sp.

C14, provided it was kept at 4ºC for several days prior to assay.

Discussion:

The isolation and characterization of freeze-tolerant bacteria have long intrigued

ecological microbiologists interested in adaptations that permit survival under extreme

conditions. A newly-developed technique designed for the purification of AFPs has now

been adapted to select for microbes within soil consortia that have an affinity for ice.

Initial experiments showed that of three bacteria, P. syringae with an INP,

Chryseobacterium sp. C14 with a putative AFP, and E. coli with no known freeze-

adaptive macromolecules, only P. syringae and Chryseobacterium sp. C14 adsorbed to

ice (Table 2.1). This was most dramatically demonstrated when a mixed culture

containing 100-fold less numerous Chryseobacterium sp. C14 than E. coli cells was

subjected to ice-affinity fractionation, and proportionately more Chryseobacterium sp.

C14 than E. coli were recovered from the ice fraction (PI of 0.16 and 0.006, respectively

in these experiments). Even though Chryseobacterium sp. C14 is five-six fold more

tolerant to a single freeze-thaw cycle than E. coli (Walker et al., 2006), this difference

alone cannot account for the selective recovery seen here.

The variation in the percent recovery and PI values (Table 2.1) may be due to

differences in physiological states. Consistent with these observations, it is thought that 73

INA may be partially dependent upon cell physiology (Pooley & Brown, 1991; Ruggles

et al., 1993; Guiran-Sherman & Lindow, 1995; Barrett, 2001). Preliminary studies also

suggested that ice-affinity selection was influenced by the initial cell density and thus

these parameters should be further explored. When partially purified protein preparations

are subjected to ice-affinity, PI values of AFPs normally range from 0.9 – 1.3 (Marshal et

al., 2004). As far as we are aware, ice-affinity has not been used hitherto to purify INPs,

but its feasibility is currently being explored. Thus, although conditions may not have

been optimal, ice-affinity certainly reduced the diversity and the abundance of the

microbial population (Fig. 2.1) and resulted in the independent recovery of individual

species more than once. Indeed, 70% (26/37) of the selected isolates were from five

different genera, a result that in itself strongly suggests that ice-associating microbes can be obtained using this technique (Fig 2.1; Table 2.2).

It is notable that many of the bacteria recovered by ice selection have previously been described as associating with ice. For instance, the genera Bacillus, Paenibacillus

(including P. amylolyticus), Arthrobacter, Chryseobacterium/Flavobacterium,

Microbacterium, Stenotrophomonas, Acinetobacter (including A. radioresistens) and

Sphingomonas have all been recovered from glacial ice cores dating from 5-20,000 and

> 500,000 years old (Christner, 2000; Christner et al., 2005). Bacillus aquamarinus,

Stenotrophomonas sp., P. fluorescens, and Sphingomonas sp. were recovered from

Antarctic lakes and reported to have IRI activity (Gilbert et al., 2004). Paenibacillus,

Arthrobacter, Microbacterium, Sphingomonas and Pseudomonas have been isolated from

Greenland ice cores (Miteva et al., 2004). Pseudomonas isolates represented half of the

selected species from the Daring Lake samples, were found in the Gould Lake consortia,

and this genus has been reported in Arctic sea ice (Groudieva et al., 2004). Indeed, it is

an intriguing possibility that ice cores and sea ice may also select for microbes with ice-

affinity analogous to the ice selection method described here.

It was originally hypothesized that certain microbes could preferentially absorb to

the ice finger by surface-exposed AFPs. AFPs have previously been shown to be periplasmic or intracellular (Gilbert et al., 2004) or secreted into the growth media

(Gilbert et al., 2004; Kawahara et al., 2001; Yamashita et al., 2002); our results indicate

that AFPs may also be outer membrane-bound. As a corollary, INP-bearing bacteria

might also be selected. Representative isolates recovered from the ice fraction were thus

tested for IRI and INA. The results of the INA assay with P. syringae demonstrated that

the modified INA capillary-based assay yielded results in accordance with prior reports

(Cochet & Widehem, 2000; Maki et al., 1974). P. syringae and commercial INPs have

been classified as type I ice nucleators with nucleation temperatures of approximately

-3ºC. The most active ice nuclei produced by P. borealis DL7 are active at

approximately -3.7ºC, and thus are also classified as type I ice nucleators. Previous

research has identified a number of Pseudomonas species with INA, however, to the best

of our knowledge, this is the first report to associate INA with P. borealis. Indeed,

subsequent preliminary work shows that this strain has a DNA sequence with some

homology to known INP encoding genes (Z. Wu & V.K. Walker, unpublished). Another selected isolate, Flavobacterium sp. GL7, exhibited INA at approximately -8ºC (type III).

Again, this appears to be the first report to associate INA with

Flavobacterium/Chryseobacterium. When P. borealis DL7, P. syringae and commercial

INPs were used in the IRI assay, large ice crystals were observed after freezing, which did not recrystallize at temperatures close to the melt. In contrast, Flavobacterium sp.

GL7 showed no IRI activity perhaps suggesting that ice crystal size could be used as an assay for very active (type I) INPs.

A. radioresistens DL5, Chryseobacterium sp. C14, and Chryseobacterium sp.

GL8 demonstrated IRI activity, although the IRI activity of the latter isolate was variable.

It is likely that culture conditions and the physiological state affect the expression of this activity as they do for INPs. The presence of slowly growing ice in the ice-affinity selection procedure may have been optimal for the expression of these activities, and thus resulted in the recovery of these cells by ice-affinity. Finally, Paenibacillus sp. C8 did not exhibit INA or IRI. In all, 80% of the isolates chosen for further characterization

(Table 2.3; Fig. 2.2, 2.3, 2.4), or 21% (4/19) of the total number of distinct isolates identified by sequencing showed IRI and/or INA.

Since some bacteria were repeatedly selected that did not show either IRI or INA, ice crystal morphology in the presence of a few chosen isolates was also examined.

Flavobacterium sp. GL7, which we classified as a type III ice nucleator with no IRI activity, produced oblong-shaped ice crystals. In contrast, P. syringae, a type I ice nucleator had no effect on ice crystal morphology (Fig. 2.3, 2.4; Table 2.3). Because the presence of IRI activity coupled with ice-shaping activity is almost invariably associated with AFP activity, it is curious that P. borealis DL7, a type I nucleator, also exhibited IRI and ice-shaping, similar to that reported for P. putida under optimal conditions (Muryoi

76 et al., 2004; Xu et al., 1998). It is thus conceivable that P. borealis DL7 might simultaneously produce an INP and an AFP. Paenibacillus sp. C8 that was obtained from the stored Calgary soil did not affect ice crystal morphology. In initial trials, this bacterium was independently isolated from the Daring Lake soil exposed to two sets of

48 freeze-thaw cycles (data not shown). This bacterium may therefore possess some mechanism to ensure survival at low temperatures and may also have an affinity for ice, possibly related to sporulation (Raymond & Hayama, 1997). Perhaps ice-affinity selection renders three classes of bacteria; those with INA, IRI, or ice-shaping; the latter of which may be a distinct or complementary classification.

The ice finger selected for cells with ice-affinity, irrespective of their geographical origin. Although the most distant locations were separated by 20° latitude, there was overlap in some of the genera selected. Indeed IRI, INA and ice-shaping activities were found in bacteria from both these distant sites, suggesting that ice-active microorganisms may be globally distributed and a selective technique is all that is required to efficiently recover them from non-polar regions. The results presented here demonstrate the extraordinary power of ice to select for a few culturable microorganisms in consortia, even when collected from temperate locations at moderate, non-freezing temperatures. Further, these isolates can have INA, IRI and/or ice-shaping activities, properties long considered to be rather rare.

Acknowledgements:

This work was financially supported by NSERC. We also wish to thank Dr. G.

Palmer for his assistance in the development of the INA assay, and continued technical 77 support. Drs. G. Voordouw, D. Guttman, and P. Grogan are thanked for providing the E. coli TG-2, P. syringae, and for collecting the Daring Lake soil samples, respectively. Dr.

P. Davies is acknowledged for the use of the nanolitre osmometer.

Literature Cited:

Barrett, J. (2001). Thermal hysteresis proteins. Int. J. Biochem. Cell Biol. 33: 105-117.

Broadbent, J.R., and Lin, C. (1999). Effect of heat shock or cold shock treatment on the resistance of Lactococcus lactis to freezing and lyophillization. Cryobiology 39: 88-102.

Chakrabartty, A., and Hew, C.L. (1991). The effect of enhanced α-helicity on the activity of a winter flounder antifreeze polypeptide. Eur. J. Biochem. 202: 1057-1063.

Christner, B.C., Mosley-Thompson, E., Thompson, L.G., Zagorodnov, V., Sandman, K., and Reeve, J.N. (2000). Recovery and identification of viable bacteria immured in glacial ice. Icarus 144: 479-485.

Christner, B.C., Kvitko, B.H. II., and Reeve, J.N. (2003). Molecular identification of bacteria and Eukarya inhabiting an Antarctic cryoconite hole. Extremophiles 7: 177-183.

Christner, B.C., Mosley-Thompson, E., Thompson, L.G., and Reeve, J.N. (2005). Recovery and identification of bacteria from polar and non-polar glacial ice. In Life in Ancient Ice. Rogers, S.O. and Castello, J., ed., Princeton University Press, Princeton, NJ, pp. 209-227.

Cochet, N., and Widehem, P. (2000). Ice crystallization by Pseudomonas syringae. Appl. Microbiol. Biotechnol. 54: 153-161.

Denninger, C.A., Mueller, G.M., and Wolber, P.K. (1988). Immunological characterization of ice nucleation proteins from Pseudomonas syringae, Pseudomonas fluorescens, and Erwinia herbicola. J. Bacteriol. 170: 669-675.

Duman, J.G., and Olsen, T.M. (1993). Thermal hysteresis protein activity in bacteria, fungi and phylogenetically diverse plants. Cryobiology 30: 322-328.

Gilbert, J.A., Davies, P.L., and Laybourn-Parry, J. (2005). A hyperactive, Ca2+ - dependent antifreeze protein in an Antarctic bacterium. FEMS Microb. Letters 245: 67- 72. 79

Gilbert, J.A., Hill, P.J., Dodd, C.E.R., and Laybourn-Parry, J. (2004). Demonstration of antifreeze protein activity in Antarctic lake bacteria. Microbiology 150: 171-180.

Groudieva, T., Kambourova, M., Yusef, H., Royter, M., Grote, R., Trinks, and Antranikian, G. (2004). Diversity and cold-active hydrolytic enzymes of culturable bacteria associated with Arctic sea ice, Spitzbergen. Extremophiles 8: 475-488.

Gurian-Sherman, D., and Lindow, S.E. (1995). Differential effects of growth temperature on ice nuclei active at different temperatures that are produced by cells of Pseudomonas syringae. Cryobiology 32: 129-139.

Inada, T., and Modak, P.R. (2006). Growth control of ice crystals by poly(vinyl alcohol) and antifreeze protein in ice slurries. Chem. Eng. 61: 3149-3158.

Kawahara, H. (2002). The structures and functions of ice crystal-controlling proteins from bacteria. J. Biosci. Bioeng. 94: 492-496.

Kawahara, D., Li, J., Griffith, M., and Glick, B.R. (2001). Relationship between antifreeze protein and freezing resistance in Pseudomonas putida GR12-2. Curr. Microbiol. 43: 375-380.

Knight, C.A. and Duman, J.G. (1986). Inhibition of recrystallization of ice by insect thermal hysteresis proteins: a possible cryoprotective role. Cryobiology 23: 256-262.

Knight, C.A., D. Wen, and Laursen, R.A. (1995). Nonequilibrium antifreeze peptides and the recrystallization of ice. Cryobiology 32: 23-34.

Kozloff, L.M., Lute, M., and Westaway, D. (1984). Phosphatidylinositol as a component of the ice nucleating site of Pseudomonas syringae and Erwinia herbicola. Science 226: 845-846.

Kozloff, L.M., Schofield, M.A., and Lute, M. (1983). Ice nucleating activity of Pseudomonas syringae and Erwinia herbicola. J. Bacteriol. 153: 222-231.

Kuiper, M.J. Lankin, C., Gauthier, S.Y., Walker, V.K., and Davies, P.L. (2003). Purification of antifreeze proteins by adsorption to ice. Biochem. Biophys. Res. Commun. 300: 645-648.

Maki, L.R., Galyan, E.L., Chang-Chen, M., and Caldwell, D.R. (1974). Ice nucleation induced by Pseudomonas syringae. Appl. Microbiol. 28: 456-459.

Marshall, C.B., Tomczak, M.M., Gauthier, S.Y., Kuiper, M.J., Lankin, C., Walker, V.K., and Davies, P.L. (2004). Partitioning of fish and insect antifreeze proteins into ice suggests they bind with comparable affinity. Biochemistry 43: 148-154.

Mazur, P. (1984). Freezing of living cells: mechanisms and implications. Am. J. Physiol. (Cell Physiol.) 247: C125-C142.

Mindock, C.A., Petrova, M.A., and Hollingsworth, R.I. (2001). Re-evaluation of osmotic effects as a general adaptive strategy for bacteria in sub-freezing conditions. Biophys. Chem. 89: 13-24.

Miteva, V.I., Sheridan, P.P., and Brenchley, J.E. (2004). Phylogenetic and physiological diversity of microorganisms isolated from a deep Greenland glacier ice core. Appl. Environ. Microb. 70: 202-213.

Muryoi, N., Matsukawa, K., Yamade, K., Kawahara, H., and Obata, H. (2003). Purification and properties of an ice-nucleating protein from an ice-nucleating bacterium, Pantoea ananatis KUIN-3. J. Biosci. Bioeng. 95: 157-163.

Muryoi, N., Sato, M., Kaneko, S., Kawahara, and Obata, H. (2004). Purification and properties of an antifreeze protein from an ice-nucleating bacterium, Pantoea ananatis KUIN-3. J. Bacteriol. 186: 5661-5671.

Pooley, L., and Brown, T.A. (1991). Effects of culture condition of expression of the ice nucleation phenotype of Pseudomonas syringae. FEMS Microb. Letters 77: 229-232.

Raymond, J. A., and Fritsen, C.H. (2001). Semipurification and ice recrystallization inhibition activity of ice-active substances associated with Antarctic photosynthetic organisms. Cryobiology 43: 63-70.

Raymond, J.A., and Hayama, T. (1997). Sporicidal activities of disinfectants on Paenibacillus larvae. J. Vet. Med. Sci. 59: 953-954.

Ruggles, J.A., Nemecek-Marshall, M., and Fall, R. (1993). Kinetics of appearance and disappearance of classes of bacterial ice nuclei support an aggreagation model for ice nucleus assembly. J. Bacteriol. 175: 7216-7221.

Tegos, G., Vargas, C., Perysinakas, A., Koukkou, A.I., Christogianni, A., Nieto, J.J., Ventosa, A., and Drainas, C. (2000). Release of cell-free ice nuclei from Halomonas elongata expressing the ice nucleation gene inaZ of Pseudomonas syringae. J. Appl. Microbiol. 89: 785-792.

Telang, A.J., Ebert, S., Foght, J.M., Westlake, D.W.S., Jenneman, G.E., Gevertz, D., and Voordouw, G. (1997). Effect of nitrate injection on the microbial community in an oil 81

field as monitored by reverse sample genome probing. Appl. Environ. Microb. 63: 1785- 1793.

Thompson, J.D., Higgins, D.G., and Gibson, T.J. (1994). CLUSTAL W: improving the sensitivity of progressive multiple sequence alignment through sequence weighting, position-specific gap penalties and weight matrix choice. Nucleic Acids Res. 22: 4673- 4680.

Vali, G. (1971). Quantitative evaluation of experimental results on the heterogenous freezing nucleation of supercooled liquids. J. Atmos. Sci. 28: 402-409.

Walker, V.K., Palmer, G.R., and Voordouw, G. (2006). Freeze-thaw tolerance and clues to the winter survival of a soil community. Appl. Environ. Microb. 72: 1784-1792.

Wilson, S.L., Buckeridge, K., Lui, J., Kamei, J., and Henessey, R. (2005). The effects of latitude and culture techniques on the assessment of microbial diversity. Genet. Soc. Can. Bull. 36: 54-55.

Xu, H., Griffith, M., Patten, C.L., and Holmes, B. (1994). New perspectives in the classification of the Flavobacteria: description of Chryseobacterium gen. nov., Bergeyella gen. nov., and Empedobacter nom. rev. Int. J. Syst. Bacteriol. 44: 64-73.

Yankofsky, S.A., Nadler, T., and Kaplan, H. (1997). The presence of complete but masked freezing nuclei in various artificially constructed ice nucleation-active Proteobacteria. Current Microbiol. 34: 318-325.

Chapter 3: Frequency of Freeze-Thaw Resistance in Microbial Communities Derived from Latitudinally-Distant Soils

Abstract:

Enrichment cultures were established from soil samples taken from two sites at the latitudinal extremes, more than 3000 km apart, of the Canadian Shield plateau. Late summer-collected soil consortia were subjected to consecutive freeze-thaw cycles, and

surviving bacteria were putatively identified on the basis of 16S rRNA gene sequence.

Semi-quantitative, real time PCR was used to determine the frequency of a subset of the

resistant microbes in the original communities. Even though the more northerly site was

exposed to longer winters at lower temperatures and large spring-time temperature

fluctuations, freeze-thaw selection similarly decreased the diversity and abundance

profile of both soil communities. Only 8 isolates, representing 6 genera, survived and

these were not abundant prior to selection. When survivors were assayed for antifreeze

and ice nucleation protein activities, over 60% of the isolates had one or both of these

properties. These ice-association phenotypes were not dominated by the northern

consortia, indicating that the two enrichment cultures originating from the distant sites

were seemingly functionally similar, at least with respect to these low temperature and

freeze-thaw adaptations.

Introduction: Becking’s well known principle of “everything is everywhere [but] the

environment selects” (Baas-Becking, 1934; translation from de Wit & Bouvier, 2006), is

viewed as rather deterministic for today’s enthusiasm for stochastic processes in

environmental microbiology. Perhaps in keeping with this notion, the literature relating

to psychrotolerant or psychrophilic microbes indicates that, for the most part, these

microbes are sought in cold habitats. This has lead to the rather consistent recovery of a number of genera, or members thereof, from extremely cold locations such as glaciers and the Arctic or Antarctic. However, given the ubiquity of microbes, and Becking’s principle, we hypothesized that cold-hardy microbes could readily be isolated from soil samples exposed to warmer continental climates. A sufficiently stringent selection regime, such as ice-affinity isolation or multiple sequential freeze-thaw cycles (Walker et

al., 2006; Wilson et al., 2006), should allow the recovery of rare microbes with low

temperature resistant phenotypes.

Psychrotolerant and psychrophilic microbes have evolved a number of

adaptations to counter the challenges posed by freeze-thaw stress, including the

production of cold shock or cold adaptive proteins (CSPs/CAPs) to act as chaperones

(Beaufils et al., 2007), or aid transcription and translation (Jones et al., 1987; La Teana et

al., 1991). Cryoprotectants, including sugars such as glycerol and trehalose (Mazur,

1966; Calcott & MacLeod, 1975b; Panoff et al., 2000), can accumulate to aid freeze

tolerance by: limiting permeability barrier damage, maintaining protein hydration and

acting as osmolytes to counter freezing desiccation (Calcott & MacLeod, 1975a, 1975b;

Izawa et al., 2004). Alterations in membrane fatty acid composition, including increased

proportions of unsaturated and branched chain fatty acids, help maintain membrane

fluidity (Klein et al., 1999; Gomez Zavaglia et al., 2000; Chattopadhyay & Jagannadham,

2001). Even in light of a number of adaptations, freezing and thawing remains detrimental to microbes and microbial communities (eg. Yergeau & Kowalchuk, 2008),

though to an extent, these communities are able to recover (Sawicka et al., 2010). This

however, requires rapid responses and reversals of those responses upon freezing and

subsequent thawing (Schimel et al., 2007). The detrimental impact is in part due to ice

formation, which in turn is dependant on temperature as well as cellular water content

(Mazur, 1963). Thus adaptations directly related to the presence of ice are also

important. Microbes can avoid freezing and the physiological and physical damage due

to ice crystal growth by decreasing extracellular ice nucleation temperatures and/or

regulating the temperature of freezing and the shape and growth of external ice crystals

(Xu et al., 1998).

Antifreeze proteins (AFPs) and ice nucleation proteins (INPs) directly associate

with ice, and have both been reported in microbes (eg. Deninger et al., 1998; Gilbert et

al., 2004). AFPs adsorb to ice thereby decreasing the freezing temperature relative to the

melting temperature, creating a thermal hysteric gap, and also modifying the ice crystal phenotype. Ice recrystallization, or the energetically-favoured growth of larger crystals at the expense of smaller ones, is also inhibited (Raymond & DeVries, 1977; Knight &

Duman, 1986). Finally, AFPs may also aid in the stabilization of membranes (Kun et al.,

2008). INPs in contrast, precipitate extracellular ice formation by acting as a template for

85 ice growth at high subzero temperatures, thus decreasing the likelihood of intracellular ice formation (Zachariassen & Hammel, 1976).

It is not known if low temperature adaptative proteins such as AFPs and INPs are disadvantageous under non-selective conditions in temperate climates during the summer months, or if they have other roles. It is, however, predicted that the frequency of freeze- thaw periods in high latitudes will increase with climate change (Henry, 2008; Joseph &

Henry, 2008) and, therefore, an increased understanding of the effects of this stress on soil microbial communities and possible resistance mechanisms, such as AFP and/or INP production is also timely. We, therefore, collected soil samples from broadly the same geological formation but at two different latitudes and subjected them to multiple freeze- thaw cycles. Our aims were to estimate the relative abundance of any resistant microbes, survey survivors for ice-association properties such as AFP and INP production, and to demonstrate that microbes with these activities are present in soils from different latitudes.

Materials and Methods:

Soil Samples and Culture Conditions: Soil samples were obtained in the late summer from the top 1-2 cm of organic soil from two locations in the Canadian Shield, a plateau where Precambrian granite and gneiss outcrops have been exposed by glaciers and are characterized by very thin soil. Soil from the Gould Lake watershed, Ontario (44º26´N,

76º35´W; pH ~6.5) overlays bedrock and the Daring Lake watershed, Northwest

Territories (64º52´N, 111º35´W; pH ~6.5) is underlain by permafrost and bedrock. Both soils are subjected to a continental climate, but with the latter experiencing a significantly 86

longer and colder winter [~156 vs. ~82 frost-free days, with mean February temperatures

of approximately -6°C (Floyd Connor, QUBS weather station, unpublished data) and

-27°C (Bob Reid, INAC, Water Resources Division, unpublished data), respectively].

Samples, with large roots removed, were collected in sterile tubes. For logistical reasons, the samples were transported and stored briefly at 4°C (Gould Lake) or -20°C (Daring

Lake).

Soil enrichment cultures were obtained by culturing 1 g of soil in > 3 mL 10%

Tryptic Soy Broth (TSB) [3 g tryptic soy broth (Bacto, Dickinson and Company, Sparks,

MD, USA), 0.1 g KNO3, 0.1 g (NH4)2SO4 and 0.1 g K2HPO4 in 1L deionized water], and

shaken at 100 rpm (Gyrotory shaker G2, New Brunswick Scientific Co. Inc. Edison, NJ,

USA) for ~ 48 h at 22°C. Isolates derived from the enrichment culture and control

bacteria (E. coli TG-2 and Chryseobacterium sp. C14) were cultured in 10% TSB as

above. Subsequently, all cultures were placed at 4°C overnight prior to freeze-thaw

treatment, or for ~48 h prior to assay. Transformed E. coli (see below) were cultured in

Lauria Broth [LB; 10g NaCl, 10g Bacto-Tryptone and 5g yeast extract in 1L deionized

water, pH 7.0] overnight at 37°C, with agitation.

Clone Library Construction and Analysis: Clone libraries were constructed to estimate

the community richness of the soil enrichment cultures prior to selection. These libraries

were assembled in duplicate by extracting DNA from the enriched soil samples

(SoilMasterTM DNA extraction kit; Epicentre, Madison, WI, USA). The 16S rRNA gene

sequence was PCR amplified from the extracted DNA with ‘universal’ bacterial 16S

rRNA gene primers (8F: 5´-AGAGTTTGATCCTGGCTCAG-3´ and r1406 5´-

ACGGGCGGTGTGTAC-3´; Sigma-Genosys, Oakville, ON, Canada; Hicks et al., 1992;

Lane et al., 1985) Taq polymerase (1 unit, Fermentas, Burlington, ON, Canada), and the cycling conditions of Telang et al., (1997), or the T7Select System (without hot start;

Novagen, Madison, WI, USA). The expected 1.4 kb amplified 16S rRNA gene sequence product was confirmed by agarose gel electrophoresis. The amplified and purified

(Qiagen, Mississauga, ON, Canada) product was cloned into the pCR2.1 vector, and subsequently transformed into E. coli DH5-α Top 10´ cells (TOPO cloning, Invitrogen,

Carlsbad, CA, USA). The inserted 16S rRNA gene of randomly selected clones was

PCR amplified either as above from isolated vector (Miniprep, Qiagen, Mississauga, ON,

Canada), or directly from the clone with the M13 forward (5´-

GTAAAACGACGGCCAG-3´) and reverse (5´-CAGGAAACAGCTATGAC-3´) primers

(TOPO cloning, Invitrogen, Carlsbad, CA, USA; Cortec, Kingston, ON, Canada).

The libraries were screened by restriction fragment length polymorphism (RFLP) analysis, which consisted of two single restriction digests, Hae III (5´-GG/CC-3´) or HhaI

(5´-G/CGC-3´) per clone (as per recommendations, Fermentas, Burlington, ON, Canada).

After digestion (37°C for 1 h), the DNA was visualized on a 1.0 – 1.5% agarose gel and the cloned inserts categorized according to their RFLP fingerprint. Clones with unique

RFLP patterns were sequenced on both strands (Plateforme de Génomique at the Centre de recherché du CHUL, Quebec, QC, Canada; or Mobix, Hamilton, ON, Canada) using the 8F and r1406 or the M13 forward and reverse primers. Sequence quality was assessed with CodonCode Aligner (http://www.codoncode.com/ aligner/trial.htm). The

88 inverse complement of the reverse sequence was obtained using “Manipulate and Display a DNA sequence” by Molecular Toolkit (http://arbl.cvmbs. colostate.edu/molkit/manip).

Completed sequences were entered into BLASTn from NCBI

(http://www.ncbi.nlm.nih.gov/blast; Altschul et al., 1997), or the Ribosomal Database

Project II (http://rdp.cme.msu.edu; Cole et al., 2005, 2009) and the putative identity of the microbe from which the insert was derived was determined based on the closest phylogenetic relative with > 97% identity.

Freeze-Thaw Selection: The enrichment cultures, without added cryoprotectants, were treated with multiple freeze-thaw cycles using an automated cryocycler wherein samples were frozen (-18°C) and thawed (5°C) on an hourly basis, with a single cycle consisting of one hour at each temperature. In practice, the samples were at 0°C or below for ~95 min each cycle (Walker et al., 2006). Following the first 48 cycles, the cultures were used to inoculate fresh media and treated with a subsequent 48 cycles. Freeze-thaw tolerance was monitored periodically during the selection by removing samples and determining their viable cell counts (colony-forming units per mL; CFU/mL).

Chryseobacterium sp. C14, with known freeze-thaw resistance (Walker et al., 2006) and

E. coli TG-2 with no known ice-association properties (Wilson et al., 2006) were used as the positive and negative freeze-thaw cycling controls, respectively. Silver iodide crystals were added to the controls and freeze-thaw resistant isolates (see below) to ensure nucleation of samples (Walker et al., 2006).

Cells that survived both sets of 48 freeze-thaw cycles were deemed to be freeze-

thaw resistant and were isolated as monocultures. The 16S rRNA gene from

morphologically distinct (with redundancy) isolates was PCR amplified and sequenced

with universal bacterial 16S rRNA gene primers 8F and r1406, or 530F (5´-GTGCCAG-

CMGCCGCGG-3´) and R1492 (5´-TACGGYTACCTTGTTACGACT-3´; Cortec,

Kingston, ON, Canada; Lane, 1991 & http://openwetware.org/wiki/Bacterial_species_ identification#Universal_Bacterial_Primers). Isolates with similar sequences were compared using Clustal W (http://www.ebi.ac.uk/clustalw/; Thompson et al., 1994), isolates with sequence similarity of > 98% were deemed to be the same and were

recorded once per genus. These isolates were subsequently subjected to one set of 48

freeze-thaw cycles to assess their respective resistance.

Phylogenetic Tree: 16S rRNA gene sequences obtained from the clone libraries and from

the freeze-thaw resistant isolates were aligned with Clustal W (http://www.ebi.ac.uk/

clustalw/; Thompson et al., 1994). PAUP* (Swofford, 2002) was used to construct the

phylogenetic tree (neighbour joining, kimura 2).

Real-Time Semi-Quantitative PCR (q-PCR): Semi-quantitative PCR (modified from

Shelburne et al., 2000) was performed (QuantiTect SYBR Green PCR Kit, Qiagen,

Mississauga, ON, Canada) to estimate the relative abundance of genera corresponding to

certain isolates in the original enrichment cultures. Universal, Bacillus, Paenibacillus

and Pseudomonas specific primers were used (Table 3.1; primers from Cortec, Kingston,

Table 3.1. QPCR primer sequences and threshold cycles. rRNA genea Forward Primer Reverse Primer Primer Gould Lake Daring Lake (µM) Percent of Percent of Universalb Universalb (Ct)c (Ct)c Universal CCATGAAGTCGGAATCGCTAG GCTTGACGGGCGGTGT 0.4 100.00+0.00 100.00+0.00 (89 bp) (Shelburne et al., 2000) (Shelburne et al., 2000) (17.43+1.10) (14.66+0.65) Universal TCCTACGGGAGGCAGCAGT GGACTACCAGGGTATCTA… 0.3 100.00+0.00 100.00+0.00 (466 bp) (Nadkarni et al., 2002) (Nadkarni et al., 2002) (14.78+0.28) (12.76+0.54) Bacillus CTAACCAGAAAGCCACGGC CCCAGTTTCCAATGACCCT 0.1 20.47+1.91 1.09+0.33 (Chiang et al., 2006)d (Merrill et al., 2003)e (19.19+0.33) (21.38+1.03) Paenibacillus GCTCGGAGAGTGACGGTACCTGAGA CTACCAGGGTATCTAATCC 0.2 0.48+0.21 0.05+0.03 (Shida et al., 1997) (Steven et al., 2007) (31.68+0.93) (33.24+1.66) Pseudomonas GACGGGTGAGTAATGCCTA CACTGGTGTTCCTTCCTATA 0.3 4.72+1.84 20.96+1.14 (Spilker et al., 2004) (Spilker et al., 2004) (18.79+0.72) (24.41+0.22) a 16S rRNA genes were amplified using either ‘universal’ or genus-specific primers. b Means and standard deviations of the amplified product as a percent of the amplified product obtained with the universal primers. c Amplified product was estimated using the threshold cycle (Ct) determined by the level of intercalated fluorescent dye. Means and standard deviations shown. d Inverse compliment of primer No.7. e Inverse compliment of primer Bac629F.

ON, Canada). Primers specific to other isolated genera were unavailable and could not

be generated even after numerous attempts. The primer concentrations used yielded the most efficient reaction of the primer titrations. Standard curves [cycle threshold (CT) versus DNA concentration] were determined in duplicate using the 16S rRNA gene from the isolates. For the experimental reactions (in triplicate), template DNA were dilutions of the original consortia used to construct the Daring and Gould Lake clone libraries. In all cases, each reaction condition was duplicated within each QPCR reaction, and no template controls were included. The average CT was calculated, as was the proportion of the total 16S rRNA gene present, which in turn was determined by averaging the CTs from both pairs of universal primers.

Characterization of the Freeze-Thaw Resistant Isolates: Ice recrystallization inhibition

(IRI) and ice-shaping assays were performed as previously described with whole cell

cultures (Wilson et al., 2006), except the capillaries, containing whole cell culture, for the

IRI assay were snap frozen at ~ -25 to -30°C in a dry ice and ethanol bath. Samples in

which there was no visible growth of ice crystals over the >16 h incubation at -6°C were deemed to exhibit IRI activity (Tomczak et al., 2003). Ice-shaping activity was assessed using a nanolitre osmometer (Clifton Technical Physics, Hartford, NY, USA) and a Leitz

Dialux 22 microscope. Ice-shaping was indicated by the resulting ice crystal having a shape other than a thin disk (Chakrabartty & Hew, 1991). Ice nucleation activity (INA) assays consisted of loading one microlitre droplets of cultures (approx 108 CFU/mL) onto

a polarizing film that rested in a covered ethylene glycol bath. The bath temperature was

92 decreased by approximately 0.2°C/min from -1°C to -20°C. As a semiquantitative estimate, the temperature at which >90% of the droplets froze (T90) was deemed to be the ice nucleation temperature and isolates were classified according to INA type, with a lower limit of activity at –9°C; quantified results of a subset of nucleation profiles are shown as log nuclei/mL vs. temperature (modified from Maki et al., 1974; Vali, 1971).

To confirm that the cryocycler selected for microbes demonstrating AFP and or

INP activity, microbes (n=8) were picked at random from each Daring and Gould Lake unselected consortia plates. These 16 microbes were assayed for IRI, ice-shaping and

INA activity. For all assays, Chryseobacterium sp. C14, fish Type III AFP (10mg/mL;

A/F Protein, Inc., Waltham, MA, USA) or purified recombinant AFPs, prepared as described in (Gordienko et al., 2010) were the positive controls for the IRI assays, while commercial INPs (Pseudomonas syringae preparation; Wards Natural Science

Establishment, Rochester, NY, USA) was the positive controls for the INA assays. In all cases, E. coli TG-2 and 10% TSB were used as the negative controls.

Statistical Analysis: The efficacy (coverage) of the clone libraries was determined on the basis of the total cultured richness captured by the random sampling of clones, equal to

[1-(n/N)], where (n) is the number of clone types encountered once, and (N) is the total number of clones (Good, 1953). The Shannon-Weaver Index was used to assess the diversity of the initial consortia, based on the clone libraries, as well as the diversity retained after freeze-thaw selection, based on observed morphologies and sequencing

93 results. Finally, Student’s T-tests were used to identify significant differences (p < 0.05) between and within the freeze-thaw selection curves, and for the diversity reduction.

Results:

Clone Library Construction and Analysis: Clone libraries were constructed to estimate the richness of the Daring and Gould Lake soil derived enrichment cultures prior to freeze-thaw selection. Table 3.2 summarizes of the results of the clone library, showing consortia richness before selection, with redundancy removed (ie. each clone type is represented once), as well as a complete list of the isolates recovered post freeze-thaw treatment. The putative identity (based on NCBI BLASTn or RDP II search results), sequence length used, accession number and percent similarity to the most likely match are indicated. Table 3.2 also summarizes the results of the RFLP analysis, which was used to determine the efficacy or coverage of the clone libraries. The clone libraries

(Table 3.2, Fig. 3.1) for the cultured Daring and Gould Lake consortia accounted for 0.63 and 0.43 respectively, of the total cultured richness captured by the random sampling of clones (coverage; Table 3.3) containing the 16S rRNA gene inserts.

Freeze-Thaw Selection: Cryocycler selection was used to obtain freeze-thaw resistant members of the initial enriched consortia, as indicated by survival following two sets of

48 freeze-thaw cycles. The positive control, Chryseobacterium sp. C14, retained high viability (as determined by CFU/mL) over the duration of the treatment, whereas the negative control, E. coli TG-2, lost all viability. There was no significant difference in the survival curves between the first and second cycle set for the soil consortia from 94

Table 3.2. Consortia richness before and after freeze-thaw selection.

Daring Lake – Before Selectiona Daring Lake - After Selectionb Putative Identity Accession N (% # RFLP Putative Identity Accession N (% and Sample IDc Numberd Match)e Patterns f and Sample ID Number Match) Actinobacterium sp. D3C AY661610 1169 (97) NA C. piscium DL11* DQ862541.1 1224 (97) Arthrobacter sp. D3E AB167248.7 1164 (98) NA Paenibacillus sp. YIB AJ495806.1 1238 (98) Bacillus sp. D5B AY748912.1 1236 (100) NA P. borealis YIC AJ012712.1 1257 (100) Unidentified Y12 AJ575723.1 1309 (95) 1 Pseudomonas sp. DL13*† DQ011923.1 1248 (100) Unidentified Y3 AF532770.1 1306 (89) 1 B. fungorum Y7 AJ544690.1 1293 (98) 2 Burkholderia sp. DL17 DQ118949.1 1388 (98) 2 Burkholderia sp. Y2 AY178076.1 1305 (98) 1 Burkholderia sp. Y8 DQ118949.1 1241 (98) 1 Burkholderia sp. Y9 AF215704.1 1243 (98) 1 Enterobacteriaceae DL9 NA 1228 (100) 2 E. coli DL3 AB305017.1 1146 (99) 7 Unidentified Y1 EF111071.1 1320 (96) 1 L. rhizovicina D7D AJ580498.1 1247 (99) NA Paenibacillus sp. DL4 DQ339607.1 1261 (99) 1 Photorhabdus sp. DL16 AM084246 1189 (100) 2 Propionibacterium sp. Y4 AM410900.1 1239 (100) 1 P. borealis DL7 AJ012712.1 1219 (99) 2 Pseudomonas sp. D5A DQ011926.1 1179 (99) NA Unidentified Y6 AF409002.1 1247 (96) 1 H = 2.53 H = 0.21 Gould Lake – Before Selection Gould Lake – After Selection Acinetobacter sp. GL5 EF103570.1 1113 (99) 1 Bacillus sp. G1a1*† AY965249.1 1156 (99) Arthrobacter sp. G3B AJ785759.1 1135 (99) NA Buttiauxella sp. G2b1* DQ223872.1 1151 (99) B. pocheonensis GL3 AB245377.1 1235 (98) 1 Enterobacteriaceae G3b1* NA 1266 (NA) Bacillus sp. K2 AJ920000.1 1281 (99) 5 P. agglomerans GLY AF157694 854 (99) Bacillus sp. K6 DQ985273.1 1285 (99) 1 95

C. indoltheticum K8 ATCC27950 1307 (97) 1 Chryseobacterium sp. GL8 DQ673675.1 1220 (99) 1 Clostridium sp. GL7 DQ479415.1 1207 (99) 1 E. ludwigii K4 AJ853891 1341 (98) 1 E. persicina G3E AJ937837.1 1229 (99) NA Flavobacterium sp. GL10 EU707556.1 1213 (99) 1 Oxalobacteraceae GL4 AY429715.1 1233 (97) 1 Paenibacillus sp. GL6 AM162313.1 1121 (99) 1 P. acnes K11 AB108483.1 1322 (99) 1 Pseudomonas sp. G3F AY263482.1 1175 (99) NA S. plicatus G3H AY966000.1 1178 (99) NA Unidentified K1 DQ819169.1 1266 (93) 1 H = 2.71 H = 0.25 * and † denote IRI and Type 1 INA activity respectively. a Richness prior to selection based on the 16S rDNA clone libraries. Diversity is indicated as calculated with the Shannon-Weaver index (H; Shannon & Weaver, 1949). b Richness following selection based on 16S rDBA sequencing of morphologically distinct isolates. Diversity is indicated as calculated with the Shannon-Weaver index (H; Shannon & Weaver, 1949). c Putative identity of the isolate, based on the database search results, followed by the sample identifier assigned in the laboratory. d Accession number corresponding to the closest match in the database. e N is the number of nucleotides entered into the databases, the percent identity between the isolate and it’s closest match is indicated in parenthesis. f Number times the RFLP banding pattern corresponding to this isolate occurred in the clone libraries, used for statistical analysis of the clone library (Shannon-Weaver index and Coverage).

Burkholderia sp. DL17 53 54 B. fungorum Y7 10 Burkholderia sp. Y8 10 Burkholderia sp. Y9 99 Burkholderia sp. Y2 Oxalobacteraceae GL4 96 98 S. plicatus G3H Unidentified Y1 55 L. rhizovicina D7D Unidentified Y3 98 P. borealis DL7 91 P. borealis YIC 10 Pseudomonas sp. G3F Pseudomonas sp. D5A 53 Pseudomonas sp. DL13* Acinetobacter sp. GL5 58 Enterobacteriaceae G3b1* 87 E. ludwigii K4 10 Buttiauxella sp. G2b1* 76 P. agglomerans GLY E. coli DL3 10 10 Enterobacteriaceae DL9 Photorhabdus sp. DL16 98 54 E. persicina G3E Unidentified K1 93 C. piscium DL11* 10 C. indoltheticum K8 99 Chryseobacterium sp. GL8

10 Flavobacterium sp. GL10 Unidentified Y6

10 63 Unidentified Y12 10 Bacillus sp. D5B 86 Bacillus sp. K6 B. pocheonensis GL3 10 89 Bacillus sp. G1a1* 10 10 Bacillus sp. K2 10 Paenibacillus sp. DL4 97 10 Paenibacillus sp. YIB Paenibacillus sp. GL6

Clostridium sp. GL7 10 Arthrobacter sp. D3E 10 Arthrobacter sp. G3B Actinobacterium sp. D3C Propionibacterium sp. Y4 P. acnes K11

0.1 Figure 3.1. Phylogenetic tree representing consortia richness prior to and following freeze-thaw selection, based on clone library or isolate sequencing, respectively. constructed using the PAUP software package. Gould Lake freeze-thaw resistant isolates are bolded, while Daring Lake isolates are underlined and bolded. The asterisk indicates ice-association activity (ice nucleation activity, ice recrystallization inhibition and/or ice- shaping). Ps. denotes Pseudomonas.

Table 3.3. Coverage represented by the clone libraries. Sample # of Clones # of RFLP # of Unique Coveragea Analyzed Fingerprint Groups RFLP Patterns Daring Lake 27 16 10 0.63 Gould Lake 28 19 16 0.43 a Coverage: C=1-(n/N). Where C = coverage, n = # of clone types encountered once, N = total # of clones (Good, 1953).

either location (Fig. 3.2, first 48 cycles shown, T-test; p=0.19-0.59 and 0.11-0.52 over 3-

48 cycles, with p=0.26 and 0.49 at 48 cycles, respectively, for the Gould and Daring Lake curves). Although there was little loss in viability in the soil cultures after the first three freeze-thaw cycles, the average viable cell number was reduced 50-fold after 12 cycles and almost three orders of magnitude after 24 cycles. At the end of 48 cycles the viable cell number was reduced more than 1000-fold for both populations with a significantly higher recovery of colonies in the Gould Lake samples (T-test; p=0.04; first set of 48 cycles).

As well as a decrease in cell abundance, the number of genera and diversity present in both consortia over the course of each set of experiments also decreased.

Overall, after two cycle sets of freeze-thaw selection, the diversity of identified species in each consortium significantly decreased by ~10-fold (Table 3.2; Shannon-Weaver index,

H values). A phylogenetic tree (Fig. 3.1) constructed using the 16S rRNA gene sequences from the original cultures and the recovered, resistant isolates (Table 3.2) shows that surviving bacteria were not phylogenetically grouped. As well, they were not all represented in the original libraries, suggesting their low initial abundance. Finally, when individual isolates were subjected to 48 freeze-thaw cycles, Pseudomonas sp. DL13 lost 5 logarithms of viability, Paenibacillus sp. Y1B lost 7 logarithms of viability, and the remaining 6 isolates lost ~1-3 logarithms of viability.

Real-Time Semi-Quantitative PCR (Q-PCR): Real time Q-PCR using universal- and genus-specific primers was used to assess the relative abundance of the genera in the

Figure 3.2. Average colony forming units (CFU) per mL for the Daring Lake (NT) and Gould Lake (ON) consortia over 48 freeze-thaw cycles. Chryseobacterium sp. C14 and E. coli TG-2 were used as relatively freeze-thaw resistant and susceptible controls, respectively. Each culture was tested in triplicate; in several cases, error bars on the data points are too small to be depicted.

100

original unselected soil, for a subset of the isolates. Paenibacillus accounted for 0.001

and 0.005 of the initial Daring and Gould Lake consortia, respectively (Table 3.1). As

well, Bacillus represented 0.01 and 0.21 and Pseudomonas 0.21 and 0.05 of the

unselected cultured population from the Daring and Gould Lake samples. Combined,

these genera represent less than 0.26 of the total 16S rRNA genes in the initial

enrichment cultures. Multiple attempts to estimate the proportion of the initial cultures

represented by another isolate, Chryseobacterium, using a variety of primers (both

published and designed) as well as reaction conditions (annealing temperatures and

MgCl2 concentrations) did not yield quantifiably-reliable amplified products and subsequent sequencing did not consistently include Chryseobacterium. These observations indicate that Chryseobacterium was present at a rather low abundance in the initial cultures. Indeed, sequences corresponding to this isolated genus were not found in the Daring Lake library (Table 3.2).

Characterization of Freeze-Thaw Selected Isolates: Freeze-thaw resistant isolates were assayed (n=3-5) for ice-association activities indicative of antifreeze or ice nucleation activities. Two isolates derived from Daring Lake (C. piscium DL11 and Pseudomonas sp. DL13) and three isolates from Gould Lake (Bacillus sp. G1a1, Buttiauxella sp. G2b1 and Enterobacteraceae G3b1) inhibited ice recrystallization similar to that seen for Type

III AFP or Chryseobacterium sp. C14 (Fig. 3.3). A number of these isolates initially

yielded large ice crystals, and since these did not appear to recystallize further, they were

classified as exhibiting IRI activity. Buttiauxella sp. G2b1 demonstrated ice-shaping

101

Figure 3.3. Representative ice-association activity assays. I, ice recrystallization inhibition assay of a freeze-thaw survivor, Pseudomonas sp. DL13. Digital images were captured prior to (I A) and following (I B) 16 h incubation at -6°C, capillary tubes are 1mm in diameter. II, ice-shaping assay, frame (II A) is a representative crystal morphology in the presence of Buttiauxella sp. G2b1 (~46 μm), while frame (II B) shows Chryseobacterium sp. DL1 1 (~54 μm), but represents the remainder of the isolates with the same phenotype.

102

activity (Fig. 3.3). Two isolates, Pseudomonas sp. DL13 and Bacillus sp. G1a1 had type

1 ice nucleation activity (see Yankofsky et al., 1981 for INP activity classification type),

similar to that seen for commercial P. syringae preparations (Fig. 3.4). In all, 5 of the 8

isolates recovered following freeze-thaw selection demonstrated one or more ice-

association activities (Table 3.2). Comparatively, of 16 microbes randomly picked from

the Daring and Gould Lake derived consortia before selection, none showed any ice-

association activities.

Discussion:

Prospecting for microbes with ice-association properties is typically undertaken in

extreme environments. In contrast, we made collections prior to the first frost at two

relatively exposed sites in the Canadian Shield, an extensive plateau dominated by

granite/gneiss outcroppings and nutrient-poor soils. Although both sites are characterized by continental climates, they are more than 3000 km distant with the Daring Lake

watershed samples from the northern region and the Gould Lake watershed samples from

the most southerly extent of the Shield. We had hypothesized that because Daring Lake

winters are significantly longer than Gould Lake winters, with mean midwinter

temperatures that are ~21°C colder and large temperature fluctuations particularly in the

spring (Nobrega & Grogan, 2007), a larger proportion of the enriched consortium would

have ice-association activities and show greater resistance to our selective regime. Since

one of the goals of this project was to prospect for microbes with AFPs and INPs for

103

Figure 3.4. Representative ice nucleation activity of freeze-thaw resistant isolates as indicated by the logarithm of ice nuclei per mL. Samples include; Pseudomonas sp. (DL13), Bacillus sp. (G1a1), P. syringae preparation (ice nucleation proteins; INPs), E. coli TG-2 and control 10% TSB. Assays were replicated (n=3-5).

104

subsequent protein isolation and characterization, a culture-dependent approach was

necessary.

Standard calculations with the clone libraries suggested that they represented

about half (63% and 43% from the Daring and Gould Lake collections, respectively) of the total cultured richness (Good, 1953; Table 3.3). RFLP analysis with subsequent sequencing showed that the enrichment cultures from both soils had half a dozen genera in common (Table 3.2). In response to freeze-thaw stress, the cell count of both consortia decreased in step until there was a 3-4 logarithm decline in viability after 48 freeze-thaw

cycles (Fig. 3.2), coupled with a 10 fold decrease in diversity (Table 3.2). After such

severe selection only the most hardy isolates survive (Walker et al., 2006) and in both

consortia freeze-thaw resistant bacteria were isolated. These 8 microbes represented 6

different genera (Table 3.2); C. piscium, Paenibacillus sp., and two Pseudomonas spp.

were isolated from the Daring Lake consortia, while Bacillus sp., Buttiauxella sp., P.

agglomerans (synonym Erwinia herbicola) and an Enterobacteriaceae were isolated

from the Gould Lake consortia. When treated as isolates, Chryseobacterium sp. DL11, P.

borealis Y1C, Bacillus sp. G1a1, Buttiauxella sp. G2b1, Enterobacteriaceae G3b1 and P.

agglomerans GLY were at least as freeze-thaw resistant as the consortia from which they

were recovered, while Pseudomonas sp. DL13 was slightly more susceptible.

Paenibacillus sp. Y1B, however, was highly susceptible as a monoculture and lost nearly

all viability. Thus, contrary to our hypothesis, although the Daring Lake collection sites

were a full 20° further north, resistant isolates were not more numerous than in the

southern collections.

105

Strikingly, all of the isolated genera, even those recovered from the south, have

previously been isolated from extreme low temperature environments such as glaciers

(Christner et al., 2000; Miteva et al., 2004), cryoconite sediment (Christner et al., 2003), accretion ice (Christner et al., 2001), and polar regions (Nelson & Parkinson, 1978;

Gilbert et al., 2004; Lee et al., 2004). Microbes identified from these environments presumably have adaptations allowing them to resist such harsh conditions, and similar adaptations may be responsible for the recovery of microbes after our selection.

The overlap of genera between natural environments and those recovered after freeze-thaw in this study reflects the stringency of the selective regime. Based on the phylogenetic tree (Fig. 3.1) and the RFLP data, recovered resistant isolates comprise a phylogenetically diverse group of microbes, apparently representing only a small portion of the original population. In order to ensure that the cryocycler did not simply select for the most abundant microbes, semi-quantitative PCR was used to estimate the proportion of the original cultured consortia that were represented by genera subsequently found to be included in the freeze-thaw resistant isolates. Since the Daring and Gould Lake enriched cultures were represented by the genera Bacillus (1.1% and 20.5%),

Paenibacillus (0.1% and 0.5%) and Pseudomonas (21.0% and 4.7%), respectively (Table

3.1), clearly the recovered specific species were present in the original cultures at a much lower proportion at both sites. Thus, our freeze-thaw regime allowed the rigorous selection of a few bacteria without apparent bias towards a particular phylogenetic grouping.

106

Given the overlap of genera isolated and the similarities amongst the selective pressures between natural low temperature environments and our selective regime, one would predict similar phenotypes. The genera of the bacteria recovered here, including

Bacillus (Nejad et al., 2004), Chryseobacterium (Walker et al., 2006), Pseudomonas (Xu et al., 1998) and Erwinia (Koda et al., 2000) have previously been associated with INP and/or AFP activity. In consequence, we characterized our resistant isolates to estimate the proportion having ice-association properties, since these activities presumably decrease or prevent ice-induced injury. In total, a remarkable 63% of the surviving isolates, originating from both sites and certainly not dominated by the higher latitude consortia, had these activities (Table 3.2; Fig. 3.3, 3.4). Pseudomonas sp. DL13 and

Bacillus sp. G1a1 demonstrated both type 1 INA and IRI activities (Fig. 3.1).

Buttiauxella sp. G2b1 had ice-shaping activity and C. piscium DL11, Buttiauxella sp.

G2b1 and Enterobacteraceae G3b1 all showed IRI. Of the surviving isolates, only P. borealis YIC, Paenibacillus sp. YIB and P. agglomerans GLY did not display any of these activities under the conditions used for assays. It is possible that these bacteria may have survived because of sporulation (Paenibacillus for instance is a known spore- former), cryoprotectants, cold shock proteins, or other adaptations. Nevertheless, the fact that the majority of our recovered isolates showed ice-association activity is even more striking given that none of the 16 isolates derived from the unselected Daring and Gould

Lake consortia, showed any such activity, indicating that these activities are not common at the collection sites.

107

Our results are in keeping with previous findings in which freeze-thaw selection of soil-derived consortia enabled the isolation of a few microbes, some with putative ice- association proteins (Walker et al., 2006). Although the original samples were obtained from different sites, with different environmental conditions (though the pH of each soil was 6.5, which may be a primary indicator of diversity; Chu et al., 2010), thus leading to differing enrichment culture diversity, similar trends were observed. The cryocycler decreased culturable consortia diversity throughout the multiple freeze-thaw cycles and enabled the recovery of freeze-thaw resistant microbes from both soil collections, including from the southern, more temperate climate location. Even ‘rare’ bacteria in a consortium can be isolated given sufficient selective pressure and these microbes have similar functional properties. We further speculate that seasonal temperature and freeze- thaw conditions select for microbes with ice-association activities so that they become

relatively abundant.

Acknowledgements:

This work was supported by NSERC (Canada) and an International Polar Year

(CiCAT) grant as well as a Queen’s Research Chair award to VKW and NSERC and

OGS (Canada) scholarships to SLW. The authors would like to thank Dr. K. Buckeridge

and the other students in our Molecular Microbiology class for their enthusiasm and

encouragement. Drs. P. Davies (Queen’s University), S. Lougheed (Queen’s University),

and G. Voordouw (University of Calgary) are thanked for the use of the nanolitre

108

osmometre, assistance with the phylogenetic tree, and for the E. coli TG-2, respectively.

T. Vanderveer is thanked for her assistance with the nanolitre osmometer and Dr. G.

Palmer for his technical support. Floyd Connor, Queen’s University Biological Station, and Bob Reid, INAC, Water Resources Division are thanked for the temperature data.

109

Literature Cited:

Baas-Becking, L.G.M. (1934). Geobiologie of inleiding tot de milieukunde. The Hague: Van Stockum, W.P., and Zoon.

Calcott, P.H., and MacLeod, R.A. (1975a). The survival of Escherichia coli from freeze- thaw damage: permeability barrier damage and viability. Can. J. Microb. 21: 1724-1732.

Calcott, P.H., and MacLeod, R.A. (1975b). The survival of Escherichia coli from freeze- thaw damage: the relative importance of wall and membrane damage. Can. J. Microb. 21: 1960-1968.

Chakrabartty, A., and Hew, C.L. (1991). The effect of enhanced α-helicity on the activity of a winter flounder antifreeze polypeptide. Eur. J. Biochem. 202: 1057-1063.

Chattopadhyay, M.K., and Jagannadham, M.V. (2001). Maintenance of membrane fluidity in Antarctic bacteria. Polar Biol. 24: 386-388.

Chiang, Y.-C., Yang, C.-Y., Li, C., Ho, Y.-C., Lin, C.-K., and Tsen, H.-Y. (2006). Identification of Bacillus spp., Escherichia coli, Salmonella spp., Staphylococcus spp., and Vibrio spp. with 16S ribosomal DNA-based oligonucleotide array hybridization. Int. J. Food Microb. 107: 131-137.

Christner, B.C., Kvitko, B.H.II., and Reeve, J.N. (2003). Molecular identification of bacteria and eukarya inhabiting an Antarctic cryoconite hole. Extremophiles 7: 177-183.

Christner, B.C., Mosley-Thompson, E., Thompson, L.G., and Reeve, J.N. (2001). Isolation of bacteria and 16S rDNAs from Lake Vostok accretion ice. Environ. Microb. 3: 570-577.

Christner, B.C., Mosley-Thompson, E., Thompson, L.G., Zagorodnov, V., Sandman, K., and Reeve, J.N. (2000). Recovery and identification of viable bacteria immured in glacial ice. Icarus 144: 479-485.

110

Chu, H., Fierer, N., Lauber, C.L., Caporaso, J.G., Knight, R., and Grogan, P. (2010). Soil bacterial diversity in the Arctic is not fundamentally different from that found in other biomes. Environ. Microb. doi:10.1111/j.1462-2920.2010.02277.x

Cole, J.R., Wang, Q., Cardenas, E., Fish, J., Chai, B., Farris, R.J., Kulam-Syed- Mohideen, A.S., McGarrell, D.M., Marsh, T., Garrity, G.M., and Tiedje, J.M. (2009). The ribosomal database project: improved alignments and new tools for rRNA analysis. Nucleic Acids Res. 37: D141-D145. de Wit, R., and Bouvier, T. (2006). ‘Everything is everywhere, but, the environment selects’; what did Baas Becking and Beijerinck really say? Environ. Microb. 8: 755-758.

Deninger, C.A., Mueller, G.M., and Wolber, P.K. (1988). Immunological characterization of ice nucleation proteins from Pseudomonas syringae, Pseudomonas fluorescens, and Erwinia herbicola. J. Bacteriol. 170: 669-675.

Gilbert, J.A., Hill, P.J., Dodd, C.E.R., and Laybourn-Parry, J. (2004). Demonstration of antifreeze protein activity in Antarctic lake bacteria. Microbiology 150: 171-180.

Gomez Zavaglia, A., Disalvo, E.A., and De Antoni, G.L. (2000). Fatty acid composition and freeze-thaw resistance in lactobacilli. J. Dairy Res. 67: 241-247.

Good, I.J. (1953). The population frequencies of species and the estimation of population parameters. Biometrika 40: 237-264.

Gordienko, R., Ohno, H., Singh, V.K., Jia, Z., Ripmeester, J.A., and Walker, V.K. (2010). Towards a green hydrate inhibitor: imaging antifreeze proteins on clathrates. PLoS ONE 5: e8953.

Henry, H.A.L. (2008). Climate change and soil freezing dynamics: historical trends and projected changes. Clim. Change 87: 421-434.

Izawa, S., Sato, M., Yokoigawa, K., and Inoue, Y. (2004). Intracellular glycerol influences resistance to freeze stress in Saccharomyces cerevisiae: analysis of a

111 quadruple mutant in glycerol dehydrogenase genes and glycerol-enriched cells. Appl. Microb. Biotech. 66: 108-114.

Jones, P.G., VanBogelen, R.A., and Neidhardt, F.C. (1987). Induction of proteins in response to low temperature in Escherichia coli. J. Bacteriol. 169: 2092-2095.

Joseph, G., and Henry, H.A.L. (2008). Soil nitrogen leaching losses in response to freeze- thaw cycles and pulsed warming in a temperate old field. Soil Biol. Biochem. 40: 1947- 1953.

Knight, C.A., and Duman, J.G. (1986). Inhibition of recrystallization of ice by insect thermal hysteresis proteins: a possible cryoprotective role. Cryobiology 23: 256-262.

Koda, N., Aoki, M., Kawahara, H.K., Yamade, K., and Obata, H. (2000). Characterization and properties of intracellular proteins after cold acclimation of the ice- nucleating bacterium Pantoea agglomerans (Erwnia herbicola). Cryobiology 41: 195- 203.

Kun, H., Minnes, R., and Mastai, Y. (2008). Effects antifreeze peptides on the thermotropic properties of a model membrane. J. Bioenerg. Biomem. 40: 389-396.

Lane, D.J. (1991). 16S/23S rRNA sequencing. In Nucleic acid techniques in bacterial systematics. Stackebrandt, E., and Goodfellow, M., ed., John Wiley and Sons, New York, NY, pp. 115-175.

Lee, Y.K., Kim, H.W., Cho, K.H., Kang, S.-H., Lee, H.K., and Kim, Y. (2004). Phylogenetic analysis of culturable arctic bacteria. Ocean Polar Res. 26: 51-58.

Maki, L.R., Galyan, E.L., Chang-Chien, M.-M., and Caldwell, R.D. (1974). Ice nucleation induced by Pseudomonas syringae. Appl. Microbiol. 28: 456-459.

112

Mazur, P. (1963). Kinetics of water loss from cells at subzero temperatures and the likelihood of intracellular freezing. J. Gen. Physiol. 47: 347-369.

Mazur, P. (1966). Physical and chemical basis of injury in single-celled microorganisms subjected to freezing and thawing. In Cryobiology. Merman, H.T., ed., Academic Press, New York, pp. 214-315.

Merrill, L., Richardson, J., Kuske, C.R., and Dunbar, J. (2003). Fluorescent heteroduplex assay for monitoring Bacillus anthracis and close relatives in environmental samples. Appl. Environ. Microb. 69: 3317-3326.

Nadkarni, M.A., Martin, F.E., Jacques, N.A., and Hunter, N. (2002). Determination of bacterial load by real-time PCR using a broad-range (universal) probe and primers set. Microbiology 148: 257-266.

Nejad, P., Ramstedt, M., and Granhall, U. (2004). Pathogenic ice-nucleation active bacteria in willows for short rotation forestry. For. Pathol. 34: 369-381.

Nelson, L.M., and Parkinson, D. (1978). Effect of freezing and thawing on survival of three bacterial isolates from an arctic soil. Can. J. Microb. 24: 1468-1474.

Nobrega, S., and Grogan, P. (2007). Deeper snow enhances winter respiration from both plant-associated and bulk soil carbon pools in birch hummock tundra. Ecosystems 10: 419-431.

Raymond, J.A., and DeVries, A.L. (1977). Adsorption inhibition as a mechanism of freezing resistance in polar fishes. Proc. Natl. Acad. Sci. USA 74: 2589-2593.

Sawicka, J.E., Robador, A., Hubert, C., Jørgensen, B.B., and Brüchert, V. (2010). Effects of freeze-thaw cycles on anaerobic microbial processes in an Arctic intertidal mud flat. ISME J. 4: 585-594.

Schimel, J., Balser, T.C., and Wallenstein, M. (2007). Microbial stress-response physiology and its implications for ecosystem function. Ecology 88: 1386-1394.

113

Shannon, C.E., and Weaver, W. 1949. The Mathematical Theory of Communication. University of Illinois Press, Urbana, IL, pp. 117.

Shelburne, C.E., Prabhu, A., Gleason, R.M., Mullally, B.H., and Coulter, W.A. (2000). Quantitation of Bacteroides forsythus in subgingival plaque comparison of immunoassay and quantitative polymerase chain reaction. J. Microbiol. Methods 39: 97-107.

Shida, O., Takagi, H., Kadowaki, K., Nakamura, L.K., and Komagata, K. (1997). Transfer of Bacillus alginolyticus, Bacillus chondroitinus, Bacillus curdlanolyticus, Bacillus glucanolyticus, Bacillus kobensis, and Bacillus thiaminolyticus to the genus Paenibacillus and emended description of the genus Paenibacillus. Int. J. Syst. Bacteriol. 47: 289-298.

Spilker, T., Coenye, T., Vandamme, P., and Li Puma, J.J. (2004). PCR-based assay for differentiation of Pseudomonas aeruginosa from other Pseudomonas species recovered from cystic fibrosis patients. J. Clin. Microb. 42: 2074-2079.

Swofford, D.L. (2002). PAUP*. Phylogenetic analysis using parsimony (*and other methods). Version 4. Sunderland, MA.

Vali, G. (1971). Quantitative evaluation of experimental results on the heterogeneous freezing nucleation of supercooled liquids. J. Atmos. Sci. 28: 402-409.

Walker, V.K., Palmer, G.R., and Voordouw, G. (2006). Freeze-thaw tolerance and clues to the winter survival of a soil community. Appl. Environ. Microb. 72: 1784-1792. 114

Wilson, S.L., Kelley, D.L., and Walker, V.K. (2006). Ice-active characteristics of soil bacteria selected by ice-affinity. Environ. Microb. 8: 1816-1824.

Yankofsky, S.A., Levin, Z., Bertold, T., and Sandlerman, N. (1981). Some basic characteristics of bacterial freezing nuclei. J Appl. Meteor. 20: 1013-1019.

Yergeau, E., and Kowalchuk, G.A. (2008). Responses of Antarctic soil microbial communities and associated functions to temperature and freeze-thaw cycle frequency. Environ. Microb. 10: 2223-2235.

Zachariassen, K.E., and Hammel, H.T. (1976). Nucleating agents in the haemolymph of insects tolerant to freezing. Nature 262: 285-287.

115

Chapter 4: Community Level Cross-Tolerance between Osmotic and Freeze-Thaw Stresses in Temperate Lakes

Abstract:

Hypersaline and low temperature environments are similar in that both can result in osmotic stress due either to constant high extracellular salt concentrations or the

temperature-dependent increase in solute levels after freezing. In order to determine if microbes from environments with high osmotic potential would be preadapted to freezing stress, enriched cultures from water and sediments collected from temperate freshwater,

brine and alkali lakes, were subjected to repetitive freeze-thaw cycles. Microbial

abundance after selection paralleled lake salinity in that enrichments from alkali lakes

were typically most resistant, and those from freshwater lakes were most susceptible.

Microbial consortia derived from the lake water samples showed equilavent or even

higher levels of resistance than their sediment-derived counterparts. Freeze-thaw

resistant isolates included known psychro-, halo- and alkali-tolerant bacteria. Although few of the isolates demonstrated ice-association activities, a higher proportion of bacteria from brine and alkali lakes had higher intracellular levels of osmolytes and appeared more likely to form biofilms, than the freshwater derived isolates, consistent with our hypothesis of cross-tolerance between osmotic and freeze-thaw stress.

116

Introduction:

Frozen environments present a number of challenges for microbes, not the least

being the physical damage associated with ice formation. Certain species produce

cryoprotectants such as sugars or specialized cold shock and antifreeze and/or ice

nucleation proteins (Lindow et al., 1978; Jones et al., 1987; Xu et al., 1998; Panoff et al.,

2000). As a consequence, in the presence of some of these molecules, the shape and size of ice crystals can be maipulated, or the temperature at which freezing initiates is altered.

Ice formation is associated with osmotic stress since solutes and cells become

concentrated in brine pockets (Mader et al., 2006; Amato et al., 2009). Survival under

hyperosmotic or low temperature conditions is accompanied by changes in gene

expression (Li et al., 2006), membrane compostition (Allakhverdiev et al., 1999;

Sakamoto & Murata, 2002), and the accumulation of osmoprotectants such as compatible

solutes (eg. amino acids and sugars) or inorganic salts, such as potassium chloride (eg.

Yancey et al., 1982; Shahjee et al., 2002).

Biofilms also help confer stress resistance. Indeed, microbes within biofilms are

widely considered to be much more resistant to a multitude of agents, up to three orders

of magnitude more than their planktonic counterparts (McAuliffe et al., 2008). For example, hyperosmotic stress has been shown to upregulate Staphylococcus and Bacillus

biofilm formation, and a Rhizobium defective in its ability to form biofilms was more

susceptible to osmotic and desiccation stresses (Knobloch et al., 2001; Vanderlinde et al.,

117

2009). Biofilms also appear to protect microbes from cold and freeze stress (Williams et

al., 2009; Wu et al., unpublished).

Similar responses to different environmental stresses has been designated as microbial cross-tolerance. For instance, salt pretreatment of bacterial strains resulted in

the production of cold or heat shock proteins which were correlated with low temperature

survival (Schmidt & Zink, 2000; Le Blanc et al., 2003). Schmid et al., (2009) further

showed that either low temperature or salt exposure could induce cold shock protein

genes that were required for optimal survival under low temperature and sodium chloride

osmotic stress. In our experiments, the hypothesis that microbial consortia from hyperosmotic environments would be more resistant, in general, to freeze-thaw stress was

investigated. This was achieved by comparing the relative survival amongst microbial communities derived from temperate lakes (water and sediment) of different salinities.

Materials and Methods:

Sample Sites and Culture Conditions: Samples were obtained from six temperate lakes

within the British Columbia interior (Canada; GPS coordinates in Table 4.1), in late

summer and stored at 4°C. Two freshwater (Fly and Leeches Lakes), two brine (Lake

East of 83 and Liberty Lake) and two alkali (TR Ranch and Spotted Lakes) lakes were

sampled. Water and sediment samples (top 2-3 cm) were obtained from all lakes, with

the exception of TR Ranch Lake, from which no water sample was obtained due to

seasonal drying. The ion content of the five lakes was determined (Table 4.1; Analytical

Services Unit, Queen’s University, Kingston, ON), and growth media was chosen or

formulated as appropriate. Earlier sampling of the sixth lake was used analogously. The 118

Table 4.1. Lake Sample Location and Lake Water Chemistry.

Lake: Type: GPS Major Ions1: pH: Lake Water Culture Media Location: Salinity (mOsm2): (mOsm2): Fly Lake Freshwater 51°54'N, None 8.9 0 10% TSB 121°19'W (36) Leeches Lake Freshwater 52°2'N, None 8.2 0 10% TSB 122° 19'W (36) Lake East of 83 Brine 51°28'N, Na+ 10 885 E83 Medium 121°23'W (170.5) Liberty Lake Brine 51°17'N, Na+, Cl- 10 2074 Liberty Medium 121°43'W (1293.5) 3 2+ 2- TR Ranch Lake Alkali 51°34'N, Mg , SO4 8.2 NA 50% Marine Broth 121°39'W (478.5) Spotted Lake Alkali 49°5'N, Na+, K+, Mg2+, 7 6016 50% Marine Broth - 2- 2- 119°34'W Cl , SO4 , S (478.5) 1 Greater than 10 mM. 2 As measured with a vapour pressure osmometer. 3 Generously provided by Dr. Cumming.

119

freshwater and alkali lakes were cultured in 10% TSB (3 g tryptic soy broth (Bacto,

Dickinson and Company, Sparks, MD, USA), 0.1 g KNO3, 0.1 g (NH4)2SO4 and 0.1 g

K2HPO4 per liter of deionized water; pH 7) or 50% Marine Broth (pH 7; Difco, Becton,

Dickinson and Company, Sparks, MD, USA), respectively. Media suitable for the brine

lakes were formulated: Lake East of 83 medium (0.01g CaCl2·2H2O, 0.34g

K2HPO4·3H2O, 0.03g MgSO4, 1.15g NaCl, 2.15g Na2CO3, 0.78g NaHCO3, per liter of

deionized water; pH 10) and Liberty Lake medium (0.01g CaCl2·2H2O, 1.36g

K2HPO4·3H2O, 0.05g MgSO4, 3.90g NaCl, 44.17g Na2CO3, 16.06g NaHCO3, per liter of

deionized water; pH 10). In the case of one isolate, PYA (peptone, yeast extract, alkaline

buffer) broth was used (Yumoto et al., 2004; pH 10). In all cases, 1.5% agar was added

for semi-solid media.

Consortia derived from the water samples were obtained by inoculating the

appropriate medium (6 mL) with the water samples (1 mL) and culturing for 72 h at

22°C, followed by subculturing (1 mL culture in 6 mL of fresh medium). Sediments were similarly enriched by inoculating the respective medium (10 mL) with the sediment

sample (0.5g) and culturing for 72 h at 22°C, followed by subculturing (1 mL culture in 6

mL of fresh medium). Bacterial strains used as controls (E. coli TG-2 and

Chryseobacterium sp. C14), were cultured in 10% TSB. All subcultures and controls

were incubated for 48 h at 22°C, and then transferred to 4°C overnight prior to freeze- thaw treatments. After treatments, recovered individual isolates were cultured in the

appropriate media prior to sequencing. Those characterized (see ‘isolate

characterization’ below) were similarly cultured, but after growth at 22°C for at least 48 h

120

they were moved to 4°C for 48 h or 72 h prior to assay. All cultures were shaken at

~100rpm while at 22°C, however those at 4°C were not shaken.

Freeze-Thaw Selection: Aliquots (2 mL) of the enriched samples, in triplicate, were subjected to 48 consecutive freeze-thaw cycles (+5°C to -18°C). Selection used a cryocycler (Walker et al., 2006). In order to determine CFUs, the water and sediment samples were diluted with diluted lake water, as appropriate. Diluted lake water was obtained by using 50% sterilized lake water from each of the lakes and 50% sterile distilled water, for all but Spotted Lake, which was diluted 1:4. The TR Ranch-derived cultures were prepared with diluted Spotted Lake water. To compare freeze-thaw

resistance in the different enrichments, each sample was normalized to a starting density

of 1 x 108 CFU/mL. A one-way analysis of variance (ANOVA; α=0.05), followed by

Tukey-Kramer honestly (least) significant difference (HSD;

http://udel.edu/~mcdonald/anova.xls) tests were used to determine the statistical

relationships of survival between the lake samples.

Isolate Recovery and Identification: Morphologically-distinct colonies were isolated as

monocultures, a subset of which was putatively identified on the basis of 16S rRNA gene

sequence analysis. These isolates were PCR amplified (from broth culture or lysates)

with the universal primers 8F and r1406 (Lane et al., 1985; Hicks et al., 1992). Standard reaction conditions were used, and cycling conditions were according to a manufacturer’s protocol (Novagen T7 Select phage display protocol; without hot start, Madison, WI,

121

USA). The PCR products (~1.4 kb) were visualized on 1% agarose gels, purified

(Qiagen, Mississauga, ON, Canada), sequenced in both directions (Plateforme de

Génomique at the Centre de recherché du CHUL, Quebec, QC, Canada) with primers 8f

and r1406. Sequences were analyzed using ‘Manipulate and Display a DNA Sequence’

from Molecular Toolkit (http://arbl.cvmbs.colostate.edu/molkit/manip) and CodonCode

Aligner (http://www.codoncode .com/aligner/trial.htm). The putative identities of the isolates were determined based on the nearest phylogenetic relative in the BLASTn

(NCBI; http://www.ncbi.nlm.nih.gov/blast; Altschul et al., 1997) or RDP II (Ribosomal

Database Project II; http://rdp.cme.msu.edu (Cole et al., 2005, Cole et al., 2009) databases. PAUP* (Swofford, 2002) was used to construct a phylogenetic tree

(neighbour joining, kimura 2) of the isolates.

Isolate Characterization: A subset of the isolates, which had been sequenced, were characterized with respect to their ice recrystallization inhibition (IRI) and ice nucleation activity (INA) phenotypes, their osmolyte content, as well as their ability to form biofilms.

IRI assays were as previously described (Tomczak et al., 2003; Wilson & Walker,

2010). Briefly, after snap freezing (~ -35°C), capillaries loaded with whole cell cultures were annealed at -6°C overnight, with images obtained prior to and following incubation used to assess ice crystal growth. INA was estimated by cooling whole cell droplets (~

+2°C to -12°C) and recording the sample nucleation temperature (9 of the 10 droplets frozen; modified from Vali, 1971; Maki et al., 1974) as described in Wilson & Walker

122

(2010). Prepared, recombinant antifreeze proteins (AFPs; Gordienko et al., 2010) and an

ice nucleation protein preparation (Wards Natural Science Establishment, Rochester, NY,

USA) were used as the positive controls for the IRI and INA assays, respectively. E. coli

TG-2 and culture media were used as negative controls for both assays.

A vapour pressure osmometer (model 3MOplus, Advanced Instruments Inc.,

Norwood, MA, USA) was used to estimate the intracellular osmolyte content of the

isolates. As indicated, isolates were cultured for 48 h at 22°C, followed by 72 h at 4°C.

Cells were harvested via centrifugation (15 min at 10,060 xg). Pellets were resuspended

in 500 μL dH2O and the optical density at 600 nm was used for normalization. Cells

were subsequently disrupted by sonification (Belgrader et al., 1999), and the osmolyte content, in milliosmoles (mOsm), was determined in duplicate. E. coli TG-2 and the culture media were used as negative controls. One-way ANOVAs (α=0.05) followed by

Tukey-Kramer HSD tests were used to distinguish between samples.

Biofilm assays were done as described by O’Toole & Kolter (1998) and

Balestrino (2008). Briefly, isolates were cultured for 48 or 72 h at 22°C in 96-well plates

(4 μL starter culture, 100 μL appropriate media), in duplicate. After overnight incubation

at 4°C, samples were stained with gentian violet (50 μL; 0.5% w/v) for 15 min. The

wells were subsequently washed 5 or more times with water, and the dye was released

upon addition of 95% ethanol (200 μL). Following transfer of the alcohol extract to new

96-well plates, the OD570 was determined after subtraction of values of the media alone.

E. coli TG-2 and the culture media were used as negative controls. T-tests (α=0.05) were

used to assess the significance of the differences between the isolates and E. coli TG-2.

123

Results:

Water Chemistry Analysis: Chemical analysis (Table 4.1) indicated that the lakes were distinct with Fly and Leeches Lakes classified as freshwater, with no major salt content and with modest differences in H+ ion concentration (pH 8.2-8.9). Those classified as

brine lakes were high in Na+ or Na+ and Cl- ions with the alkali lakes high in Mg2+ and

2- + + - 2- SO4 and for one lake Na , K , Cl , S . Overall salinity of the freshwater lakes was low

(0 mOsm), while the brine lakes were more saline at 885 mOsm and 2,074 mOsm for

Lake East of 83 and Liberty Lake, respectively. Spotted Lake was highly saline, at 6016 mOsm.

Freeze-Thaw Selection: Freeze-thaw cycling was used to eliminate freeze-susceptible

microbes and to select for resistant members of the original consortia. The reduction in

consortia abundance post selection appeared to be related to the salinity of the “home” lake. The consortia derived from the freshwater lakes (Fly and Leeches Lake) lost approximately 4 logarithms of viability in the water-derived enrichments, and 5-6 logarithms for the sediment-derived samples (Fig. 4.1). The microbial abundance of the brine lakes (Lake East of 83 and Liberty Lake) decreased by approximately 4 and 2 logarithms, respectively, for the water-derived samples, and approximately 4 logarithms for the sediment-derived consortia. Finally, enrichments from the alkali lakes, decreased

124

F B A F B A

b b c b

a a,d a a a a a

Figure 4.1. Average colony forming units (CFU) per mL for the water and sediment- derived consortia after 48 freeze-thaw cycles. These are compared to single isolate control cultures, Chryseobacterium sp. C14 and E. coli TG-2. ‘All’ indicates the normalized starting density of 1x108 CFU/mL for all cultures. Significant differences (α=0.05) amongst each sample type are indicated by different letters (a-d) over the bars and fresh (F), brine (B) and alkali (A) samples are indicated along the top axis.

125

by about 2 logarithms for both the water sample (Spotted Lake) and the sediment-derived

consortia (Spotted and TR Ranch Lake).

A one-way ANOVA (α=0.05) followed by a Tukey-Kramer HSD test showed that there

were statistically significant differences amongst the lake consortia with respect to their

resistance to freeze-thaw stress. The water and sediment samples from Fly and Leeches

Lakes, and Lake East of 83 were statistically similarly resistant, and were all less resistant

(p<0.05) than those samples derived from Liberty Lake water and Spotted Lake water and sediments (Fig. 4.1). The culture from Liberty Lake water was similarly resistant to

TR Ranch and Spotted Lake sediments, and more resistant (p<0.05) than all the others.

Sediment cultures from Liberty Lake were, however, were less resistant (p<0.05) than the

Spotted Lake water and sediment cultures, and similarly resistant to the others. Finally,

TR Ranch and Spotted Lake sediment-derived cultures were similarly resistant, and were

slightly less resistant (p<0.05) than the Spotted Lake water consortia. Such high viability

after 48 freeze-thaw cycles was only surpassed by Chryseobacterium sp. C14, a known

freeze-thaw resistant strain (Fig. 4.1), (Walker et al., 2006; Wilson & Walker 2010).

Isolate Recovery and Identification: Morphologically-distinct microbes, with redundancy

and recovered post freeze-thaw selection, were successfully isolated as monocultures.

The 16S rDNA of a subset of these isolates was sequenced (Table 4.2; Fig. 4.2) to

estimate consortia richness following selection. Notwithstanding the orders of magnitude

reduction in overall cell viability, known halophiles were recovered from the brine and

alkali lakes.

126

Table 4.2. Lake consortia richness following freeze-thaw.

Sample Nearest Phylogenetic Accession N (% Putative Identity and Sample Relative1 Number2 Similarity)3 ID4 Freshwater: Fly Lake: Bacillus sp. HM216203.1 1081 (100) Bacillus sp. FW3 water Pseudomonas sp. FN547413.1 1142 (100) Pseudomonas sp. FW3a Fly Lake: Acinetobacter sp. FJ405316.1 1131 (99) Acinetobacter sp. FNS8 sediment Arthrobacter gandavensis AM237357.1 1095 (99) A. gandavensis FNS27 Bacillales NA 1168 (NA) Bacillales FNS42 Bacillus sp. FJ908092.1 1067 (99) Bacillus sp. FNS14 Bacillus sp. AY660700.1 1249 (99) Bacillus sp. FNS44 Cryobacterium sp. GU733466.1 1044 (99) Cryobacterium sp. FNS46 Herbaspirillum sp. FN555399.1 1174 (99) Herbaspirillum sp. FNS6 Leifsonia sp. GU213301.1 1144 (99) Leifsonia sp. FNS18b Microbacteriaceae NA 1147 (NA) Microbacteriaceae FNS3 Paenibacillus sp. AB046422.1 1209 (100) Paenibacillus sp. FNS13 Paenibacillus sp. GQ915094.1 1170 (99) Paenibacillus sp. FNS28 Planococcaceae NA 1198 (NA) Planococcaceae FNS29 Rhodococcus erythropolis AP008957.1 1008 (98) R. erythropolis FNS16 Rhodococcus sp. EU768823.1 948 (99) Rhodococcus sp. FNS48 Sporosarcina aquimarina NR_025049.1 1004 (99) S. aquimarina FNS43 Trichococcus sp. AM933652.1 979 (99) Trichococcus sp. FNS25

Leeches Pseudomonas sp. EF111108.1 1157 (100) Pseudomonas sp. LW1 Lake: Sphingomonas sp. AM900788.1 1196 (99) Sphingomonas sp. LW6 water Leeches Bacillaceae NA 1247 (NA) Bacillaceae LNS2 Lake: Bacillaceae NA 1185 (NA) Bacillaceae LNS62

127

sediment Bacillaceae NA 1132 (NA) Bacillaceae LNS63 Bacillus sp. D84614.2 1227 (100) Bacillus sp. LNS4 Micrococcaceae NA 1052 (NA) Micrococcaceae LNS56 Paenibacillus sp. EU332823.1 1161 (99) Paenibacillus sp. LNS58 Pseudomonadaceae NA 631 (NA) Pseudomonadaceae LNS44 Pseudomonas sp. FJ715742.1 1126 (100) Pseudomonas sp. LNS11 Trichococcus sp. AM933652.1 1179 (100) Trichococcus sp. LNS35

Brine: Lake East Alteromonadales EF554911.1 1239 (98) Alteromonadales EW4 of 83: Bacillus sp. FJ607059.1 1264 (99) Bacillus sp. EW5 water Bacillus saliphilus NR_025554.1 1133 (99) B. saliphilus EW1 Bacillus selenitireducens CP001791.1 1081 (99) B. selenitireducens EW9 Lake East Bacillus sp. AB055097.1 1182 (99) Bacillus sp. E83 11 of 83: Bacillus sp. FJ373038.1 1213 (98) Bacillus sp. E83 14 sediment Bacillus sp. FJ764772.1 1254 (100) Bacillus sp. E83 15 Halomonas variabilis AY204638.1 1228 (99) H. variabilis E83 6 Micrococcaceae NA 1145 (NA) Micrococcaceae E83 IV Nesterenkonia halotolerans NR_029073.1 1158 (100) N. halotolerans E83 III Pseudomonas sp. GQ417894.1 1016 (99) Pseudomonas sp. E83 3

Liberty Bacillus sp. FJ764772.1 1162 (99) Bacillus sp. LLW3 Lake: Halomonas sp. FJ950737.1 1152 (98) Halomonas sp. LLW2 water Halomonas sp. GU228483.1 1203 (99) Halomonas sp. LLW14 Halomonas sp. EF554888.1 1235 (99) Halomonas sp. LLW36 Halomonas sp. EU432575.1 1125 (99) Halomonas sp. LLW42 Idiomarina sp. EF554920.1 1170 (99) Idiomarina sp. LLW27 Nesterenkonia sp. EU432564.1 1057 (99) Nesterenkonia sp. LLW16 Roseinatronobacter monicus DQ659236.1 1069 (100) R. monicus LLW20

128

Liberty Halomonas sp. GU228481.1 591 (99) Halomonas sp. LLNS4 Lake: Idiomarina sp. EF554872.1 1118 (99) Idiomarina sp. LLNS12 sediment

Alkali: TR Ranch Bacillaceae NA 1028 (NA) Bacillaceae RNS26 Lake: Bacillus sp. EU004572.1 1223 (99) Bacillus sp. RNS24 sediment Gillisia sp. EU196339.1 966 (98) Gillisia sp. RNS1 Marinobacter sp. AB167042.1 1141 (98) Marinobacter sp. RNS5 Pseudomonas sp. EU365515.1 1067 (98) Pseudomonas sp. RNS10

Spotted Halomonas sp. GU447290.1 1261 (99) Halomonas sp. SW10 Lake: Halomonas sp. AB085656.1 1207 (99) Halomonas sp. SW13 water Lysinibacillus fusiformis HM068891.1 1187 (100) L. fusiformis SW22 Nesterenkonia halotolerans NR_029073.1 1158 (99) N. halotolerans SW17 Salegentibacter sp. GQ452870.1 1182 (98) Salengentibacter sp. SW28 Spotted Bacillus sp. HM045841.1 1052 (100) Bacillus sp. SNS3 Lake: Halomonas sp. EU308353.1 1100 (97) Halomonas sp. SNS4 sediment 1 Each ‘redundant’ isolate per genus is <98% similar, therefore, each isolate reported above is ‘unique’. 2 Accession number corresponding to the closest match in the database. NA is present for those defined at the family level. 3 The number (N) of bases sequenced and the overall % similarity to the corresponding accession number. NA is present for the % sililarity of the isolates identified at the family level. 4 Letters and numbers following the genus and species (in some cases) refer to the home lake designation.

129

70 Bacillaceae 100 L. fusiformis 89 Bacillaceae

99 Bacillaceae 100 Planococcaceae 63 74 S. aquimarina

Bacillus sp.

Bacillaceae Trichococcus sp. FNS25 100 Trichococcus sp. LNS35 55 99 Bacillus sp. 91 Bacillus sp.

100 Bacillus sp.

Bacillus sp. 95 Bacillus sp. 91 100 100 Bacillus sp. E83

Bacillus sp. E83

Bacillus sp. E83 99 Bacillus sp. 100 95 100 Bacillus sp.

100 B. saliphilus

B. selenitireducens EW9 100 Bacillales Paenibacillus sp. 100 Paenibacillus sp.

Paenibacillus sp. 96 96 67 Micrococcaceae 75 Micrococcaceae E83

A. gandavensis 100 100 N. halotolerans E83 100 N. halotolerans

97 Nesterenkonia sp. 77 Cryobacterium sp. 100 100 100 Microbacteriaceae

Leifsonia sp.

90 100 R. erythropolis Rhodococcus sp.

Gillisia sp. 100 88 Salengentibacter sp. SW28

100 Sphingomonas sp. LW6 R. monicus LLW20 Herbaspirillum sp. 60 Halomonas sp.

92 Halomonas sp. 56 H. variabilis E83

89 Halomonas sp. 92 Halomonas sp. Halomonas sp. 100 100 Halomonas sp. LLW14

72 84 Halomonas sp. Halomonas sp.

59 55 Marinobacter sp. 64 Alteromonadales

Acinetobacter sp.

80 66 Idiomarina sp. Idiomarina sp. LLNS12

83 Pseudomonadaceae 69 Pseudomonas sp. RNS10 Pseudomonas sp. E83 3 Pseudomonas sp. Pseudomonas sp. FW3a

Pseudomonas sp. LW1 0.1

Figure 4.2. An unrooted phylogenetic tree representing the relatedness of the lake water and sediment freeze-thaw resistant isolates, constructed using the PAUP software package. The lake sample from which the listed microbes were recovered is indicated in Table 4.2. Isolates from freshwater, brine and alkali lakes are in blue, green and red font respectively. The characterized isolates are underlined.

130

Isolate Characterization Assays: A subset of the microbes recovered following freeze- thaw selection were assayed for known low temperature resistant phenotypes (Table 4.3), including biofilm formation, osmotic content and ice-association characteristics. For some of these assays, the culture media was modified as indicated in Table 4.3 to either improve microbial growth or to facilitate the ice-association assays. Microbes were occasionally cultured and assayed in more than one media.

None of the recovered isolates demonstrated IRI activity. Only two species

(Altermonadales EW4 and R. monicus LLW20) exhibited INA activity, and this activity was only demonstrated when the bacteria were cultured in East of 83 and Liberty media

(respectively) and subsequently dialyzed (overnight at 4°C) against 50% Marine broth.

For both isolates INA activity was lower than that observed in a commercial preparation

(~ -6°C for Alteromonadales EW4 and ~ -8°C for R. monicus vs. -2°C, respectively).

The intracellular osmolyte content was determined using a vapour pressure osmometer and with E. coli TG-2 as the baseline. When washed cells were sonicated in distilled water, the osmolarity of 13 of 21 isolates were statistically higher than E. coli

(Table 4.3). Only 20% (2/9) of the freeze-thaw resistant bacteria derived from freshwater lakes had high intracellular osmolarity compared to 100% (6/6) and 83% (5/6) of those obtained by selection of brine and alkali lakes, respectively. Isolates from the more saline lakes had a statistically higher intracellular osmotic content than those from the other lakes.

The ability of the isolates to form biofilms was assessed by gentian violet staining, and values were normalized against the respective culture media and compared

131

Table 4.3. Lake isolate characterization assays. Isolated From Isolate IRI1 INA2 Biofilm Osmolyte Culture (Y/N) (Y/N) Formation3 Content4 Media5 (Y/N) (mOsm) Fly Lake water: Bacillus sp. FW3 N N N 1.06 10% TSB Pseudomonas sp. FW3a N N N 1.24 10% TSB Fly Lake sediment: Cryobacterium sp. FNS46 N N N 1.87 10% TSB Leifsonia sp. FNS18b N N N 1.35 10% TSB Paenibacillus sp. FNS28 N N N 0 10% TSB S. aquimarina FNS43 N N N 7.54* 10% TSB

Leeches Lake water: Sphingomonas sp. LW6 N N N 1.91 10% TSB Leeches Lake sediment: Bacillaceae LNS62 N N N 12.66* 50% MB Trichococcus sp. LNS35 N N N 0 10% TSB

Lake East of 83 water: Alteromonadales EW4 N N Y 13.29* E83 Alteromonadales EW4 N Y Y 13.49* 50% MB Lake East of 83 sediment: H. variabilis E83 6 NA N Y 19.8* E83 H. variabilis E83 6 N N Y 6.71* 50% MB

Liberty Lake water: Halomonas sp. LLW2 NA N Y 52.68* LM Halomonas sp. LLW2 N N N 26.89* 50% MB Halomonas sp. LLW42 NA N Y 46.61* LM Halomonas sp. LLW42 N N N 137.2* 50% MB R. monicus LLW20 NA Y Y 50.05* LM R. monicus LLW20 N N Y 29.27* PYA Liberty Lake sediment: Idiomarina sp. LLNS12 NA N N 32.27* LM Idiomarina sp. LLNS12 N N N 11.67* 50% MB

132

TR Ranch Lake sediment: Gillisia sp. RNS1 N N Y 14.23* 50% MB Marinobacter sp. RNS5 N N Y 7.13* 50% MB Pseudomonas sp. RNS10 N N Y 11.37* 50% MB

Spotted Lake water: L. fusiformis SW22 N N N 10.99* 50% MB N. halotolerans SW17 N N Y 4.57 50% MB Salengentibacter sp. SW28 N N N 10.71* 50% MB

Control: E. coli TG-2 N N N 0 10% TSB 1 Ice recrystallzation activity (Yes, No or NA if not assayed for instance, if the media media was not conducive to assay). 2 Ice nucleation activity (Yes or No). 3 Biofilm formation is indicated by Yes (Y) if it is greather than that shown by E. coli TG-2, or No (N) if formation is less than or equal to that shown by E. coli TG-2. 4 mOsm readings are significantly greater (Tukey-Kramer HSD test) then that of the control, E. coli TG-2, which under the assay conditions used here was 0 mOsm are indicated by an asterisk. 5 Culture media is listed in the methods as Tryptic Soy Broth (TSB), Marine Broth (MB), East of 83 medium (E83), Liberty Lake medium (LM) and Peptone, Yeast Extract, Alkaline buffer Broth (PYA).

133

to E. coli TG-2. Since E. coli typically has low levels of biofilm production (Reisner et

al., 2006), any isolate giving a statistically greater value (p<0.05) was considered to form

a biofilm. Of the tested recovered bacteria, 6 of 21 formed biofilms with another 2

isolates showing media-dependent biofilm formation (Table 4.3). None of the freshwater

lake bacteria appeared to produce biofilms (0/9). In contrast, at least 50% (3/6 or 5/6,

depending on the media) of the freeze-thaw selected brine lake isolates tested positive for

biofilms and 67% (4/6) of the alkali lake samples were positive.

Discussion:

While freeze and osmotic stresses pose a number of challenges for microbes, they

share the common challenge of desiccation. Cells must cope with a higher extracellular

solute concentration no matter if they inhabit a saline environment or if ice formation

increases the external solute concentration. Thus we reasoned that there was potential for

cross-tolerance between freeze-thaw and osmotic stresses, in whole communities, as has

previously been explored in certain bacteria (Schmidt & Zink, 2000; Leblanc et al., 2003;

Schmid et al., 2009). The stress of repetitive freeze-thaw cycles on microbial consortia from six lakes located in the same region but of varying salinities, resulted in distinctly different profiles. Freshwater sediment and water-derived cultures were the most susceptible and viabilities were reduced at least 100,000-fold (Fig. 4.1). Alkali lake- derived water and sediment samples were generally the most resistant with 100-fold losses in viability, a remarkable difference. Thus, the hypothesis that a cross protection phenotype should apply to communities as a whole, appears to be reasonable. It should 134 be noted that overall, culturable consortia derived from the brine lake samples showed an intermediate resistance between these two extremes.

Often enrichments of water versus sediment communities were concordant in viability loss after 48 freeze-thaw treatments, but when they differed (no more than 10- to

100- fold), those from sediments were more susceptible than their water counterparts.

This may in part be because at least a portion of the water column experiences seasonal freezing, while the sediments do not. Although collections were made in late summer to mitigate these differences, such seasonal pre-selection may be responsible for this variation in some lakes. Alternatively, there were differences in the identified community members in water and sediments from the same lake, with generally less richness noted in the water samples (Table 4.2; Fig. 4.2), this too could explain such modest variation.

The surviving freeze-thaw isolates from freshwater lakes were largely spore formers such as Bacillus and Paenibacillus, as well as Pseudomonas. All three of these genera, and others recovered from the fresh water and sediment samples, including

Acinetobacter, Arthrobacter, and Sphingomonas have previously been recovered with freeze-thaw or ice-affinity selection from soil samples (Walker et al., 2006; Wilson &

Walker, 2010). Many of these, as well as Rhodococcus, have been previously been reported from glaciers, Antarctica, and alpine ice caves (eg. Christner et al., 2000; Gilbert et al., 2004; Lee et al., 2004; Margesin et al., 2004; Miteva et al., 2004).

Surviving microbes recovered from the brine lakes also contained some spore formers (Bacillus), and those previously associated with psychrotolerance such as

135

Pseudomonas (Wilson et al., 2006). However, since these lakes were alkaline (pH 10), it is not surprising that alkali- and halotolerant microbes such as Alkalibacterium,

Halomonas, Idiomarina, Nesterenkonia, and Roseinatronobacter were also identified.

Genera with similar environmental preferences were even more apparent amongst the

survivors from the alkali lakes (pH 7 and 8.2; Table 4.1), which included Gillisia,

Halomonas, Marinobacter, Nesterenkonia, and Salegentibacter, in addition to spore

formers (Bacillus and Lysinibacillus) and the near ubiquitous Pseudomonas. The vast

majority of these freeze-thaw resistant bacteria too, have previously been reported in cold

(or frozen), hypersaline and/or alkaline environments (eg. Brinkmeyer et al., 2003;

Stougaard et al., 2003; Steven et al., 2007, 2008; Niederberger et al., 2010), underscoring the success of the rigorous freeze-thaw selective regime.

Freeze-thaw resistance has previously been shown to be frequently affiliated with ice-association phenotypes (Chapter 1 and 3), which would presumably increase the likelihood of survival. Thus a subset of the isolates recovered from each lake were assayed to obtain insight into the types of adaptations found in lakes of different chemistries. Remarkably, none of the isolates demonstrated IRI, but Alteromonadales

EW4 and R. monicus LLW20 demonstrated low INA activity (Table 4.3). In contrast,

63% of isolates recovered after repetitive freeze-thaw cycles, but derived from soils showed some ice-association phenotype (Chapter 3). It is not known why there should be a difference between soil and lake samples, but in the case of the brine and alkali lakes we speculate that ice-association phenotypes may not provide suitable cross-tolerance.

136

Although ice-association characteristics were very rare amongst survivors, more than half of them showed elevated intracellular osmolyte content compared to controls

(Table 4.3). As anticipated, a number of the genera recovered from the brine and alkali lakes have previously been associated with osmoprotectants or hyperosmotic environments, including Acinetobacter (Mogilnaya et al., 2005), Arthrobacter

(Yamamoto et al., 2001), Bacillus (Kuhlmann & Bremer, 2002), Halomonas, Idiomarina, and Marinobacter (Naganuma et al., 2005), Nesterenkonia (Zhang et al., 2008), and

Paenibacillus (Sokhansanj et al., 2005). However, two isolates from the freshwater lakes also showed elevated osmolarity, with one showing levels as high as some isolates from brine and alkali lake samples. Since these bacteria survived multiple freeze-thaw cycles, these results are consistant with the known association of osmolytes with cryoprotection

(eg. Shahjee et al., 2002).

To further test the hypothesis of cross-tolerance, the ability of the isolates to form biofilms was tested since biofilms are reported to provide resistance to both osmotic and low temperature stress (Knobloch et al., 2001; Williams et al., 2009). Of the 21 isolates assayed, 9 formed biofilms in at least one of the tested growth conditions and all of these were derived from the brine or alkali lakes. Strikingly, some of the isolates showed more than one adaptation with two brine lake isolates demonstrating INA, elevated osmolarity, and biofilm formation. Such multiple adaptations would presumably serve to ensure survival under the most extreme conditions and, in fact, Alteromonadales EW4is related to one of the most “extreme psychrophiles” known with multiple membrane, protein and metabolite adaptations (Riley et al., 2008).

137

To recapitulate, the observed resistance of microbial consortia to freeze-thaw

stress increased with the osmotic stress posed by the “home lake” chemistry, indicating

that adaptation to osmotic stress increases freeze-thaw survival. Resistant phenotypes to

freeze-thaw stress in lakes were not frequently ice-association activities but rather

seemed to be dominated by spore formation, high intracellular osmolyte concentrations

and biofilm formation, all of which are known to facilitate survival against freeze-thaw

and osmotic stresses (Knobloch et al., 2001; Shahjee et al., 2002; Williams et al., 2009).

Taken together then, the data presented here strongly support the notion of community

cross-tolerance between freeze-thaw and osmotic stresses, as has been previously shown for individual isolates (Schmidt & Zink, 2000; Le Blanc et al., 2003; Schmid et al.,

2009).

Acknowledgements:

An NSERC (Canada) grant to V.K.W and NSERC scholarship to S.L.W.

financially supported this work. Drs. B. Cumming, S. Lougheed, and B. Tufts are

thanked for encouragement and information on TR Ranch Lake, assistance with the

phylogenetic tree, as well as the vapour pressure osmometer, respectively. Dr. G. Palmer

is greatly thanked for technical assistance and for facilitating the collection expedition in

British Columbia. M. Chalifoux is acknowledged for her invaluable assistance and Dr.

G. Voordouw for the E. coli TG-2. T. Vanderveer assisted with some data collection. A.

Kanawaty and S. Franchuk are acknowledged for some preliminary analysis (including

lake water chemistry at the Analytical Services Unit; Queen’s University) and B. 138

Momciu, R. Murray and A. Stanczak are also acknowledged for preliminary analysis of a pilot study.

139

Literature Cited:

Amato, P., Doyle, S., and Christner, B.C. (2009). Macromolecular synthesis by yeasts under frozen conditions. Environ. Microb. 11: 589-596.

Balestrino, D., Ghigo, J.-M., Charbonnel, N., Haagensen, J.A.J., and Forestier, C. (2008). The characterization of functions involved in the establishment and maturation of Klebsiella pneumoniae in vitro biofilm reveals dual roles for surface exopolysaccharides. Environ. Microb. 10: 685-701.

Belgrader, P., Hansfor, D., Kovacs, G.T.A., Venkateswaran, K., Mariella, R., Milanovich, F., Nasarabadi, S., Okuzumi, M., Pourahmadi, F., and Northrup, M.A. (1999). A minisonicator to rapidly disrupt bacterial spores for DNA analysis. Anal. Chem. 71: 4232-4236.

Brinkmeyer, R., Knittel, K., Jürgens, J., Weyland, H., Amann, R., and Helmke, E. (2003). Diversity and structure of bacterial communities in Arctic versus Antarctic pack ice. Appl. Environ. Microb. 69: 6610-6619.

Christner, B.C., Mosley-Thompson, E., Thompson, L.G., Zagorodnov, V., Sandman, K., and Reeve, J.N. (2000). Recovery and identification of viable bacteria immured in glacial ice. Icarus 144: 479-485.

Gilbert, J.A., Hill, P.J., Dodd, C.E.R., and Laybourn-Parry, J. (2004). Demonstration of antifreeze protein activity in Antarctic lake bacteria. Microbiology 150: 171-180.

140

Gordienko, R., Ohno, H., Singh, V.K., Jia, Z., Ripmeester, J.A., and Walker, V.K. (2010). Towards a green hydrate inhibitor: imaging antifreeze proteins on clathrates. PLoS ONE 5: e8953.

Jones, P.G., VanBogelen, R.A., and Neidhardt, F.C. (1987). Induction of proteins in response to low temperature in Escherichia coli. J. Bacteriol. 169: 2092-2095.

Knobloch, J.K.-M., Bartscht, K., Sabottke, A., Rohde, H., Feucht, H.-H., and Mack, D. (2001). Biofilm formation by Staphylococcus epidermidis depends on functional RsbU, an activator of the sigB operon: differential activation mechanisms due to ethanol and salt stress. J. Bacteriol. 183: 2624-2633.

Kuhlmann, A.U., and Bremer, E. (2002). Osmotically regulated synthesis of the compatible solute ectoine in Bacillus pasteurii and related Bacillus spp. Appl. Environ. Microb. 68: 772-783.

Leblanc, L., Leboeuf, C., Leroi, F., Hartke, A., Auffray, Y. (2003). Comparison between NaCl tolerance response and acclimation to cold temperature in Shewanella putrefaciens. Current Microbiol. 46: 157-162.

Lee, Y.K., Kim, H.W., Cho, K.H., Kang, S.-H., Lee, H.K., and Kim, Y. (2004). Phylogenetic analysis of culturable arctic bacteria. Ocean Polar Res. 26: 51-58.

Lindow, S.E., Arny, D.C., and Upper, C.D. (1978). Erwinia herbicola: a bacterial ice nucleus active in increasing frost injury to corn. Phytopathology 68: 523-527.

Mader, H.M., Pettitt, M.E., Wadham, J.L., Wolff, E.W., and Parkes, R.J. (2006). Subsurface ice as a microbial habitat. Geology 34: 169-172.

Maki, L.R., Galyan, E.L., Chang-Chien, M.-M., and Caldwell, R.D. (1974). Ice nucleation induced by Pseudomonas syringae. Appl. Microbiol. 28: 456-459. 141

Margesin, R., Schumann, P., Spröer, C., and Gounot, A.-M. (2004). Arthrobacter psychrophenolicus sp. nov., isolated from an alpine ice cave. Int. J. Sys. Evol. Microb. 54: 2067-2072.

McAuliffe, L., Ayling, R.D., Ellis, R.J., and Nicholas, R.A.J. (2008). Biofilm-grown Mycoplasma mycoides subsp. subsp. mycoides SC exhibit both phenotypic and genotypic variation compared with planktonic cells. Vet. Microb. 129: 315-324.

Mogilnaya, O.A., Lobova, T.I., Kargatova, T.V., and Popova, L.Y. (2005). Biofilm formation by bacterial associations under various salinities and copper ion stress. Biofouling 21: 247-255.

Naganuma, T., Hua, P.N., Okamoto, T., Ban, S., Imura, S., and Kanda, H. (2005). Depth distribution of euryhaline halophilic bacteria in Suribati Ike, a meromictic lake in East Antarctica. Polar Biol. 28: 964-970.

Niederberger, T.D., Perreault, N.N., Tille, S., Sherwood Lollar, B., Lacrampe-Couloume, G., Andersen, D., Greer, C.W., Pollard, W., and Whyte, L.G. (2010). Microbial characterization of a subzero, hypersaline methane seep in the Canadian high Arctic. ISME J. doi:10.1038/ismej.2010.57.

O’Toole, G.A., and Kolter, R. (1998) . Initiation of biofilm formation in Pseudomonas fluorescens WCS365 proceeds via multiple, convergent signaling pathways: a genetic analysis. Mol. Microb. 28: 449-461.

Reisner, A., Krogfelt, K.A., Klein, B.M., Zechner, E.L., and Molin, S. (2006). In vitro biofilm formation of commensal and pathogenic Escherichia coli strains: impact of environmental and genetic factors. J. Bacteriol. 188: 3572-3581.

Riley, M., Staley, J.T., Danchin, A., Zhang Wang, T., Brettin, T.S., Hauser, L.J., Land, M.L., and Thompson, L.S. (2008). Genomics of an extreme psychrophile, Psychromonas ingrahamii. BMC Genomics 9: 210-228.

Sakamoto, T., and Murata, N. (2002). Regulation of the desaturation of fatty acids and its role in tolerance to cold and salt stress. Curr. Opin. Microbiol. 5: 206-210. 142

Schmid, B., Klumpp, J., Raimann, E., Loessner, M.J., Stephan, R., and Tasara, T. (2009). Role of cold shock proteins in growth of Listeria monocytogenes under cold and osmotic stress conditions. Appl. Environ. Microb. 75: 1621-1627.

Schmidt, G., and Zink, R. (2000). Basic features of the stress response in three species of bifidobacteria: B. longum, B. adolescentis, and B. breve. Int. J. Food Microb. 55: 41-45.

Sokhansanj, A., Asghar Karkhane, A., Rastgar Jazii, F. (2005). Identification and characterization of salt-inducible polypeptide in Paenibacillus sp., a moderately halophilic bacterium. J. Biosci. Bioeng. 100: 573-575.

Steven, B., Pollard, W.H., Greer, C.W., and Whyte, L.G. (2008). Microbial diversity and activity through a permafrost/ground ice core profile from the Canadian high Arctic. Environ. Microb. 10: 3388-3403.

Stougaard, P., Jørgensen, F., Johnsen, M.G., and Hansen, O.C. (2002). Microbial diversity in ikaite tufa columns: an alkaline, cold ecological niche in Greenland. Environ. Microb. 4: 487-493.

Swofford, D.L. (2002). PAUP*. Phylogenetic analysis using parsimony (*and other methods). Version 4. Sunderland, MA.

Vali, G. (1971). Quantitative evaluation of experimental results on the heterogeneous freezing nucleation of supercooled liquids. J. Atmos. Sci. 28: 402-409.

Vanderlinde, E.M., Muszyński, A., Harrison, J.J., Koval, S.F., Foreman, D.L., Ceri, H., Kannenberg, E.L., Carlson, R.W., and Yost, C.K. (2009). Rhizobium leguminosarum biovar viciae 3841, deficient in 27-hydroxyoctacosanoate-modified lipopolysaccharide, is impaired in desiccation tolerance, biofilm formation and motility. Microbiology 155: 3055-3069. 143

Walker, V.K., Palmer, G.R., and Voordouw, G. (2006). Freeze-thaw tolerance and clues to the winter survival of a soil community. Appl. Environ. Microb. 72: 1784-1792.

Williams, H.N., Turng, B-F., Kelley, J.I. (2009). Survival response of Bacteriovorax in surface biofilm versus suspensions when stressed by extremes in environmental conditions. Microb. Ecol. 58: 474-484.

Wilson, S.L., Kelley, D.L., and Walker, V.K. (2006). Ice-active characteristics of soil bacteria selected by ice-affinity. Environ. Microb. 8: 1816-1824.

Wilson, S.L., and Walker, V.K. (2010). Selection of low-temperature resistance in bacteria and potential applications. Environ. Tech. 31: 943-956.

Yamamoto, T., Maruta, K., Watanabe, H., Yamashita, H., Kubota, M., Fukuda, S., and Kurimoto, M. (2001). Trehalose-producing operon treYZ from Arthrobacter ramosus S34. Biosci. Biotechnol. Biochem. 65: 1419-1423.

Yancey, P.H., Clark, M.E., Hand, S.C., Bowlus, R.D., and Somero, G.N. (1982). Living with water stress: Evolution of osmolyte systems. Science 217: 1214-1222.

Yumoto, I., Hirota, K., Nodasaka, Y., Yokota, Y., Hoshino, T., and Nakajima, K. (2004). Alkalibacterium psychrotolerans sp. nov., a psychrotolerant obligate alkaliphile that reduces an indigo dye. Int. J. Sys. Evol. Microb. 54: 2379-2383.

Zhang, B., Bao, X., Wang, L., and Yang, S.S. (2008). Cloning and characterization of the gene cluster for biosynthesis of ectoine from Nesterenkonia halobia DSM 20541. J. Microbiol. 46: 309-318.

144

Chapter 5: Freeze-Thaw Selection of a Cultured Oil Well-Derived Community

Abstract:

Gas hydrates form spontaneously given sufficient moisture and small gas molecules as

well as high pressures and near freezing temperatures. This can result in blockages of oil

wells and gas pipelines, with consequential safety, environmental and economic

concerns. Currently large amounts of thermodynamic inhibitors such as methanol are

used to prevent hydrate growth. However, "green inhibitors", such as antifreeze proteins

(AFPs), have recently shown promise in lab-scale experiments. In an attempt to enrich for microbes with potentially useful properties, we selected for freeze-thaw resistant culturable microbes from a produced water (crude oil and water mixture) sample.

Alternating freeze and thaw cycles decreased consortia abundance approximately 10,000-

fold. Selection also resulted in a shift within the dominant community members as

determined by DNA analysis. Some ice recrystallization inhibition activity was present

in the recovered cultures, indicating the presence of ice-association properties. Given

that a subset of the selected microbes may be able to inhibit the growth of ice crystals,

they may also have the potential to inhibit hydrate growth.

145

Introduction:

Gas hydrates or clathrates are composed of stabilized water molecules and enclathrated gas molecules, such as methane or propane gas hydrates (Kvenvolden,

1999). Hydrates form spontaneously under high pressure and low temperature conditions

(Sloan, 1998 in Kvenvolden, 1999), including in gas and oil pipelines with potentially disastrous results. For example, hydrates were involved in the Piper Alpha (1988) and the recent Deepwater Horizon (2010) disasters. Oil companies routinely add large amounts of thermodynamic hydrate growth inhibitors, such as methanol, to exploratory, recovery and transport pipelines (Koh et al., 2002; Sloan, 2003). However, methanol is toxic, highly flammable and costly. Furthermore, the methanol must be extracted prior to refining crude oil.

Alternative hydrate growth inhibitors, including antifreeze proteins (AFPs) have recently shown promise. Type I AFP (a fish AFP) and an insect AFP (CfAFP) were both shown to absorb to and alter the growth of a model hydrate, tetrahydrofuran (THF; Zeng et al., 2003; Zeng et al., 2006). The former (Type I) was more effective in preventing growth than polyvinylpyrrolidone (PVP), which is a known chemical kinetic inhibitor

(Sloan, 1995; Lederhos et al., 1996). It seems likely that the mechanism of action was adsorption-inhibition (Zeng et al., 2003), as is the case with ice growth inhibition by

AFPs. This was further investigated by Gordienko et al., (2010), who found that Type III

AFP (a fish AFP) and a grass AFP (LpAFP) were both able to adsorb to the model hydrate and prevent growth more efficiently than PVP. These researchers also showed these same AFPs inhibited “real” gas hydrates including methane, propane and/or natural

146

gas hydrate formation (Zeng et al., 2003; Zeng et al., 2006; Gordienko et al., 2010; Ohno

et al., 2010), thus piquing the interests of industrial and academic researchers alike.

However, while promising in lab-scale experiments, AFP use for industrial-scale

applications is not yet feasible as it is challenging to obtain large quantities of these

eukaryotic proteins.

Previously, freeze-thaw selection of cultured communities has resulted in the

selection of microbes with ice-association properties including ice recrystallization

inhibition and ice-crystal morphology activities (Walker et al., 2006; Chapter 3).

However, these consortia were sampled aerobically, whereas gas hydrate formation is

often under oxygen-limiting conditions. Oxygen does not readily form hydrates

(Hammerschmidt et al., 1934). In contrast, methane forms the most commonly found

hydrate and a large portion of the methane is from methanogenic microbes (Davidson et

al., 1978; Kvenvolden, 1999). These, in turn, associate with anaerobic methane oxidizers

and sulphate reducing bacteria (SRB; Hoehler et al., 1994; Hinrichs et al., 1999 and

Boetius et al., 2000). Thus, in an effort to identify suitable, culturable microbes with

AFPs or similar properties, a crude oil and water mixture (produced water) sample was

obtained from a production well. Anaerobic culture, suitable for SRB, coupled with

cyclic freeze-thaw treatments were used in a novel effort to identify microbes with ice

and/or hydrate association properties, as a first step for their eventual use as inhibitors.

Materials and Methods:

Sample Enrichment Conditions and Freeze-Thaw Selection: A sample mixture of oil and

water from a production well (produced water) was obtained from well #18 of an Alberta 147

(Canada) oil field. The water fraction was separated as much as possible from the visible

hydrocarbon layer and then used to inoculate saline Postgate C (sPGC) media, in order to

enrich for anaerobic sulphate reducing bacteria (Postgate, 1979), with 1mg/L resazurin

(Forsberg, 1980). The media was supplemented with oxyrase (~0.03 mL per 2 mL media)

and kept at 37°C for at least 30 min (Oxyrase Inc, Mansfield, OH, USA) in order to

maintain an anaerobic state during bench top manipulation. An enrichment culture (5 mL

produced water and 20 mL sPGC) was grown in an anaerobic chamber (without shaking)

for 9 days at room temperature. This was subcultured (2 mL culture; 20 mL fresh sPGC)

for a subsequent 19 days. From this subculture, five replicate cultures were initiated,

which were cultured anaerobically for 3 days at 22°C and subsequently incubated at 4°C

for 4 days prior to freeze-thaw selection. All five replicate enrichments were kept

anaerobic while being subjected to 48 consecutive cycles of alternating +5°C and -18°C

using a cryocycler (Walker et al., 2006; Walker et al., 2008). Silver iodide was added as

a precaution to the cultures to ensure consistent nucleation (Walker et al., 2006).

Estimation of Microbial Abundance and Diversity: A 96-well plate most probable

number (MPN; modified from McCrady, 1915; Cochran, 1950) technique was used to

estimate microbial abundance before and after selection. Briefly, 96-well plates were

inoculated (starting sample to 108 dilution in sPGC with oxyrase) using 5 replicate wells per dilution. The plates were subsequently incubated in an anaerobic chamber at ~22°C for 3 days and standard calculations for MPN were used.

148

Samples (400-500 μL) of each culture were removed prior to and subsequent to freeze-thaw cycle treatments. These were treated with ethidium monoazide bromide

(EMA) and light according to Pistz et al., (2007) in order to decrease the probability of polymerase chain reaction amplification of DNA from dead or moribund cells in subsequent analysis. Genomic DNA was extracted using a DNeasy kit (Gram positive protocol, Qiagen, Mississauga, ON, Canada), and was used for clone library construction and denaturing gradient gel electrophoresis (DGGE) analysis.

Clone libraries of partial gyrase subunit B (gyrB) sequences, as an alternative to

16S rDNA sequence analysis, were constructed by PCR amplification of gyrB from the consortia DNA obtained prior to and following selection from a randomly chosen two of the five replicate, freeze-thaw selected cultures. Primers UP1 (5´-

GAAGTCATCATGACCGTTCTGCAYGC-NGGNGGNAARTTYGA; Yamamoto and

Harayama, 1995; Delmas et al., 2006) and 181r (5´CAGGAAACAGCTATGACCARRT-

GNGTNCCNCC; Delmas et al., 2006) were employed, using standard PCR conditions

(annealing temperature of 40°C) and the amplified DNA was subsequently cloned using

pCR2.1 and E. coli DH5-α Top 10´ cells (TOPO cloning kit; Invitrogen, Carlsbad, CA,

USA). Ten clones from each of the replicates, before and after treatment, were randomly

selected (40 clones in total). The insert from these clones were PCR amplified as

described above.

In order to assess the richness of the clone libraries, restriction fragment length

polymorphism (RFLP) analysis was performed. Briefly, the 40 sequences derived from

the clones were subjected to restriction enzyme digestion [HhaI (5´-G/CGC-3´) or HaeIII

149

(5´-GG/CC-3´)] for 1 h at 37°C, as suggested by the manufacturers (Fermentas,

Burlington, ON, Canada). After visualization on a 1-1.5% agarose gel, each fragment was categorized according to its RFLP fingerprint. At least one clone per fingerprint was sequenced (Plateforme de Génomique at the Centre de recherché du CHUL, Quebec, QC,

Canada) using UP1 and/or 181r primers. Sequences were prepared and assembled using

CodonCode Aligner (http://www.codoncode.com/aligner/trial.htm) and “Manipulate and

Display a DNA sequence” (http://arbl.cvmbs.colostate.edu/molkit/manip). The putative identity of the organism from which the gyrB gene originated was determined following

BLASTn searches (NCBI; http://www.ncbi.nlm.nih.gov/blast; Altschul et al., 1997).

Denaturing gradient gel electrophoresis (DGGE) was also used to visualize DNA from abundant species. Partial bacterial 16S rRNA genes from the five replicate consortia prior to and following selection were PCR amplified using primer 338f (5´-

ACTCCTACGGG-AGGCAGCAG) containing a GC clamp (Øvreås et al., 1997) and

907r (5´-CCGTCAA-TTCMTTTRAGTTT; Lane et al., 1985) using standard conditions

(Buckeridge et al., 2010). An equimolar amount of DNA was loaded into each well of

6% polyacrylamide gels, with 16-24% formamide and 2.8-4.2M urea as the denaturing gradient (Buckeridge et al., 2010). Amplicons were separated over the course of a 20 h migration (12 min at 120V then 65V for the remainder of the time) in 1x Tris-acetate-

EDTA buffer which was heated to 60°C with a D-Code universal mutation detection system (BIO-RAD Laboratories, Hercules, CA, USA). The gels were subsequently stained with SYBR green (10x; Invitrogen, Burlington, ON, Canada) and imaged using a

Chemi Genius, Bio Imaging System and GeneSnap (Syngene, Cambridge, UK).

150

GeneTools (Syngene, Cambridge, UK) was used to analyze the similarity amongst the

lanes. The resultant similarity matrix was analyzed statistically, without the use of the

Monte Carlo method (Kropf et al., 2004). DGGE analysis was conducted in triplicate.

Ice-Association Characterization: Ice recrystallization inhibition (IRI) and ice nucleation

activity (INA) assays were as previously described (Chapter 1.1; Wilson & Walker,

2010), with the enrichment cultures and the appropriate controls. Each of the five

replicate post-selection cultures were recovered anaerobically in fresh media (60 μL from

a frozen stock and 1.5 mL sPGC; 3 days at ~22°C, followed by 4-5 days at 4°C). The

recovered enrichments were subsequently subcultured twice (60 μL culture and 1.5 mL

sPGC; 8 days at room temp and 4 days at 4°C) for a total of 3 cultures per replicate, each

of which were assayed for IRI and INA activity in duplicate (n=6). Assays were

performed aerobically.

Results:

Microbial Abundance and Diversity: After 48 freeze-thaw cycles, the consortia

population was reduced from 107 cells per mL by approximately four logarithms of viability, to ~103 cells per mL (Fig. 5.1). Gyrase B clone libraries indicated that freeze-

thaw selection also decreased microbial richness. However, with only three RFLP

fingerprints prior to selection, and only one following (Fig. 5.2), it suggested limited initial richness. This was confirmed by sequencing of the clonal inserts (8 and 7 clones from the 0 and 48 cycle libraries, respectively) which resulted in only two distinct DNA sequences (~ 11% identity) within the libraries. The closest match in the NCBI BLASTn 151

1.00E+08 1.00E+07 1.00E+06 1.00E+05 1.00E+04

1.00E+03 Avg. MPN 1.00E+02 1.00E+01 1.00E+00 0 48 Freeze-Thaw Cycles

Figure 5.1. Average viable cells per mL, as estimated by the most probable number (MPN; anaerobically cultured), of the five replicate produced water consortia prior to (0 cycles) and following (48 cycles) freeze-thaw treatment. Replicates were normalized to a starting density of 1.0 x 107. Standard deviation of the replicates post-selection is indicated by error bars.

152

1 kb

400 bp 300 bp

200 bp

100 bp A B C B A A A

Figure 5.2. Restriction fragment length polymorphism (RFLP) screening of the gyrase B clone libraries. The first five clones (two lanes per clones) following the DNA ladder are from the clone libraries prior to freeze-thaw treatment, while the final two are post- treatment. Each sample was digested with HaeIII or HhaI (left and right lane per paired lanes, respectively). The first, fifth, six, and seventh clones (from the left) represent RFLP fingerprint pattern A. The second and fourth clones represent RFLP fingerprint pattern B, while the third represents pattern C.

153 database for all of the sequenced clones (representing all three RFLP fingerprints), was to

Desulfomicrobium baculatum (accession number CP001629.1, 92-93% identity, ~400-

500 nucleotides per sequence). Lesser matches in the database were gyrB specific.

DGGE analysis indicated that there was a shift in the microbial community diversity as a result of freeze-thaw stress (Fig. 5.3). Several minor bands were more prominent in the unselected samples. Of three major bands observed, one increased, one decreased and one appeared unchanged following selection, relative to its unselected counterpart. This assessment was tested by analysis of the Pearson correlation coefficients (similarity matrices) amongst the gel lanes which showed that the replicates before and after selection were more similar amongst themselves than between the conditions. Thus the lanes grouped based on pre- or post-freeze-thaw selection were statistically different (p <

0.01; data not shown).

Ice-Association Characterization: The five recovered replicate consortia were each assayed (n=6) for ice-association. Enrichments were cultured anaerobically, but for logistical reasons were assayed aerobically. None of the consortia demonstrated INA, as indicated by a nucleation temperature below -10°C (not shown). Three of the five consortia demonstrated some IRI activity when originally cultured. This activity was apparently lost upon subculturing, since the assays with the subcultures failed to show

IRI activity.

154

Figure 5.3. 16S rRNA gene sequences visualized after denaturing gradient gel electrophoresis (DGGE) analysis. Lanes A-C are from replicates 5, 4 and 3 (respectively) prior to freeze-thaw selection, while lanes D-F are from replicates 5, 4 and 3 (respectively) following 48 freeze-thaw cycles. Arrows indicate bands which appear to change in abundance by the treatment. Image is a composite, however, all lanes are from the same representative DGGE gel.

155

Discussion:

Freeze-thaw selection has previously been used to recover microbes with ice- association characteristics, including IRI and INA activity (Walker et al., 2006; Wilson &

Walker, 2010). However, this technique has not been previously applied to anaerobic cultures. Because of technical challenges, the cultures enriched for SRB were not sampled periodically during selection, but only analyzed after 48 cycles. At the conclusion of the experiments, however, the abundance of the oil well-derived consortia had been reduced approximately 10,000-fold (Fig. 5.1).

Given the dramatic decrease in cell viability, it was expected that clone libraries of partial gyrB sequences screened using RFLP analysis would show a reduction in microbial diversity. Although there was a decrease from three RFLP fingerprints to one subsequent to freeze-thaw treatment (Fig. 5.2), all RFLP patterns corresponded to a single sequence that showed limited identity to a single SRB, D. baculatum, which is a mesophilic, strict anaerobe (Sharak Genthner & Devereux, 2005). However, the recovery of a sequence derived from possibly a single organism indicates that either the sample was a near monoculture to begin with, or that the library construction was flawed, possibly due to limitations of the ‘universal’ gyrase primers for these types of samples or alternatively, by a disproportionate selection of D. baculatum related sequences. That selection did indeed result in a shift within the predominant community members as well as viable cells was shown by DGGE analysis and multiple bands (Fig. 5.3) indicated that a broader range of microbes were present prior to selection. Thus our hypothesis is that 156

the gyrase primers used were not ‘universal’. In order to quantify the DNA differences

seen in the DGGE analysis, and to identify species, deep sequencing analysis will be

done with DNA from two of the five replicate cultures, pre and post freeze-thaw stress. It

is anticipated that this will allow for a more detailed analysis of the microbial diversity

within the consortia.

Despite a lack of information on the identities of the surviving microbes, replicate

cultures recovered following selection were assayed for INA and IRI activities. None of

the cultures demonstrated INA activity, but three of the five replicate cultures recovered

post selection demonstrated some IRI activity. When they were subsequently subcultured no IRI activity was detected. It is not known if subculturing resulted in the loss of one or more members of the consortia. Clearly, multiple rounds of subculturing without freeze- thaw selection should be avoided in future experiments.

In conclusion, it was hypothesized that anaerobic freeze-thaw stress would

decrease consortia abundance, resulting in a shift in the community structure, and

allowing the recovery of ice-active oil well-derived species. We have now partially achieved that goal. Since it is believed that a subset of the microbes post-selection showed IRI activity, future experiments may have the potential for the discovery of individuals or communities that may inhibit the growth of gas hydrates.

157

Acknowledgements:

This work was financially supported by a Queen’s Research Chair and an NSERC grant

to V.K.W. and an NSERC scholarship to S.L.W. Thanks to Drs. J. Ramsay (Queen’s

University), and G. Voordouw (University of Calgary), for the anaerobic chamber and

produced water sample respectively. Thanks also to M. Chalifoux, S. Cornish and S.

Johnston for their assistance and to N. Kumar for passing along his DGGE knowledge.

Dr. Allen-Vercoe and J. McDonald (University of Guelph) are thanked for their help with

the analysis of the DGGE images.

158

Literature Cited:

Buckeridge, K.M., Zufelt, E., Chu, H., and Grogan, P. (2010). Soil nitrogen cycling rates in low arctic shrub tundra are enhanced by litter feedbacks. Plant Soil 330: 407-421.

Cochran, W.G. (1950). Estimation of bacterial densities by means of the “most probable number.” Biometrics 6: 105-116.

Delmas, J., Breysse, F., Devulder, G., Flandrois, J.-P., and Chomarat, M. (2006). Rapid identification of Enterobacteriaceae by sequencing DNA gyrase subunit B encoding gene. Diagn. Microbiol. Infect. Dis. 55: 263-268.

Forsberg, C.W. (1980). Sulfide production from cysteine by Desulfovibrio desulfuricans. Appl. Environ. Microb. 39: 453-455.

Gordienko, R., Ohno, H., Singh, V.K., Jia, Z., Ripmeester, J.A., and Walker, V.K. (2010). Towards a green hydrate inhibitor: imaging antifreeze proteins on clathrates. PLoS ONE 5: e8953.

Hammerschmidt, E.G. (1934). Formation of gas hydrates in natural gas transmission lines. Ind. Eng. Chem. 26: 851-855.

Hinrichs, K.-U., Hayes, J.M., Sylva, S.P., Brewer, P.G., and DeLong, E.F. (1999). Methane-consuming archaebacteria in marine sediments. Nature 398: 802-805.

Koh, C.A., Westacott, R.E., Zhang, W., Hirachand, K., Creek, J.L. and Soper, A.K. (2002). Mechanisms of gas hydrate formation and inhibition. Fluid Phase Equilibr. 194- 197: 143-151.

159

Kropf, S., Heuer, H., Grüning, M., and Smalla, K. (2004). Significance test for comparing complex microbial community fingerprints using pairwise similarity measures. J. Microbiol. Methods 57: 187-195.

Kvenvolden, K.A. (1999). Potential effects of gas hydrate on human welfare. Proc. Natl. Acad. Sci. USA 96: 3420-3426.

Lederhos, J.P., Long, J.P., Sum, A., Christiansen, R.L., and Sloan, E.D. Jr. (1996). Effective kinetic inhibitors for natural gas hydrates. Chem. Eng. Science 51: 1221-1229.

McCrady, M.H. (1915). The numerical interpretation of fermentation-tube results. J. Infect. Dis. 17: 183-212.

Ohno, H., Susilo, R., Gordienko, R., Ripmeester, J., and Walker, V.K. (2010). Interaction of antifreeze proteins with hydrocarbon hydrates. Chem. Eur. J. DOI: 10.1002/chem.200903201.

Øvreås, L., Forney, L., Daae, F.L., and Torsvik, V. (1997). Distribution of bacterioplankton in Meromictic Lake Sælenvannet, as determined by denaturing gradient gel electrophoresis of PCR-amplified gene fragments coding for 16S rRNA. Appl. Environ. Microb. 63: 3367-3373.

Pisz, J.M., Lawrence, J.R., Schafer, A.N., and Siciliano, S.D. (2007). Differentiation of genes extracted from non-viable versus viable micro-organisms in environmental samples using ethidium monoazide bromide. J. Microbiol. Methods 71: 312-318.

Postgate, J.R. (1979). The sulphate-reducing bacteria (2nd Edition). Cambridge University Press, Cambridge, UK.

Sharak Genthner, B.R., and Devereux, R. (2005). Genus 1. Desulfomicrobium Rozanova, Nazina and Galushko 1994, 370 (Effective publication: Rozanova, Nazina and Galushko 1988, 518). Bergey’s Manual of Systematic Bacteriology (2nd Edition) Vol. 2, part C. Garrity, G.M., ed., Springer Science + Business Media, Inc., New York, New York. pg. 944-948.

Sloan, E.D. (1995). A method for controlling clathrate hydrates in fluid systems. US Patent 5,420,370 (for PVP). International patents also filed.

Sloan, E.D. (1998). Clathrate Hydrates of Natural Gas (2nd Edition), Dekker, New York.

160

Sloan, E.D. (2003). Fundamental principles and applications of natural gas hydrates. Nature 426: 353-359.

Walker, V.K., Palmer, G.R., and Voordouw, G. (2006). Freeze-thaw tolerance and clues to the winter survival of a soil community. Appl. Environ. Microb. 72: 1784-1792.

Walker, V.K., Wilson, S.L., Wu, Z., Mia, D.N., Zeng, H., Ripmeester, J.A., and Palmer, G.R. (2008). Screening microbes for ice-association proteins with potential applications as ‘Green Inhibitors’ for Gas Hydrates. In Emerging Environmental Technologies. Shah, V., ed., Springer Science + Business Media, New York, NY, pp. 29-42.

Wilson, S.L., and Walker, V.K. (2010). Selection of low-temperature resistance in bacteria and potential applications. Environ. Tech. 31: 943-956.

Yamamoto, S., and Harayama, S. (1995). PCR amplification and direct sequencing of gyrB genes with universal primers and their application to the detection and taxonomic analysis of Pseudomonas putida strains. Appl. Environ. Microb. 61: 1104-1109.

Zeng, H., Wilson, L.D., Walker, V.K., and Ripmeester, J.A. (2003). The inhibition of tetrahydrofuran clathrate-hydrate formation with antifreeze protein. Can. J. Phys. 81: 17- 24.

161

Chapter 6: General Discussion

Low Temperature and Osmotic Stresses - Challenges and Resistance:

Low temperature and hyperosmotic environments induce a number of physiological changes in microbes, including protein, nucleic acid and membrane stability and function, as well as changes in cellular volume (Bennett et al., 1981; Roth et al., 1985; Meury, 1988; Chattopadhyay & Jagannadham, 2001; Beaufils et al., 2007).

The presence of ice can also cause physical damage and potential membrane rupture

(Gomez Zavaglia et al., 2000). Microbes can resist these physiological effects by producing stress-induced proteins (eg. cold shock proteins), altering the cellular membrane structure, and the accumulation of solutes to maintain osmotic balance and cell volume (Klein et al., 1999; Gomez Zavaglia et al., 2000; Panoff et al., 2000;

Chattopadhyay & Jagannadham, 2001). Microbes within biofilms are also shielded from stress (Knobloch et al., 2001; Williams et al., 2009).

Additionally, microbes in low temperature environments may also express one of

two types of proteins capable of directly associating with ice, antifreeze proteins (AFPs)

and ice nucleation proteins (INPs). AFPs decrease the freezing point of a solution and

inhibit the growth (recystallization) of small ice crystals into larger, potentially damaging

crystals. The latter is referred to as ice recrystallization inhibition (IRI; Raymond &

DeVries, 1977). Conversely, INPs nucleate ice formation at relatively high subzero

temperatures, thereby preventing supercooling (Cochet & Widehem, 2000). These

protein activities, as well as the ability to change the morphology of single ice crystals,

162

are present in a growing number of microbes, and will collectively be described as ‘ice-

association’ activities.

Selecting for Microbes from Terrestrial Environments – the Role of Ice-Association

Activities:

A substantial amount of previous research has observed the microbial diversity in cold or frozen environments and the impact of freeze-thaw cycling on microbial communities (eg. Yergeau & Kowalchuk, 2008). Except for a few (Walker et al., 2006;

Wilson & Walker, 2010), most have done so in ‘extreme’ environments, and to the best of our knowledge, none have looked at the role of ice-association activities for community-level resistance. Here, in contrast, selection of resistant microbes from more temperate environments, and the role of ice-association activities in resistance were evaluated. Two selection regimes, ice-affinity and freeze-thaw cycling, were used with microbial consortia derived from temperate and boreal soils.

Ice-affinity selection resulted in the isolation of a number of psychrotolerant microbes from both temperate and boreal soil enrichments (Chapter 2). Many of the isolated microbes demonstrated ice-association activities. This suggests that these activities, at least in part, conferred the observed resistance and allowed for microbial incorporation into growing ice. Furthermore, a large proportion of the microbes typically found within sea ice and glaciers, have also been, at least at the genera level, affiliated with ice-association activities (Chapter 1; Table 6.1), possibly promoting their incorporation into the ice as it grows. Table 6.1 summarizes the isolates natural history. 163

Table 6.1 Genera isolated with ice-affinity and/or freeze-thaw selection, a summary. Genus Isolated From1 Psychrotolerant2 Halotolerant2 Alkalitolerant2 INP/INA2 AFP/IRI2 Acinetobacter DL and GL (IA); Fly Yamahira et al., Lobova et al., Yamahira et al., No Chapter 2 2008 2004 2008 Arthrobacter GL (IA); Fly Reddy et al., 2000 Reddy et al., 2000 Qvit-Raz et al., No No 2008 Bacillus C and DL (IA); GL Stougaard et al., Yumoto et al., Yumoto et al., Nejad et al., 2006 Gilbert et al., (FT); all lakes 2002 2003 2003 Chapter 3 2004 Chapter 3 Buttiauxella GL (FT) Amarita et al., NA3 NA3 No Chapter 3 1995 Chryseobacterium GL (IA); Bai et al., 2006 Park et al., 2008 Chaudhari et al., No Walker et al., DL (FT) 2010 2006 Chapter 2, 3 Cryobacterium Fly Mosier et al., Mosier et al., Dastager et al., No No 2007 2007 2009 Flavobacterium GL (IA) Mosier et al., Mosier et al., Qvit-Raz et al., Chapter 2 Kawahara et al., 2007 2007 2008 2001 Gillisia Ranch Neiderberger et Neiderberger et Aislabie et al., No No al., 2010 al., 2010 2006 Halomonas E83; Liberty; Neiderberger et Neiderberger et Wani et al., 2006 No Gilbert et al., Spotted al., 2010 al., 2010 2004 Herbaspirillum Fly NA3 Kirchhof et al., NA3 No No 2001 Idiomarina Liberty Gilbert et al., Gilbert et al., Navarro et al., No Gilbert et al., 2004 2004 2009 2004 Leifsonia Fly Mosier et al., Mosier et al., Dastager et al., No No 2007 2007 2009 Lysinibacillus Spotted Ahmed et al., Ahmed et al., Ahmed et al., No No 2007 2007 2007 Marinobacter RL Neiderberger et Neiderberger et Navarro et al., No No

164

al., 2010 al., 2010 2009

Microbacterium GL (IA) Miteva et al., Yatrebova et al., Dastager et al., No No 2004 2009 2009 Nesterenkonia E83; Liberty; Brinkmeyer et al., Qvit-Raz et al., Qvit-Raz et al., No No Spotted 2003 2008 2008 Paenibacillus C and GL (IA); DL Stougaard et al., Sokhansanj et al., Stougaard et al., No No (FT); Fly; Leeches 2002 2005 2002 Pantoea GL (FT) Selvakumar et al., NA3 NA3 Lindow et al., No 2008 1978 Pseudomonas DL and GL (IA); DL Lobova et al., Lobova et al., Wani et al., 2006 Green & Warren, Xu et al., 1998 (FT); Fly; Leeches; 2004 2004 1985 Chapter 3 E83; Ranch Chapter 2, 3 Roseinatronobacter Liberty NA3 Sorokin et al., Sorokin et al., Chapter 4 No 2000 2000 Rhodococcus Fly Steven et al., 2007 Plotnikova et al., Mertingk et al., No Duman & Olsen, 2006 1998 1993 Salengentibacter Spotted Brinkmeyer et al., NA3 NA3 No No 2003 Sphingomonas DL (IA); Leeches Brinkmeyer et al., Mosier et al., Qvit-Raz et al., Nejad et al., 2006 Gilbert et al., 2003 2007 2008 2004 Sporosarcina Fly Steven et al., 2007 NA3 NA3 No No Stenotrophomonas GL (IA) Mosier et al., Mosier et al., NA3 No Gilbert et al., 2007 2007 2004 Streptomyces DL (IA) Li et al., 2002 NA3 Dastager et al., No No 2008 Trichococcus Fly; Leeches Bai et al., 2006 NA3 NA3 No No 1Daring and Gould Lake soil samples are indicated as DL and GL, respectively. IA or FT denote ice-affinity or freeze-thaw selection, respectively. Fly, Leeches, East of 83 (E83), Liberty, TR Ranch (Ranch) and Spotted Lake-derived genera are also noted. 2 Association with the phenotype in question, either as characterized isolates or identified within the appropriate environment, is denoted by the respective reference.

165

3 NA indicates that the genus has not been found associated with the environmental condition, or that, to the best of my knowledge, a suitable reference is not available.

166

Most of these same soil samples were also subjected to freeze-thaw selection. As

mentioned above, freeze-thaw cycling has previously been shown to be detrimental to

microbial community structure and function. Similarly, here, as with ice-affinity, freeze-

thaw selection greatly decreased the microbial consortia abundance, enabling the

isolation of a few of microbes, many of which were also recovered from at least one of

the soil enrichments by ice-affinity. Also similar to those recovered post ice-affinity selection, a strikingly high proportion of the microbes recovered following freeze-thaw

demonstrated ice-association activities (a phenotype which was not observed in randomly

chosen isolates from the pre-selected consortia; Chapter 3). Surprisingly, the responses

of the temperate and boreal soil communities to freeze-thaw stress were rather similar.

Lastly, a higher proportion of microbes recovered post ice-affinity selection demonstrated

these protein activities than did those isolated post freeze-thaw selection. This

emphasizes the important role these activities likely play in microbial incorporation into

growing ice, and the possibility for community interactions amongst microbes exposed to

freezing.

Selecting for Microbes from Lake Environments – the Role of Cross-Tolerance:

Given that freezing and osmotic stresses exert similar physiological effects,

namely those associated with desiccation (Ko et al., 1994), there is the possibility of

cross-tolerance between these two stresses (Schmidt and Zink, 2000; Leblanc et al.,

2003; Schmid et al., 2009). Cross-tolerance was evaluated by exposing microbial

consortia from lakes with varying degrees of salinity (freshwater, brine and alkali lakes)

to freeze-thaw stress (Chapter 4; preliminary results can be found in Appendix A). 167

Consortia from more saline lakes tended to have a higher degree of resistance then those

from the freshwater lakes. The proportion of resistant isolates with ice-association

activity was remarkably low (~10%) compared with those isolated from soil (~63%),

indicating that these activities did not confer resistance. This may be because microbes

in temperate aquatic environments are not as likely to encounter ice, and therefore do not

express these activities. It might also indicate that these activities are not beneficial in

these environments.

Since ice-association activities are probably specific to low-temperature stress, the lack of these activities suggests that resistance was not conferred by low-temperature specific responses, thus supporting the possibility of cross-tolerance. As such, osmolyte concentration and biofilm formation, which are important for both low temperature and hyperosmotic stresses were evaluated. Microbes from the brine and alkali lakes tended to contain a higher solute concentration then those from the freshwater lakes, indicating that previous exposure to hypersaline environments, and thus a higher osmoprotectant concentration, increased resistance to freeze-thaw stress. Similarly, microbes from the brine and alkali lakes had a greater propensity for biofilm formation. Taken together then, these data strongly indicate that there was cross-tolerance between osmotic and freeze-thaw stresses.

Selecting for Microbes with Hydrate-Association Activity:

Freeze-thaw selection, at least with some environments, has been shown to select for microbes with ice-association activities, including those possibly resulting from

AFPs. Some AFPs have been shown to inhibit the growth of gas hydrates (Zeng et al., 168

2003; Zeng et al., 2006; Gordienko et al., 2010; Ohno et al., 2010). Hydrates are composed of stabilized water molecules and enclathrated gas molecules (Kvenvolden,

1999), and can form spontaneously in gas and oil pipelines with negative consequences

(Koh et al., 2002). As such, microbes enriched from produced water (crude oil and water mixture) were subjected to freeze-thaw stress in an attempt to recover microbes capable of hydrate growth inhibition and survival in oil well conditions (Chapter 5).

As with the previously discussed environments, selection greatly decreased microbial abundance and diversity. Selection also resulted in the initial recovery of microbes with ice-association (IRI) activity, possibly indicating that consortia members may have AFP activity. While the ability of these consortia to inhibit hydrate growth has yet to be assessed, the possible presence of microbes with such activities is promising for future industrial applications.

Recovered Isolates:

A number of isolates were recovered over the research projects discussed herein.

Of the 92 recovered isolates, 34 were subsequently characterized, as described in the relevant chapters. These 92 recovered isolates represent 27 genera, a large portion of which have previously been associated with cold or hyperosmotic environments and some of the characteristics assayed for here. As mentioned, Table 6.1 summarizes the relevant natural history of these genera, illustrating the remarkable reserve of strategies these microbes have acquired to survive such ‘harsh’ conditions.

169

Conclusions:

Microbial consortia enriched from a number of environments were subjected to

low temperature stress, either using ice-affinity or freeze-thaw selection. The phenotypes of the recovered isolates indicate that ice-association activities and sporulation seem to be

important factors conferring resistance in terrestrial environments. In consequence,

preliminary attempts were made to clone the putative antifreeze protein from one

recovered isolate (P. borealis DL7; Chapter 2) and from Chryseobacterium sp. C14, with

phage display. The results of this can be found in Appendix B. Conversely, ice-

association does not seem to be important in aquatic environments, wherein increased

osmolyte concentrations and biofilm formation seem to be the major factors enabling

resistance via cross-tolerance. Sporulation is also likely an important factor in aquatic

environments given the number of spore-formers that were isolated.

The difference in the role of ice-association activities between the terrestrial and

lake environments studied here may be that microbes within soils are more likely to

freeze then those in the temperate lakes. Thus, while cryoprotectants and biofilms are

important for soil microbes, ice-association activities may be more important. The

converse appears to be true of aquatic microbes.

Overall, microbial communities from three distinct environments (soil, lakes and

produced water) were subjected to stress in order to assess stress resistance mechanisms.

The possibility of finding ‘extremophiles’ in moderate environments, and the possibility

of using low temperature stress to select for microbes with industrially relevant

170

properties; such as ice- and/or hydrate-association activities was also successfully investigated.

Future Directions: Effects of Freeze-Thaw Stress on Soil Microbial Communities: Pyrosequencing of the

consortia prior to selection is currently underway to thoroughly assess the effects of this

stress on microbial community diversity. This will also enable a more thorough analysis

of the effects of ice-affinity selection on the microbial communities since the same soil

samples, with one exception, were separately enriched for each selective regime.

Cross-Tolerance Between Osmotic and Freeze-Thaw Stresses (Lake Samples):

Metabolomics could be used to further assess the effects of freeze-thaw stress on

osmolyte accumulation, including which osmolytes (salts, sugars and amino acids) were

affected and to what degree. Denaturing gradient gel electrophoresis (DGGE) analysis to

estimate the effects of freeze-thaw stress on microbial diversity would be informative,

however, pyrosequencing on one lake sample from each lake type is currently underway

to more thoroughly assess microbial diversity prior to and following freeze-thaw

selection.

Searching for Hydrate Inhibitors: The effects of the post-freeze-thaw selected microbial

consortia on hydrate growth will be tested (by a collaborator), using tetrahydrofuran, a

model hydrate. Deep sequencing of the consortia prior to and following selection is

underway. Additional future directions could include the isolation of the survivors and 171

analyses such as biofilm or aggregate formation and osmolyte content could be done to

further investigate why it is that these microbes may have survived. Finally, given the symbiosis between sulfate reducing bacteria and archaea, the archaeal content of the produced water samples could be assessed. This could be done by DGGE, sequencing, or by quantitative polymerase chain reaction (QPCR) of the genes involved in methane oxidation.

For all projects discussed herein, the role of microbial interactions could be studied. Finally, the role of cold induced proteins could be assessed using microarrays or reverse transcriptase, quantitative polymerase chain reaction (RT-QPCR) analysis.

If these on going and proposed studies were completed, a complete story would unfold. We would know the culturable diversity within these environments (at the time of sampling), and the full effects of selection on that diversity. We would also know with a greater degree of certainty how it is that microbes become incorporated into growing ice and, potentially, gas hydrates, and how they resist freeze-thaw stress. The latter may include yet unknown strategies, proteins or protein functions, metabolites or community interactions.

172

Literature Cited:

Ahmed, I., Yokota, A., Yamazoe, A., and Fujiwara, T. (2007). Proposal of Lysinibacillus boronitolerans gen. nov. sp. nov., and transfer of Bacillus fusiformis to Lysinibacillus fusiformis comb. nov. and Bacillus sphaericus to Lysinibacillus sphaericus comb. nov. Int. J. Sys. Evol. Microb. 57: 1117-1125.

Aislabie, J.M., Chhour, K.-L., Saul, D.J., Miyauchi, S., Ayton, J., Paetzold, R.F., and Balks, M.R. (2006). Dominant bacteria in soils of Marble Point and Wright Valley, Victoria Land, Antarctica. Soil Biol. Biochem. 38: 3041 -3056.

Amarita, F., Alkorta, F., Lescan du Plessix, M., Cantabrana, T., Rodriguez-Fernandez, C. (1995). Isolation and properties of free and immobilized beta-galactosidase from the psychrotrophic enterobacterium Buttiauxella agrestis (strain NC4). J. Appl. Bacteriol. 78: 630-635.

Bai, Y., Yang, D., Wang, J., Xu., S., Wang, X., An, L. (2006). Phylogenetic diversity of culturable bacteria from alpine permafrost in the Tianshan Mountains, northwestern China. Res. Microb. 157: 741-751.

Chattopadhyay, M.K., and Jagannadham, M.V. (2001). Maintenance of membrane fluidity in Antarctic bacteria. Polar Biol. 24: 386-388.

Chaudhari, P.N., Chaudhari, B.L., and Chincholkar, S.B. (2010). Cholesterol biotransformation to androsta-1,4-diene-3,17-dione by growing cells of Chryseobacterium gleum. Biotechnol. Lett. 32: 695-699.

173

Cochet, N., and Widehem, P. (2000). Ice crystallization by Pseudomonas syringae. Appl. Microbiol. Biotechnol. 54: 153-161.

Dastager, S.G., Dayanand, A., Li, W.-J., Kim, C.-J., Lee, J.-C., Park, D.-J., Tian, X.-P., and Raziuddin, Q.S. (2008). Proteolytic activity from an alkali-thermotolerant Streptomyces gulbargensis sp. nov. Curr. Microbiol. 57: 638-642.

Dastager, S.G., Pandey, A., Lee, J.-C., Li, W.-J., and Kim, C.-J. (2009). Polyphasic taxonomy of novel Actinobacteria showing macromolecule degradation potentials in Bigeum Island, Korea. Curr. Microbiol. 59: 21-29.

Davidson, D.W., El-Defrawy, M.D., Fulgem, M.O., and Judge, A.S. (1978). In ‘Proceedings of the 3rd International Conference on Permafrost’, National Research Council of Canada 1: 938–943.

Duman, J.G., and Olsen, T.M. (1993). Thermal hysteresis protein activity in bacteria, fungi, and phylogenetically diverse plants. Cryobiology 30: 322-328.

Gilbert, J.A., Hill, P.J., Dodd, C.E.R., and Laybourn-Parry, J. (2004). Demonstration of antifreeze protein activity in Antarctic lake bacteria. Microbiology 150: 171-180.

Gomez Zavaglia, A., Disalvo, E.A., and De Antoni, G.L. (2000). Fatty acid composition and freeze-thaw resistance in lactobacilli. J. Dairy Res. 67: 241-247.

Gordienko, R., Ohno, H., Singh, V.K., Jia, Z., Ripmeester, J.A., and Walker, V.K. (2010). Towards a green hydrate inhibitor: imaging antifreeze proteins on clathrates. PLoS ONE 5: e8953.

Green, R.L., and Warren, G.J. (1985). Physical and functional repetition in a bacterial ice nucleation gene. Nature 317: 645-648.

Kawahara, H., Li, J., Griffith, M. and Glick, B.R. (2001). Relationship between antifreeze protein and freezing resistance in Pseudomonas putida GR12-2. Current Microbiol. 43: 365-370.

Kirchhof, G., Eckert, B., Stoffels, M., Baldani, J.I., Reis, V.M., and Hartmann, A. (2001). Herbaspirillum frisingense sp. nov., a new nitrogen-fixing bacterial species that occurs in C4-fibre plants. Int. J. Sys. Evol. Microb. 51: 157-168.

174

Ko, R., Smith, L.T., and Smith, G.M. (1994). Glycine betaine confers enhanced osmotolerance and cryotolerance on Listeria monocytogenes. J. Bacteriol. 176: 426-431.

Koh, C.A., Westacott, R.E., Zhang, W., Hirachand, K., Creek, J.L. and Soper, A.K. (2002). Mechanisms of gas hydrate formation and inhibition. Fluid Phase Equilibr. 194- 197: 143-151.

Kvenvolden, K.A. (1999). Potential effects of gas hydrate on human welfare. Proc. Natl. Acad. Sci. USA 96: 3420-3426.

Li, W.-J., Zhang, L.-P., Xu, P., Cui, X.-L., Lu, Z.-T., Xu, L.-H., and Jiang, C.-L. (2002). Streptomyces beijiangensis sp. nov., a psychrotolerant actinomycete isolated from soil in China. Int. J. Sys. Evol. Microb. 52: 1695-1699.

Lindow, S.E., Arny, D.C., and Upper, C.D. (1978). Erwinia herbicola: a bacterial ice nucleus active in increasing frost injury to corn. Phytopathology 68: 523-527.

Lobova, T.I., Listova, L.V., and Popva, L.Y. (2004). Distribution of heterotrophic bacteria in Lake Shira. Microbiology 73: 89-93.

Mertingk, H., Müller, R.H., and Babel, W. (1998). Etherolytic cleavage of 4-(2,4- dichlorophenoxy)butyric acid and 4-(4-chloro-2-methylphenoxy)butyric acid by species of Rhodococcus and Aureobacterium isolated from an alkaline environment. J. Basic Microb. 38: 257-267.

Meury, J. (1988). Glycine betaine reverses the effects of osmotic stress on DNA replication and cellular division in Escherichia coli. Arch. Microb. 149: 232-239.

Mosier, A.C., Murray, A.E., and Fritsen, C.H. (2007). Microbiota within the perennial ice cover of Lake Vida, Antarctica. FEMS Microb. Ecol. 59: 274-288.

175

Navarro, J.B., Moser, D.P., Flores, A., Ross, C., Rosen, M.R., Dong, H., Zhang, G., and Hedlund, B.P. (2009). Bacterial succession within an ephemeral hypereutrophic Mojave Desert playa lake. Microb. Ecol. 57: 307-320.

Nejad, P., and Ramstedt, M. (2006). Presence of quorum-sensing-mediated gene regulation in pathogenic ice-nucleation-active (INA) bacteria. World J. Microbiol. Biotechnol. 22: 1373-1375.

Ohno, H., Susilo, R., Gordienko, R., Ripmeester, J., and Walker, V.K., (2010). Interaction of antifreeze proteins with hydrocarbon hydrates. Chem. Eur. J. DOI: 10.1002/chem.200903201.

Park, S.C., Kim, M.S., Baik, K.S., Kim, E.M., Rhee, M.S., and Seong, C.N. (2008). Chryseobacterium aquifrigidense sp. nov., isolated from a water-cooling system. Int. J. Sys. Evol. Microb. 58: 607-611.

Plotnikova, E.G., Rybkina, D.O., Anan’ina, L.N., Yastrebova, O.V., and Demakov, V.A. (2006). Characteristics of microorganisms isolated from technogenic soils of the Kama region. Russ. J. Ecol. 37: 233-240.

Qvit-Raz, N., Jurkevitch, E., and Belkin, S. (2008). Drop-size soda lakes: transient microbial habitats on a salt-secreting desert tree. Genetics 178: 1615-1622.

Raymond, J.A., and DeVries, A.L. (1977). Adsorption inhibition as a mechanism of freezing resistance in polar fishes. Proc. Natl. Acad. Sci. USA 74: 2589-2593.

Reddy, G.S.N., Aggarwal, R.K., Matsumoto, G.I., and Shivaji, S. (2000). Arthrobacter flavus sp. nov., a psychrophilic bacterium isolated from a pond in McMurdo Dry Valley, Antarctica. Int. J. Sys. Evol. Microbiol. 50: 1553-1561.

Roth, W.G., Leckie, M.P., and Dietzler, D.N. (1985). Osmotic stress drastically inhibits active transport of carbohydrates by Escherichia coli. Biochem. Biophys. Res. Commun. 126: 434-441.

176

Schmidt, G., and Zink, R. (2000). Basic features of the stress response in three species of bifidobacteria: B. longum, B. adolescentis, and B. breve. Int. J. Food Microb. 55: 41-45.

Selvakumar, G., Kundu, S., Joshi, P., Nazim, S., Gupta, A.D., Mishra, P.K., and Gupta, H.S. (2008). Characterization of a cold-tolerant plant growth-promoting bacterium Pantoea dispersa 1A isolated from a sub-alpine soil in the North Western Indian Himalayas. World J. Microb. Biotechnol. 24: 955-960.

Sorokin, D.Y., Tourova, T.P., Kuznetsov, B.B., Bryantseva, I.A., and Gorlenko, V.M. (2000). Roseinatronobacter thiooxidans gen. nov., sp. nov., a new alkaliphilic aerobic bacteriochlorophyll α-containing bacterium isolated from a soda lake. Microbiology 69: 75-82.

Stougaard, P., Jørgensen, F., Johnsen, M.G., and Hansen, O.C. (2002). Microbial diversity in ikaite tufa columns: an alkaline, cold ecological niche in Greenland. Environ. Microb. 4: 487-493.

Walker, V.K., Palmer, G.R., and Voordouw, G. (2006). Freeze-thaw tolerance and clues to the winter survival of a soil community. Appl. Environ. Microb. 72: 1784-1792.

Wani, A.A., Surakasi, V.P., Siddharth, J., Raghavan R.G., Patole, M.S., Ranade, D., and Shouche, Y.S. (2006). Molecular analyses of microbial diversity associated with the Lonar soda lake in India: an impact crater in a basalt area. Res. Microb. 157: 928-937.

177

Wilson, S.L., and Walker, V.K. (2010). Selection of low-temperature resistance in bacteria and potential applications. Environ. Tech. 31: 943-956.

Yamahira, K., Hirota, K., Nakajima, K., Morita, N., Nodasaka, Y., and Yumoto, I. (2008). Acinetobacter sp. strain Ths, a novel psychrotolerant and alkalitolerant bacterium that utilizes hydrocarbon. Extremophiles 12: 729-734.

Yastrebova, O.V., Plotnikova, E.G., Anan’ina, L.N., and Demakov, V.A. (2009). Aerobic spore-forming bacteria from the region of salt mining. Russ. J. Ecol. 40: 516-521.

Yergeau, E., and Kowalchuk, G.A. (2008). Responses of Antarctic soil microbial communities and associated functions to temperature and freeze-thaw cycle frequency. Environ. Microb. 10: 2223-2235.

Yumoto, I., Yamaga, S., Sogabe, Y., Nodasaka, Y., Matsuyama, H., Nakajima, K., and Suemori, A. (2003). Bacillus krulwichiae sp. nov., a halotolerant obligate alkaliphile that utilizes benzoate and m-hydroxybenzoate. Int. J. Sys. Evol. Microb. 53: 1531-1536.

Zeng, H., Wilson, L.D., Walker, V.K., and Ripmeester, J.A. (2003). The inhibition of tetrahydrofuran clathrate-hydrate formation with antifreeze protein. Can. J. Phys. 81: 17- 24.

178

Appendix A: Preliminary Assessment for Cross-Tolerance Between Osmotic and

Freeze-Thaw Stresses.

Cultured microbial communities from sediments obtained from temperate climate lakes with varying degrees of salinity were subjected to freeze-thaw selection in order to ascertain the possibility of resistance being conferred by cross-tolerance between osmotic and freeze-thaw stress. Initial experiments were carried out with archived sediment samples which were kindly provided by Dr. B. Cumming (Queen’s University).

Sediment samples were collected from freshwater (Fly and Leeches Lakes), brine

(Liberty Lake), and alkali (TR Ranch Lake) lakes in the 1990’s, and were stored at 4°C.

These samples were enriched, subjected to freeze-thaw cycling and survivors were isolated as monocultures. A subset of resistant isolates were putatively identified on the basis of 16S rDNA sequence analysis. Sample treatment and microbe identification protocols were the same as those described in Chapter 4.

The results of these preliminary experiments indicated that freeze-thaw resistance was greater for the microbial consortia from brine and alkali lakes than those from the freshwater lakes (Fig. A1). Sequencing of a small portion of the recovered isolates from

Leeches (freshwater), Liberty (brine) and TR Ranch (alkali) Lakes also indicated a shift in the resistant microbes towards those previously affiliated with halotolerance (Table

A.1; Fig. A.2). These results were suggestive of resistance to freeze-thaw stress being conferred by cross-tolerance and fresh samples (water and sediment) were collected to

179 further test this notion. The results of the experiments with fresh samples can be found in

Chapter 4.

180

1.00E+09

1.00E+08

1.00E+07

1.00E+06 0 Cycle 1.00E+05 Fly Sed. Leeches Sed. 1.00E+04 Liberty Sed. TR Ranch Sed.

CFU/mL (Avg.) 1.00E+03 Chryseo. sp. C14 E. coli 1.00E+02

1.00E+01

1.00E+00

1.00E-01

Figure A.1. Average colony forming units (CFU) per mL of the consortia versus 48 freeze-thaw cycles. These are compared to the positive and negative controls, Chryseobacterium sp. C14 and E. coli TG-2, respectively. ‘0 Cycle’ indicates a normalized starting density of 1x108 CFU/mL.

181

Table A.1. Lake consortia richness following freeze-thaw.

Sample Nearest Phylogenetic Accession Number N (% Relative (Sample Identifier)1 Similarity) Freshwater: Leeches Bacillus sp. (LS11) FN395284.1 1030 (100) Lake: Paenibacillus wynnii (LS9) AJ633647.1 1127 (992)

Brine: Liberty Bacillus sp. (LSA) FJ764772.1 1254 (99) Lake: Bacillus sp. (LSD) GQ404472.1 1226 (99) Halomonas sp. (LSU) GU228483.1 1178 (99) Idiomarina sp. (LSC) EF554872.1 1208 (99)

Alkali: TR Ranch Bacillus alkalitelluris (RS1) AY829448.1 1246 (95) Lake: Bacillus sp. (RS12) HM045841.1 1234 (99) Bacillus sp. (RS21) FJ764772.1 1177 (99) Bacillus sp. (RS25a) FJ607059.1 1180 (100) Pseudidiomarina sp. (RS6) GQ180187.1 1248 (96) Pseudomonas sp. (RS25) DQ480134.1 1098 (99)

1Each ‘redundant’ isolate per genus is <98% similar, therefore, each isolate reported above is unique. 296% query coverage, all others 100%.

182

Figure A.2. Phylogenetic tree representing partial consortia richness following freeze- thaw selection, constructed using the dnadist, neighbour and drawgram programs within the PHYLIP software package (Felsenstein, 1993). The lake sample from which the microbe was recovered is indicated in Table 1.

183

Appendix B: The Use of Phage Display for Isolating and Cloning Putative-AFPs from Chryseobacterium sp. C14 and Pseudomonas borealis

Chryseobacterium sp. C14 and Pseudomonas borealis both demonstrate AFP activity, the former exhibits IRI, while the latter has both IRI and ice-shaping activities.

Phage display was used to attempt to isolate, with the intent of subsequent cloning, the putative AFP from each of these bacteria.

The T7 Select 10-3b phage display system (Novagen T7 Select phage display;

Madison, WI, USA) was used as per the manufacturers suggestions. An AFP from a perennial rye grass, Lolium perenne (LpAFP) was the positive control, and restriction enzyme digested genomic DNA from each of these microbes was shotgun cloned into the phage vector. Ice-affinity purification was used to select for recombinant phages containing inserts capable of directly associating with ice, from within the clone libraries.

Phages were isolated following the ice-affinity purification and the inserts were screened using PCR with T7 select primers and/or LpAFP primers, as appropriate.

PCR amplification of the library constructed with the LpAFP gene, post ice- affinity selection, yielded fragments that were considerably different than the expected size (Fig. B1), and the inserts were inconsistently amplifiable. None of the clones screened contained an inserted LpAFP. PCR amplification of phages from the P. borealis genomic DNA clone libraries were also inconsistently amplifiable and rather small. Likewise, PCR amplification of phages isolated from the Chryseobacterium sp.

184

C14 library were far too small (~100bp). Overall, it appeared as though none of the screened phages from any of the constructed libraries contained an insert that either was, or could be used to isolate the respective putative AFP.

185

10 kb

1 kb 750 bp 500 bp 250 bp

Figure B.1. A typical screen of phages from the LpAFP clone library. Each phage containing an insert, as indicated by PCR products, had two products (<500 bp and ~750 bp -1 kb). The expected insert size was 540bp.

186