Cellular and Molecular Mechanisms of Environmental Stress Tolerance in Insects

Home , Alaskozetes antarcticus, Belgica (midge), Belgica antarctica, Cold hardening, Cryptobiosis, Fat body

DISSERTATION

Presented in Partial Fulfillment of the Requirements for the Degree Doctor of Philosophy in the Graduate School of The Ohio State University

Nicholas M. Teets

Graduate Program in Entomology

The Ohio State University

2012

Dissertation Committee:

Professor David L. Denlinger, Advisor

Professor Thomas G. Wilson

Professor P. Larry Phelan

Professor Sandor Gyorke

Professor Peter Piermarini

Nicholas M. Teets

2012

Abstract

The ability to tolerate environmental stress is a key adaptation for insects across the world. As ectotherms, insects are unable to regulate internal temperature, and their small body size makes them particularly susceptible to extremes in temperature and water availability. Insects rely on numerous physiological adaptations to cope with environmental stress, and recent advances in molecular biology and “omics” technologies have made it possible to study these mechanisms in detail. In this dissertation, I explore molecular mechanisms of stress tolerance in both temperate and polar insects.

In a process called rapid cold-hardening (RCH), brief exposure (i.e. minutes to hours) to nonlethal low temperatures significantly enhances tolerance to cold-shock conditions. While the ecological context of RCH is well-established, the underlying mechanisms are poorly understood. Using cDNA microarrays, we measured changes in gene expression accompanying RCH (2 h at 0°C) in the flesh fly, Sarcophaga bullata. To our surprise, no transcripts were differentially expressed during RCH, suggesting RCH occurs without the need for new gene products. Rather, cold-sensing and RCH appear to be primarily governed by second messenger systems, including calcium signaling. In

Chapter 3, we show that chilling evoked an increase in intracellular calcium and activated calcium/calmodulin-dependent protein kinase II. Blocking calcium signaling

ii pharmacologically prevented RCH, indicating calcium signaling is required during cold- sensing and RCH.

In the latter 4 chapters of this dissertation, I investigated physiological and molecular mechanisms of stress tolerance in the Antarctic midge, Belgica antarctica, the world’s southernmost insect and the only insect endemic to the continent. In the unpredictable climate of Antarctica, larvae are likely exposed to multiple bouts of both cold and desiccation stress, thus I quantified the survival and energetic consequences of repeated cold and dehydration exposure in B. antarctica. Larvae exposed to five diurnal freeze-thaw cycles experienced significant mortality and energy depletion. However, this was only true if larvae were frozen during repeated cold exposure; supercooled larvae exposed to the same temperatures experienced no significant mortality or energy depletion. Repeated bouts of dehydration and rehydration were also energetically costly, as 5 cycles of dehydration and rehydration caused larvae to consume 67% of their carbohydrate energy reserves.

In the final two chapters, I explored transcriptional mechanisms of extreme stress tolerance in B. antarctica. Targeted qPCR experiments revealed significant restructuring of metabolic gene expression during periods of stress. Cold stress caused upregulation of genes involved in glucose mobilization, while dehydration stress increased expression of genes required for glucose, trehalose, and proline synthesis. Finally, using RNA-seq, I measured changes in gene expression accompanying extreme dehydration in larvae of B. antarctica. Expression results identified upregulation of pathways involved in cellular recycling and energy conservation, such as ubiquitin-mediated proteasome and

iii autophagy, with concurrent downregulation of numerous genes involved in central metabolism.

Dedication

Dedicated to my wife and best friend, Julie.

Acknowledgments

I thank my advisor, Dr. David Denlinger, for his support and guidance during my graduate career. I also thank past and present members of the Denlinger lab for their camaraderie and assistance, particularly Qirui Zhang, Julie Reynolds, Alena Kobelkova,

Megan Meuti, and Justin Peyton, and Marie Bontempo, who have been here for all or most of my time at Ohio State. I am indebted to Rob Michaud for teaching me the basics of molecular biology, and I thank several undergraduate researchers, particularly Charles

Dean, for their assistance.

I thank Richard Lee, my undergraduate advisor, for his continued support and mentoring. I also thank Yuta Kawarasaki, Juanita Constible, and the Palmer Station support staff for two great field seasons in Antarctica.

I acknowledge my committee members, Thomas Wilson, Sandor Gyorke, Larry

Phelan, and Peter Piermarini for their contributions to this dissertation.

I thank various collaborators who have made substantial contributions to the results presented in this dissertation. Collaborators, with the chapters they contributed to in parentheses, include Justin Peyton (2,7), Herve Colinet (2,7), David Renault (2,7),

Greg Ragland (2), Dan Hahn (2), Shu-Xia Yi (3), Richard Lee (3-7), Yuta Kawarasaki (4-

7), and Joanna Kelley (7). Additional contributions are listed in the

“Acknowledgements” section of each chapter.

Many thanks to the College of Biological Sciences and the Department of

Entomology for financial support.

Finally, thanks to family and friends for their support and encouragement these past 5 years. Specifically, I thank my wife Julie for providing welcome distractions at home, my parents, Darcy and Kaye, for encouraging me to pursue my dreams, and my brothers, Tom, Joe, and Jacob for keeping things interesting.

vii

Vita

June 13, 1984 ...... Born, Lorain, OH

2003...... Marion L. Steele High School, Amherst, OH

2003-2004 ...... Laboratory Assistant, Department of Zoology, Miami University

2005-2007 ...... Undergraduate Research Assistant, Department of Zoology, Miami University

2007 ...... B.S. in Zoology, Miami University

2007-2012 ...... Graduate Research Fellow, Department of Entomology, Ohio State University

2009-2011 (intermittent) ...... Graduate Teaching Assistant, Department of Entomology, Ohio State University

Publications

Teets, N.M., Peyton, J.T., Colinet, H., Renault, D., Kelley, J.L., Kawarasaki, Y., Lee, R.E., Denlinger, D.L., 2012. Gene expression changes governing extreme dehydration tolerance in an Antarctic insect. Proceedings of the National Academy of Sciences of the United States of America, DOI 10.1073/pnas.1218661109

Teets, N.M., Kawarasaki, Y., Lee, R.E., Denlinger, D.L., 2012. Expression of genes involved in energy mobilization and osmoprotectant synthesis during thermal and desiccation stress in the Antarctic midge, Belgica antarctica. Journal of Comparative Physiology. B, Biochemical, Systemic, and Environmental Physiology, DOI 10.1007/s00360-012-0707-2.

viii

Teets, N.M., Peyton, J.T., Ragland, G.J., Colinet, H., Renault, D., Hahn, D.A., Denlinger, D.L., 2012. Uncovering molecular mechanisms of cold tolerance in a temperate flesh fly using a combined transcriptomic and metabolomic approach. Physiological Genomics, 44, 764-777.

Teets, N.M., Kawarasaki, Y., Lee, R.E., Denlinger, D.L., 2012a. Energetic consequences of repeated and prolonged dehydration in the Antarctic midge, Belgica antarctica. Journal of Insect Physiology, 58, 498-505.

Goto, S.G., Philip, B.N., Teets, N.M., Kawarasaki, Y., Lee, R.E., Denlinger, D.L., 2011. Functional characterization of an aquaporin in the Antarctic midge Belgica antarctica. Journal of Insect Physiology, 57, 1106-1114.

Teets, N.M., Kawarasaki, Y., Lee, R.E., Denlinger, D.L., 2011. Survival and energetic costs of repeated cold exposure in the Antarctic midge, Belgica antarctica: a comparison between frozen and supercooled larvae. Journal of Experimental Biology, 214, 806-814.

Michaud, M.R., Teets, N.M., Peyton, J.T., Blobner, B.M., Denlinger, D.L., 2011. Heat shock response to hypoxia and its attenuation during recovery in the flesh fly, Sarcophaga crassipalpis. Journal of Insect Physiology, 57, 203-210.

Benoit, J.B., Lopez-Martinez, G., Teets, N.M., Phillips, S.A., Denlinger, D.L., 2009. Responses of the bed bug, Cimex lectularius, to temperature extremes and dehydration: levels of tolerance, rapid cold hardening and expression of heat shock proteins. Medical and Veterinary Entomology, 23, 418-425.

Teets, N.M., Elnitsky, M.A., Benoit, J.B., Lopez-Martinez, G., Denlinger, D.L., Lee, R.E., 2008. Rapid cold-hardening in larvae of the Antarctic midge Belgica antarctica: cellular cold-sensing and a role for calcium. American Journal of Physiology. Regulatory, Integrative and Comparative Physiology, 294, R1938-R1946.

Fields of Study

Major Field: Entomology

Table of Contents

Abstract ...... ii Dedication ...... v Acknowledgments...... vi Vita ...... viii Publications ...... viii Fields of Study ...... ix Table of Contents ...... x List of Tables ...... xiii List of Figures ...... xv Chapter 1: Introduction ...... 1 Overview ...... 1 Mechanisms of Cold Hardening and Cold Shock Recovery ...... 2 Stress Tolerance of Antarctic Arthropods ...... 19 References ...... 34 Figures ...... 49 Chapter 2: Combined Transcriptomic and Metabolomic Approach Uncovers Molecular Mechanisms of Cold Tolerance in a Temperate Flesh Fly ...... 50 Abstract ...... 50 Introduction ...... 51 Materials and Methods ...... 54 Results and Discussion ...... 59 Conclusions ...... 71 Acknowledgements ...... 72 Grants ...... 73 References ...... 73 Tables ...... 80 Figures ...... 89 Chapter 3: Calcium Signaling Mediates Insect Cold Sensing ...... 95 Abstract ...... 95 Introduction ...... 96 Materials and Methods ...... 98 Results and Discussion ...... 102

Acknowledgements ...... 106 Grants ...... 106 References ...... 107 Figures ...... 110 Chapter 4: Survival and Energetic Costs of Repeated Cold Exposure in the Antarctic Midge, Belgica antarctica: A Comparison Between Frozen and Supercooled Larvae .. 120 Abstract ...... 120 Introduction ...... 121 Materials and Methods ...... 124 Results ...... 129 Discussion ...... 134 Conclusions ...... 142 Acknowledgements ...... 143 Grants ...... 143 References ...... 144 Figures ...... 149 Chapter 5: Energetic Consequences of Repeated and Prolonged Dehydration in the Antarctic Midge, Belgica antarctica...... 153 Abstract ...... 153 Introduction ...... 154 Materials and Methods ...... 157 Results ...... 163 Discussion ...... 168 Acknowledgements ...... 173 Grants ...... 174 References ...... 174 Tables ...... 178 Figures ...... 180 Chapter 6: Expression of Genes Involved in Energy Mobilization and Osmoprotectant Synthesis During Thermal and Dehydration Stress in the Antarctic Midge, Belgica antarctica ...... 184 Abstract ...... 184 Introduction ...... 185 Materials and Methods ...... 188 Results ...... 193 Discussion ...... 197 Conclusions ...... 204 Acknowledgements ...... 205 Grants ...... 205

References ...... 205 Tables ...... 211 Figures ...... 213 Chapter 7: Gene Expression Changes Governing Extreme Dehydration Tolerance in an Antarctic Insect ...... 219 Abstract ...... 219 Introduction ...... 220 Materials and Methods ...... 222 Results and Discussion ...... 229 Acknowledgements ...... 240 Grants ...... 240 References ...... 240 Tables ...... 245 Figures ...... 252 Chapter 8: Major Findings and Conclusions ...... 258 Part 1: Cellular and Molecular Mechanisms of Rapid Cold-Hardening ...... 258 Part 2: Mechanisms of Stress Tolerance in B. antarctica ...... 262 References ...... 267 Bibliography ...... 269

xii

List of Tables

Table 2.1 – Number of differentially expressed probes in each pairwise comparison. .... 80

Table 2.2. DAVID enrichment analysis of the differentially expressed ESTs in the Control vs. CS+2R comparison...... 81

Table 2.3. DAVID enrichment analysis of the differentially expressed ESTs in the Control vs. RCH+CS+2R comparison...... 82

Table 2.4. Gene set analysis of genes involved in recovery from cold shock...... 83

Table 2.5. Gene set analysis of genes enriched in the C vs. RCH+CS+2R comparison. . 84

Table 2.6. Expression of genes involved in Drosophila cold stress during recovery from cold shock in S. bullata...... 85

Table 2.7. Metabolite content in response to RCH and cold shock...... 86

Table 2.8. Metabolic pathways modulated by RCH...... 87

Table 2.9. Summary of biochemical pathways enriched in both the transcriptomic and metabolomic datasets during recovery from cold shock...... 88

Table 5.1. Effect of prolonged dehydration (10 d at 99% RH) and rehydration (1 d at 100% RH) on water content, osmolality, survival, and midgut cell survival in larvae of B. antarctica...... 178

Table 5.2. Effects of repeated bouts of dehydration and rehydration on whole-animal and midgut cell survival in larvae of B. antarctica...... 179

Table 6.1. Bioinformatics analysis of the 11 metabolic genes profiled in this study. .... 211

Table 6.2. Primers used for qPCR gene expression assays...... 212

Table 7.1. Summary of read statistics from Illumina sequencing...... 245

Table 7.2. Primers used for qPCR validation...... 246

xiii

Table 7.3. GO enrichment analysis of (A) upregulated and (B) downregulated genes upregulated in response to cryoprotective dehydration...... 247

Table 7.4. Gene set analysis (GSA) revealing enriched KEGG pathways during cryoprotective dehydration...... 248

Table 7.5. GO enrichment analysis of genes more highly expressed in the cryoprotective dehydration group relative to the desiccation group...... 249

Table 7.6. GO enrichment analysis of genes (A) upregulated or (B) downregulated in response to desiccation...... 250

Table 7.7. Gene set analysis (GSA) revealing enriched KEGG pathways during desiccation...... 251

xiv

List of Figures

Figure 1.1. Current model of the various cell signaling pathways involved in the cellular response to low temperature in insects...... 49

Figure 2.1. Temperature treatments (A-C) and hybridization design (D) for microarray experiment...... 89

Figure 2.2. Effect of rapid cold-hardening (RCH) on the cold tolerance of pharateadult flesh flies...... 90

Figure 2.3. Heat map showing expression patterns of the top 150 most differentially expressed ESTs...... 91

Figure 2.4. Principal components analysis (A) and hierarchical clustering (B) of the entire microarray dataset...... 92

Figure 2.5. Relative changes in metabolite contents in response to RCH and cold shock...... 93

Figure 2.6. Heat map diagram (A) and principal components analysis (B) of the entire metabolomics dataset...... 94

Figure 3.1. Rapid cold-hardening reduces freezing injury in tissues of the goldenrod gall fly, E. solidaginis and activates calcium signaling pathways...... 110

Figure 3.2. Predicted amino acid sequence and tissue-specific expression of E. solidaginis calmodulin...... 111

Figure 3.3. Predicted amino acid sequence and tissue-specific expression of E. solidaginis calcium/calmodulin dependent protein kinase II...... 112

Figure 3.4. Pharmacological inhibition of calcium signaling blocks rapid cold-hardening in tissues of E. solidaginis...... 113

Figure 3.5. Additional cell viability assays...... 114

Figure 3.6. Chilling also evokes an increase in intracellular calcium in tissues of the flesh fly, S. bullata...... 115

Figure 3.8. Developmental and tissue-specific expression of calmodulin in the flesh fly S. bullata...... 117

Figure 3.9. Developmental and tissue-specific expression of calcium/calmodulin- dependent protein kinase II (CaMKII) in the flesh fly, S. bullata...... 118

Figure 3.10. Working model for the role of calcium signaling during cold-sensing and rapid cold-hardening...... 119

Figure 4.1. Effect of RCE on larval survival (a) and cell survival (b) of B. antarctica larvae...... 149

Figure 4.2. Lipid (a), glycogen (b), trehalose (c), glucose (d), and glycerol (e) content of B. antarctica larvae during RCE...... 150

Figure 4.3. Total energy content (a) and carbohydrate energy content (b) due to the major energy stores of B. antarctica larvae during RCE...... 151

Figure. 4.4. mRNA expression of hsp70 during RCE in B. antarctica larvae...... 152

Figure 5.1. Water content (A) and osmolality (B) of larvae of B. antarctica during repeated bouts of dehydration and rehydration...... 180

Figure 5.2. Lipid (A), triglyceride (B), glycogen (C), trehalose (D), glucose (E), and glycerol (F) content in larvae of B. antarctica during repeated dehydration/rehydration cycles...... 181

Figure 5.3. Effect of prolonged dehydration on metabolite (A) and energy content (B) of larvae of B. antarctica...... 182

Figure 5.4. Effect of repeated dehydration on the total energy content (TEC) (A) and carbohydrate energy content (CEC) (B) of larvae of B. antarctica...... 183

Figure 6.1. Pathway diagram illustrating the biochemical reactions catalyzed by the genes examined in this study...... 213

Figure 6.2. Gene expression changes during thermal stress...... 214

Figure 6.3. Summary of gene expression patterns during thermal stress...... 215

Figure 6.4. Gene expression changes during dehydration stress...... 216

Figure 6.5. Summary of gene expression patterns during dehydration stress...... 217

xvi

Figure 6.6. Principal components analysis of thermal stress (A) and dehydration stress (B) gene expression datasets...... 218

Figure 7.1. Expression summary (A), dendrogram (B) and Venn diagram (C) showing degree of similarity between the desiccation (D) and cryoprotective dehydration (CD) groups...... 252

Figure 7.2. Results of qPCR validation experiment...... 253

Figure 7.3. Pathway diagrams illustrating upregulation of autophagy-related genes (A) and downregulation of carbohydrate metabolism and ATP synthesis (B)...... 254

Figure 7.4. Changes in metabolite content in response to desiccation and cryoprotective dehydration...... 255

Figure 7.5. Hierarchical clustering of the metabolomics dataset...... 256

xvii

Chapter 1: Introduction

Overview

For temperate and polar insects, low temperature poses a number of physiological challenges. As ectotherms, insects’ body temperature closely tracks environmental temperature, leading to significant perturbations in function (Denlinger and Lee, 1998).

Low temperature damages cell membranes and denatures proteins, causing ion imbalances and enzyme malfunction (Lee, 2010). Maintaining water balance is also a significant challenge in the winter; cold winter air holds very little moisture, and water consumption typically ceases in dormant, overwintering insects (Danks, 2000).

Furthermore, evaporative water loss is high in insects, due to their small body size and consequent high surface area to volume ratio (Gibbs et al., 2003). Despite these obstacles, insects thrive on every continent and are the most abundant and diverse terrestrial organisms, with roughly a million species having been described (Mayhew, 2007).

The study of insect low temperature biology began over 50 years ago, and ever since physiologists have been uncovering physiological mechanisms of cold tolerance.

Recent advances in molecular biology and “omics” technologies have extended upon the earlier work of basic physiologists. However, until recently, detailed molecular studies were restricted to model organisms such as Drosophila melanogaster, which are often subpar models for stress physiology (Hoffmann, 2010). In this dissertation, I will explore molecular and physiological mechanisms of stress tolerance in three non-model species 1

of insects, the flesh fly, Sarcophaga bullata, the goldenrod gall fly, Eurosta solidaginis, and the Antarctic midge, Belgica antarctica. There are essentially two main thrusts to this dissertation. In Chapters 2 and 3, I will explore rapid responses to low temperature in temperate insects at the cellular and molecular level. In Chapters 4-7, I will investigate mechanisms of stress tolerance in Antarctica’s only endemic insects, B. antarctica, with experiments ranging from basic ecophysiological principles to high throughput molecular studies. The rest of this chapter contains a review of our current knowledge of these two topics.

Mechanisms of Cold Hardening and Cold Shock Recovery

Basic principles of insect cold tolerance

Insects have evolved diverse strategies for coping with low temperature, and experts are continually redefining and re-categorizing these strategies. In general, most insects can be classified as either freeze-avoidant or freeze-tolerant (Hawes and Bale,

2007). Freeze-avoidant insects, which make up the majority of insects categorized thus far, succumb to internal ice formation and therefore must remain in a supercooled state at subzero temperatures. The temperature at which ice forms is the supercooling point, and this point is dependent on a number of factors, including body size and the presence or absence of ice-nucleating agents (Duman et al., 1991). In contrast, freeze-tolerant insects can survive internal ice formation, provided ice is restricted to the extracellular space.

Typically, freeze-tolerant insects have relatively high supercooling points, to prevent

mechanical damage from rapid, uncontrolled ice formation at extreme low temperatures

(Lee, 2010).

The factors that determine whether a particular species is freeze-avoiding or freeze-tolerant are largely unknown. Freeze-avoidance appears to be basal in arthropods, as a majority of freeze-tolerant insects are found in higher orders, Diptera, Coleoptera,

Lepidoptera, and Hymenoptera (Sinclair et al., 2003a). However, freeze-tolerance appears to be an example of convergent evolution, as the capacity for freeze-tolerance has evolved numerous times. For example, even within drosophilid flies, there are examples of both freeze-avoiding and freeze-tolerant species (Kostal et al., 2011). Sinclair et al.

(2003a; 2005) postulate that hemispheric differences in climatic variability may drive the evolution of freeze tolerance. In the southern hemisphere, where temperature and moisture levels tend to be more variable, freeze-tolerance is much more prevalent.

However, the sample size for southern hemisphere insects is much smaller and restricted to a few geographic locations, so additional studies are needed to test this hypothesis. An alternative hypothesis is that cold tolerance strategies are driven by water balance characteristics (Zachariassen et al., 2008). Freeze-tolerance may be favored by insects with high cuticular water permeability, as freezing would be a means of retaining water during the winter.

Mechanisms of cold injury

Depending on the severity and duration of cold exposure, insects experience several levels of cold injury. The first detrimental effect of cold is the onset of chill coma, which occurs when neuromuscular function is impaired due to an inability to regulate ion

balance (MacMillan and Sinclair, 2011). While chill coma is reversible, if these conditions continue over an extended period (days to weeks), insects are at risk of indirect chilling injury. Indirect chilling injury is a result of impaired membrane function, which causes prolonged ion imbalance and energy deficits due to an inability to sustain central metabolism (Dollo et al., 2010). At more severe temperatures, insects succumb to direct chilling injury, also known as cold shock injury (Lee, 2010). While the exact mechanisms of cold shock injury are still being sorted out, it is largely due to membrane damage resulting from phase transition from a liquid crystalline to a gel state (Steponkus,

1984). Additionally, cold causes actin depolymerization, leading to reorganization of the actin cytoskeleton at low temperatures (Kim et al., 2006). Finally, cold shock directly denatures proteins, thereby inhibiting protein function and causing harmful aggregations of denatured proteins (Feder and Hofmann, 1999). When freezing occurs, additional injury can occur as a result of freeze concentration (Lee, 2010). Since only water molecules join the growing ice lattice, extracellular solutes become concentrated, driving water out of the cell and causing cell shrinkage. Growing ice crystals can also cause physical damage to cells by fracturing cell membranes (Tursman and Duman, 1995).

Seasonal adaptations to low temperature

To combat the above challenges posed by low temperature, insects have accumulated a number of adaptations to enhance their tolerance to subzero temperatures.

The best described of these adaptations are mechanisms of seasonal cold acclimation. A majority of temperate insects enter into overwintering diapause, an environmentally programmed period of dormancy characterized by developmental arrest and reduced oxygen consumption rates (Denlinger et al., 2005). Diapause can occur at any 4

developmental stage, although a given species has a specific stage at which it enters diapause. Diapause is typically programmed by decreasing photoperiods in advance of winter, and while there are exceptions, diapause is generally associated with enhanced environmental stress tolerance (Denlinger, 1991). For example, in the flesh fly,

Sarcophaga crassipalpis diapausing pupae readily survive for 7 days at -17°C, whereas non-diapausing pupae succumb to the same conditions in as little as 20 minutes (Lee and

Denlinger, 1985).

Insects can also enhance seasonal cold tolerance by means of cold acclimation.

Cold acclimation is defined as an enhancement of cold tolerance following prolonged

(i.e. days to weeks) exposure to mild, sublethal low temperatures (Colinet et al., 2012a).

For example, non-diapausing D. melanogaster raised at 19°C are significantly more cold- tolerant than those raised at 25°C (Rako and Hoffmann, 2006). Cold acclimation is also frequently observed under natural conditions; larvae of Eurosta solidaginis are unable to survive freezing at -20° in early autumn, but in response to gradually decreasing temperatures in the fall, larvae are fully freeze tolerant at -20°C by the onset of winter

(Williams et al., 2004).

Seasonal cold acclimation involves a number of physiological adaptations, which protect cells from cold damage and help maintain function at low temperature. The earliest, and perhaps best described, mechanism of seasonal cold acclimation is the accumulation of low-molecular weight osmoprotectants (Lee, 2010). The most common osmoprotectants observed in overwintering insects are small sugar alcohols, such as glycerol, sorbitol, and inositol. Functions of these compounds include stabilizing

membranes and proteins, enhancing supercooling capacity, and preventing osmotic damage to cells (Yancey, 2005). However, the types and amounts of osmoprotective polyols vary significantly by species; E. solidaginis accumulates roughly equal amounts of glycerol and sorbitol (Storey et al., 1981), while the closely related Drosophila montana synthesizes large amounts of myo-inositol as the primary osmoprotectant

(Vesala et al., 2012a). In addition, other classes of compounds are being recognized as osmoprotectants. Trehalose, the blood sugar of insect, is accumulated in a number of desiccation-tolerant overwintering insects (Elnitsky et al., 2008b; Mitsumasu et al.,

2010), while the amino acid proline is the primary osmoprotectant in the freeze-tolerant drosophilid Chymomyza costata (Kostal et al., 2011). The identification of novel cryoprotectants has been facilitated by recent advances in metabolomics. For example, cold acclimation in D. melanogaster results in elevation of several amino acids, sugars, and polyols, although the adaptive significance of these changes remains to be seen

(Colinet et al., 2012a).

Cell membranes are particularly susceptible to cold, and consequently a number of cold acclimation mechanisms appear to involve membrane modifications. A hypothesis termed homeoviscous adaptation proposes that organisms adjust membrane composition at low temperatures to maintain a consistent level of membrane fluidity

(Kostal, 2010). In support of this, cold or diapause-induced changes in membrane fatty acid composition have been observed in numerous species of insects (Bennett et al.,

1997; Kostal et al., 2003; Kostal and Simek, 1998; Michaud and Denlinger, 2006;

Overgaard et al., 2005; Pruitt and Lu, 2008). Also, cholesterol may play a role in insect

cold acclimation, as membrane bound cholesterol can prevent coalescing of adjacent fatty acyl chains (Shreve et al., 2007).

At the molecular level, a number of genes and proteins have been implicated as key components of diapause and cold acclimation. Perhaps most conspicuous are the heat shock proteins, molecular chaperones that refold denatured proteins during periods of stress (Feder and Hofmann, 1999). Heat shock proteins are upregulated during diapause in a number of insects, including egg diapausing silk moths (Moribe et al., 2010), pupal diapausing flesh flies (Rinehart et al., 2000a; Yocum et al., 1998), and adult diapausing

Colorado potato beetles (Yocum, 2001). Knocking down heat shock proteins during diapause reduces cold tolerance, indicating a crucial role for these genes in diapause- associated thermoprotection (Rinehart et al., 2007). However, there are instances where diapause-associated cold hardening is not associated with increase heat shock protein expression, as in the mosquito Culex pipiens (Rinehart et al., 2006b).

In addition to heat shock proteins, targeted studies have revealed other classes of genes involved in diapause and cold acclimation responses. Aquaporins, water channels that promote water movement across cell membranes, appear to be essential for freeze- tolerant insects. In E. solidaginis, aquaporins increase in abundance in the weeks leading to winter (Philip and Lee, 2010), and blocking aquaporins with mercuric chloride reduces tissue freezing tolerance (Philip et al., 2008). In Drosophila, senescence marker protein-

30, a calcium binding protein, is elevated during cold hardening, perhaps to maintain intracellular calcium concentrations at low temperature (Goto, 2000). Finally, the onion maggot, Delia antiqua, upregulates a desaturate gene in response to cold acclimation,

which likely contributes to membrane restructuring in preparation for winter (Kayukawa et al., 2007).

While the above targeted studies have yielded several important candidate genes involved in diapause and seasonal cold hardening, recent transcriptomics and proteomics experiments have revealed many additional players. Using proteomics, Li et al. (2007) identified 80 proteins that were differentially regulated in brains of diapausing flesh fly pupae. While a number of these were heat shock proteins, there were a number of additional proteins that were either unidentified or have unknown function. At the mRNA levels, changes are even more conspicuous. In flesh flies, >50% of the entire transcriptome is differentially expressed during diapause (Ragland et al., 2010), including an abundance of stress related genes, metabolic genes, and cell signaling genes. While the exact physiological role of these genes during diapause remains to be seen, these results indicate that the regulation of diapause is a complex process that involves more than just a few candidate genes. Transcriptomic and proteomic studies in other overwintering species have yielded a similar picture, namely that genes across numerous functional categories and signaling pathways participate in the diapause phenotype (Colinet et al.,

2012b; Ragland et al., 2011). While there are some common themes, such as the role of heat shock proteins, and molecular adjustments in carbohydrate metabolism, a number of diapause-induced gene expression changes appear to be species specific. Thus, insects are capable of using disparate molecular mechanisms to produce a dormant, overwintering phenotype.

Rapid cold hardening – significance and mechanisms

While seasonal responses to low temperature have been well-studied, insects are also capable of responding to low temperature on a much shorter time scale. In a process called rapid cold-hardening (RCH), insects are capable of significantly increasing cold tolerance after brief (i.e. minutes to hours) exposure to nonlethal low temperature, typically around 0°C for temperate insects (Lee et al., 1987). For example, in pharate adult flesh flies, as little as 30 min at 0°C significantly enhances survival following a 2 h cold shock at -10°C. In addition to rapid induction, there is also rapid attenuation, as the protection afforded by RCH is quickly lost upon warming (Coulson and Bale, 1990).

RCH has since been described in dozens of insect species across numerous taxonomic groups (reviewed by Lee and Denlinger, 2010). Also, the ecological relevance of this phenomenon has been well-established. RCH can be induced by ecologically relevant cooling rates and thermoperiods (Kelty and Lee, 1999, 2001), while field-collected insects are significantly more cold tolerant during cool parts of the day than those sampled during warm periods of the day (Kelty, 2007; Koveos, 2001). Furthermore, RCH preserves ecologically important functions such as courtship and mating (Rinehart et al.,

2000b; Shreve et al., 2004).

Since the discovery of RCH, physiologists have been attempting to uncover the mechanisms behind it. As a starting point, many of the mechanisms associated with seasonal cold acclimation were examined in the context of RCH. Indeed, some of the

“usual suspects” appear to be at play during RCH. In response to RCH, pharate adults of

S. crassipalpis synthesize glycerol, increasing concentrations nearly 4-fold to ~80 mM

(Chen et al., 1987). While this concentration of glycerol is modest compared to some 9

overwintering insects, later experiments demonstrated that injection of small amounts of glycerol can cause immediate increases in cold tolerance (Yoder et al., 2006). Recent studies using metabolomics have provided further evidence that metabolic adjustments are a key component of RCH. In S. crassipalpis, metabolomics confirmed the accumulation of glycerol during RCH and also revealed accumulation of 2 amino acids

(alanine and glutamine), an additional polyol (sorbitol), and a sugar (glucose) (Michaud and Denlinger, 2007). Also, accumulation of pyruvate indicates increased rates of glycolysis during RCH, which may provide energy and substrates for downstream physiological functions. Similar experiments in D. melanogaster likewise indicated

RCH-induced glucose accumulation, suggesting rapid mobilization of glucose may be a common theme during RCH (Overgaard et al., 2007).

As with seasonal acclimation, the cell membrane is also a target for RCH. RCH causes elevation of oleic acid in the cell membrane (Michaud and Denlinger, 2006), which leads to a measurable increase in membrane fluidity at low temperature (Lee et al.,

2006a). Additional chemical modifications to phospholipid head groups in response to

RCH also enhance membrane fluidity (Michaud and Denlinger, 2006). A similar phenomenon occurs in D. melanogaster, although in this species it is linoleic acid rather than oleic acid that is responsible for increased membrane fluidity (Overgaard et al.,

2005). Membrane cholesterol may also be important during RCH; augmentation of cholesterol in the diet not only enhances baseline cold tolerance, it also increases the capacity to undergo RCH (Shreve et al., 2007). Nonetheless, there have been cases where

RCH is observed in the absence of detectable changes in membrane composition, indicating that other mechanisms are at play (MacMillan et al., 2009). 10

At the molecular level, heat shock proteins are an obvious candidate for conferring cold tolerance during RCH. During heat shock, transcription of heat shock protein mRNAs begins immediately, so it is conceivable something similar could be at play during RCH. However, to date, there is no evidence that RCH upregulates transcription of heat shock proteins. Sinclair et al. (2007) measured expression of two different heat shock protein genes during RCH, but neither was upregulated during the hardening period. While transcription of heat shock protein mRNA doesn’t appear to be involved in RCH, results at the protein level are more promising. Using proteomics, Li et al. identified a single heat shock protein (hsp23) that is upregulated in the brains of flesh flies that have undergone RCH (Li and Denlinger, 2008). Thus, it appears upregulation of hsp23 is occurring primarily via regulation of translation, rather than transcription.

Expression studies with other genes have further confirmed the hypothesis that transcription is not a major mechanism of RCH. In the same study measuring heat shock protein expression, Sinclair et al. (2007) measured expression of three additional cold- related candidate genes, and none were differentially expressed during RCH. A candidate-gene microarray study containing 219 diapause-associated genes in Drosophila montana identified a proline-metabolic gene upregulated during RCH, but targeted qPCR experiments failed to confirm this result (Vesala et al., 2012b). Qin et al. (2005) measured the expression of 7000 cDNAs in response to cold hardening, but they gave flies a 30 minute recovery period after hardening, making it impossible to conclude that any observed expression differences occurred during RCH. One obvious conclusion from these studies is that RCH simply doesn’t require the transcription of new genes; given the speed (as fast as 30 minutes) and low temperature at which RCH occurs, perhaps this is 11

to be expected. Nonetheless, in plants, several genes are upregulated within minutes of cold exposure (Thomashow, 1999), leaving open the possibility that yet-to-be-identified genes may be differentially expressed during RCH. Recent advances in transcriptomics will make it possible to determine which, if any, genes are differentially expressed during

RCH.

Recently, it was observed that isolated tissues of S. crassipalpis are capable of

RCH ex vivo (Yi and Lee, 2004), and the same is true for a freeze-tolerant species,

Belgica antarctica (Teets et al., 2008). This suggests that the brain and neurohormones are not required for RCH, but rather, cell-mediated signaling pathways are capable of inducing RCH. This was an important development, because it directed focus to second messenger pathways as key mediators of RCH. To date, three signaling pathways have been implicated in the RCH response: MAP kinase signaling, apoptosis signaling, and calcium signaling. MAP kinases are a group of serine/threonine protein kinases that are activated by diverse cell stimuli, most notably stress and immune signals (Pearson et al.,

2001). In S. crassipalpis, a specific MAP kinase, p38 MAP kinase, is phosphorylated immediately upon transfer to RCH-inducing conditions (Fujiwara and Denlinger, 2007).

Activation of p38 is maximal at 0°C (the optimal temperature for RCH) and is independent of the brain, making it a strong candidate as a mediator of RCH. However, the upstream signals that activate p38 during RCH are unknown, as are the downstream targets and effectors.

One deleterious effect of cold shock is activation of apoptotic cell death pathways

(Yi and Lee, 2011; Yi et al., 2007). While the underlying mechanisms of cold-induced

apoptosis are unknown, it may be a consequence of calcium imbalances and mitochondrial misfuction, both of which can trigger apoptosis (Kroemer et al., 1998). In

S. crassipalpis and D. melanogaster, RCH inhibits cold-induced apoptosis (Yi and Lee,

2011; Yi et al., 2007), permitting cell survival during otherwise lethal conditions. The mechanism appears to involve inhibition of caspase activity as well as accumulation of an anti-apoptotic protein, bcl-2. However, as with p38, the upstream mechanisms that target the apoptosis pathway during RCH are unidentified.

A third signaling pathway that has been implicated in RCH is the calcium signaling pathway (Teets et al., 2008). Isolated tissue of B. antarctica loaded with either the calcium chelator BAPTA or the calmodulin inhibitor W-7 fail to undergo RCH ex vivo. Thus, both intracellular calcium and activation of calmodulin appear to be essential for RCH. Indirect evidence for a role of calcium during cold hardening is provided from experiments on D. melanogaster mutants of an extracellular matrix protein, dystroglycan

(Takeuchi et al., 2009). Mutants lacking functional dystroglycan have elevated levels of intracellular calcium, which leads to enhanced cold tolerance and a preference for lower temperatures. Also, transient receptor protential channels in specific neurons of larval D. melanogaster directly respond to low temperature by conducting an inward calcium current (Rosenzweig et al., 2008), although the function of these calcium channels during

RCH or in peripheral tissues has not been addressed.

While Teets et al. (2008) clearly demonstrated a function for calcium during RCH in B. antarctica, a number of questions still remain. Namely, what is the mechanism of calcium entry during RCH, and what are the downstream targets of calcium signaling

during RCH? Clues for the mechanism of calcium entry during RCH can be found in the plant literature. In plants, cold triggers membrane rigidification and destabilization of actin, which activate calcium channels on the cell membrane (Orvar et al., 2000). Since calcium is kept at higher concentrations in extracellular spaces, opening of these channels allows calcium to enter the cell and trigger downstream cold acclimation processes.

Intracellular calcium stores in the endoplasmic reticulum also contribute to cold-induced calcium flux in plants (Knight et al., 1996). Once inside the cell, calcium activates a number of cold-specific protein kinases (Monroy et al., 1997), leading to upregulation of cold acclimation specific genes (Monroy and Dhindsa, 1995; Monroy et al., 1993). Other hypothetical roles for calcium during RCH include: 1) Stimulation of mitochondrial energy metabolism at low temperatures by activating mitochondrial dehydrogenase

(Denton, 2009; Takeuchi et al., 2009). 2) Activation of glycogen phosphorylase kinase, which in turn could activate glycogen phosphorylase and stimulate glycogenolysis and osmoprotectant mobilization (Johnson, 1992). 3) Cross-talk with MAP kinase (Takeda et al., 2004) and apoptosis pathways (Yano et al., 1998) to promote cell survival. 4)

Trafficking and/or activation of proteins such as aquaporins (Chou et al., 2000), which may need to be rapidly activated during cold exposure.

The above studies indicate that RCH is a dynamic process involving numerous physiological pathways, many of which have yet to be discovered. A current model of our understanding of cold-sensing and RCH signal transduction is presented in Figure

1.1. A major challenge for physiologists is to connect these pieces of the puzzle and determine which processes are vital to producing the RCH phenotype. However, recent advances in transgenics and molecular biology should make it possible to completely 14

unravel the RCH signaling pathway, from perception of cold by the cell to the effector molecules that are responsible for the observed cold hardening.

Mechanisms of cold shock recovery

Rapid cold hardening is one example of how insects quickly respond to changes in temperature. Another instance where insects rapidly respond to changes in temperature is during recovery from cold shock. The multitude of cellular injuries resulting from cold shock (discussed above) must be repaired upon return to permissible temperatures. While only a handful of studies have addressed the mechanisms of cold repair in insects, some themes are beginning to emerge.

Following a cold insult, insects show an increase in metabolic rate immediately upon return to permissible temperatures (Block et al., 1998; Lalouette et al., 2011). This increase in aerobic metabolism is necessary to replenish cellular ATP stores, since oxidative phosphorylation cannot be maintained at low temperature. Indeed, shortly after removal from cold, there is a detectable increase in ATP levels (Dollo et al., 2010). This increase in cellular respiration is necessary to support cold repair processes, many of which are energetically expensive. For example, to regain neuromuscular function following cold shock, ion gradients must be restored, which requires activation of a number of ATP-dependent ion pumps (Kostal et al., 2007). Furthermore, elevated metabolism is required to support the burst of mRNA and protein synthesis that occurs during recovery from cold shock (discussed below).

At the molecular level, a number of genes and proteins have been identified that are important for recovery from cold injury. Early on, it was discovered that there is a

burst of protein synthesis during recovery from cold shock (Joplin et al., 1990), and advances in molecular biology have made it possible to determine the exact identity of some of these genes. Once again, the heat shock proteins are one of the most conspicuous players during recovery from cold shock. While very little, if any, heat shock protein transcription occurs at subzero temperatures, upon return to permissible temperatures, several heat shock proteins are strongly upregulated during recovery (e.g. Colinet et al.,

2010c; Rinehart and Denlinger, 2000; Rinehart et al., 2006b; Rinehart et al., 2000a;

Sinclair et al., 2007). Heat shock proteins likely function to bind and repair proteins that are denatured by cold. Knocking down heat shock proteins with RNAi during recovery from cold shock significantly reduces survival (Colinet et al., 2010b), indicating these genes play a key role in the recovery process.

Aside from heat shock proteins, several other genes are induced during recovery from cold. In D. melanogaster, the gene frost is upregulated during recovery from cold

(Goto, 2001). While this gene was originally thought to be specific to cold stress, it has since been found to be upregulated during dehydration as well (Sinclair et al., 2007).

Nonetheless, frost is responsive to cold in numerous life stages and species of

Drosophila, indicating it plays an important role during cold stress in drosophilid flies

(Bing et al., 2012). Additionally, as with heat shock proteins, knocking down frost expression with RNAi impairs recovery from cold stress (Colinet et al., 2010a). However, frost does not appear to have any orthologs outside of Drosophila, so it is likely a specific cold-recovery mechanism for this genus. Also, it does not contain any conserved functional domains, so the mechanism of action of frost is uncertain at this time.

The identification of genes involved in recovery from cold stress has been aided by advances in genomics and transcriptomics. To date, there have only been two transcriptomic studies of cold recovery in insects, and both have been conducted in D. melanogaster. Using microarrays, Qin et al. (2005) identified 36 genes that were differentially expressed after 30 minutes of mild cold stress. These experiments confirmed the role of heat shock proteins and frost during cold shock recovery, and they also identified a number of new candidate genes, including several membrane proteins and genes involved in mitochondrial metabolism. A more recent study by Zhang et al.

(2011) used a longer recovery period (6 h), and found several genes related to immune function that were upregulated. However, since D. melanogaster is a tropically-adapted species that is considerably less cold-hardy than temperate species, additional studies in other species are needed to fully appreciate transcriptomic responses during cold shock recovery. At the protein level, a proteomics study of fluctuating thermal regimes in the wasp Aphidus colemani identified many proteins that were upregulated during recovery from cold exposure, including heat shock proteins, proteins involved in aerobic metabolism and ATP synthesis, and proteins involved in cytoskeletal reorganization.

Future directions

Like other branches of physiology, insect cold hardening research has benefitted from advances in molecular biology and “omics” technology. However, until recently, large-scale molecular studies were restricted to model organisms such as D. melanogaster. Now, with next-generation sequencing technology, it is possible to conduct large-scale “omics” experiments on any insect species. Several transcriptomic studies related to overwintering have now been conducted (e.g. Clark et al., 2009; 17

Ragland et al., 2010; Ragland et al., 2011), but many more studies need to be conducted to fully grasp the physiological mechanisms of cold tolerance across all insect taxa. Also, studies that combine two or more of the “omics” technologies (i.e. genomics, transcriptomics, proteomics, and metabolomics) could benefit cold tolerance research, to highlight the interplay between gene expression and downstream physiological and metabolic changes. However, this isn’t to say that targeted physiological studies should be abandoned; rather, “omics” approaches can inform targeted studies by highlighting genes and pathways that are important for cold hardening.

While considerable research has been conducted on seasonal acclimation and diapause, less attention has been focused on rapid responses to low temperature. For example, fundamental questions such as whether gene expression contributes to RCH have yet to be thoroughly addressed. Furthermore, the effect of RCH on downstream cold recovery processes is still a mystery. Finally, perhaps the biggest black box in RCH research is the cold sensing and signal transduction pathways that trigger RCH. Because

RCH occurs so quickly (i.e. on the order of minutes), it likely involves protein kinase cascades and other second messenger systems. Indeed, there is evidence that MAP kinase signaling (Fujiwara and Denlinger, 2007) and calcium signaling (Teets et al., 2008) are involved, but the picture is still very incomplete. In this regard, insect research is considerably behind the plant literature, where the cold sensing mechanisms and downstream signaling pathways have been well-defined (Mahajan and Tuteja, 2005).

Thus, considerable work is needed in insects to identify cold sensing mechanisms and determine how these pathways elicit downstream cold hardening responses.

Stress Tolerance of Antarctic Arthropods

Overview of Antarctic arthropods

While insects on all continents encounter some form of environmental stress, perhaps nowhere are the environmental onslaughts more challenging than the continent of Antarctica. Even in maritime Antarctica, where proximity to the sea limits temperature extremes, winter lows exceed -40°C, and subzero temperatures can be experienced any time of year (Baust and Lee, 1981). Furthermore, water is frozen and biologically unavailable for much of the year, thus water availability is perhaps the biggest challenge confronting terrestrial arthropods in Antarctica (Kennedy, 1993). Whereas arthropods are the predominant terrestrial life form on Earth, only a few species have successfully become established in Antarctica, and most of these are restricted to maritime regions

(Convey, 1996a).

The predominant arthropods in Antarctica are soil microarthropods, including mites and collembolans (Convey, 1996a). While a complete arthropod biodiversity inventory for Antarctica is lacking, a 1971 report listed 136 species of arthropods in

Antarctica (Gressitt, 1971). Of these, ~40% are parasitic ticks, mites, lice and fleas, while the free-living species are primarily composed of mites and collembolans. Perhaps most glaring is Antarctica’s lack of true insects; true insects native to Antarctica consist of two chironomid species, the midges Belgica antarctica and Parochlus steinenii (Convey and

Block, 1996). Since P. stenenii is also found in southern South America, B. antarctica is considered the only true insect endemic to Antarctica. A third Dipteran species,

Eretmoptera murphyi, was accidentally introduced to Signy Island in the maritime

Antarctica from subantarctic South Georgia in the 1960’s (Convey, 1992). Despite the 19

lack of species diversity in Antarctica, local abundance can be very high. For example, a mass of over 2 million collembolan eggs (from Cryptopygus antarcticus and Friesea grisea) was found under a single rock on the Antarctic peninsula in an area less than 0.1 m2 (Schulte et al., 2008).

Basic ecophysiological characteristics of Antarctic arthropods

While the life history adaptations of Antarctic invertebrates have been reviewed extensively (e.g. Convey, 1996a, 2010), they are worth mentioning here. Perhaps the most conspicuous life history adaption of Antarctic arthropods is lifespan extension. The low temperatures of Antarctica lead to extremely short growing seasons, so in many cases

Antarctic arthropods take multiple years to complete the life cycle (Convey, 1996a). For example, the Antarctic collembolan, Cryptopygus antarcticus takes an estimated 3-7 years to complete its life cycle (Burn, 1981), whereas temperate collembolans typically complete multiple generations per year. Similarly, the Antarctic midge Belgica antarctica has a 2-year life cycle, which is rare among the chironomids (Convey and Block, 1996).

In conjunction with long life cycles, Antarctic arthropods typically lack a true diapause and instead rely on quiescence to endure unfavorable periods (Convey, 1996b). This allows Antarctic species to take advantage of the extremely short, unpredictable growing seasons. However, the capacity for diapause has only been examined in non-insect taxa, which typically do not express diapause, even in temperate regions.

Aside from prolonged, flexible lifestyles, there are a couple other noteworthy features of Antarctic arthropods. While most representative species are from apterous taxa (i.e. mites and collembolans), of the three chironomid species on the continent, two

(B. antarctica and E. murphyi) are secondarily brachypterous (Convey and Block, 1996).

Brachyptery is a common adaptation in isolated, windswept regions, to prevent accidental dispersal into unfavorable environments and to conserve heat by reducing surface area to volume ratio. Because environmental stressors are thought to be the primary selective pressure in Antarctica, there is thought to be little inter- and intraspecific competition for resources. A majority of Antarctic arthropods are generalist herbivores and detritovores

(Davis, 1981), thus thriving on sporadic, inconsistent nutrient sources. A final noteworthy attribute of Antarctic arthropod life histories is the large investment of energy towards environmental stress protection. Many stress mechanisms (see below for examples) are energetically costly, thus diverting energy from growth and reproduction. For example, populations of the Antarctic mite Alaskozetes antarcticus from mild subantarctic climates are able to divert significantly more energy to reproduction than populations in maritime

Antarctica, presumably due to differential resource allocation towards environmental stress tolerance (Convey, 1998). However, for a majority of Antarctic species, the fitness costs of elevated stress tolerance have not been quantified.

Stress tolerance of Antarctic arthropods

As mentioned above, mechanisms of stress tolerance are a major physiological cost for Antarctic arthropods. Also, Antarctic arthropods are excellent models for the evolutionary physiology of stress tolerance, since so few species have become successfully established on the continent. Thus, the basic mechanisms of stress tolerance have been extensively studied in Antarctic arthropods, and recent studies have begun to unravel the molecular mechanisms of extreme stress tolerance.

With low temperature being the most conspicuous feature of Antarctic environments, the cold tolerance of Antarctic arthropods has been extensively studied.

The most primitive group, ticks and mites (Order: Acari), are all freeze-avoiding

(Somme, 1981) and rely on supercooling to survive subzero temperatures. For example, the Antarctic mite Alaskozetes arcticus has winter supercooling points around -30°C, which is accomplished in part by accumulation of glycerol and removal of ice-nucleating particles in the gut (Young and Block, 1980). Another mite, Styerotydeus mollis, has supercooling points around -20°C in early summer that increase over the course of the summer, likely due to feeding (Sjursen and Sinclair, 2002). While overwintering supercooling points have not been measured, they are assumed to be less than winter minimum microhabitat temperatures, which can reach -40°C. The seabird tick, Ixodes uriae, has a similar supercooling capacity as mites (~-30°C), but can also tolerate high temperatures (>25°C), perhaps to permit survival and activity both on- and off-host (Lee and Baust, 1987). In addition to seasonal regulation of supercooling points, summer- acclimated mites are also capable of RCH (Worland and Convey, 2001).

Like mites, all Antarctic collembolan studied thus far are freeze-avoiding.

Generally, collembolan have extensive supercooling capacities, with supercooling points below -35°C recorded in some species (Cannon and Block, 1988). Extensive supercooling in this group is achieved by a combination of extremely high hemolymph osmolality (Sinclair and Sjursen, 2001) and a moderate degree of thermal hysteresis

(Sinclair et al., 2006). Seasonal regulation of supercooling points in Antarctic collembolans appears to be primarily regulated by the molt cycle. Decreasing temperatures trigger progression into a non-feeding, pre-molt period that results in gut 22

clearance and decreased supercooling points (Worland and Convey, 2008). Like mites,

Antarctic collembolans are capable of rapidly increasing cold tolerance (Worland and

Convey, 2001). Here, gut clearance also plays a role, as rapid decreases in supercooling point are achieved primarily by gut clearance (Worland et al., 2000). In the field, diurnal variations in supercooling points can be observed as a result of daily rhythms of feeding and gut clearance; for example, in two species of Antarctic collembola, nighttime supercooling points are >10°C lower than those during the day (Sinclair et al., 2003b).

While some have referred to this phenomenon as “rapid cold hardening,” it appears to be distinct from rapid cold hardening sensu stricto, which involves cellular protection against the damaging effects of low temperature (see Lee and Denlinger, 2010).

Among the true insects on Antarctica, there are notably different mechanisms of overwintering cold tolerance. The midge P. steinenii is freeze-susceptible and has relatively modest supercooling capacity, with summer supercooling points around -15°C

(Shimada et al., 1991). Also, lower lethal temperatures of P. steinenii are only -3°C for summer larvae (and -9°C for pupae), making it significantly less cold hardy than other

Antarctic dipterans (see below). However, the winter cold hardiness of P. steinenii has not been examined. Nonetheless, as an aquatic midge that overwinters as larvae at the bottom of freshwater ponds and lakes, P. steinenii is likely buffered from extremes in temperature, perhaps explaining the modest level of cold tolerance. In contrast, the other two Antarctic insects, endemic B. antarctica and introduced E. murphyi, are both freeze- tolerant, and in fact are the only known freeze-tolerant invertebrates on the continent.

Larvae of both B. antarctica and E. murphyi are freeze-tolerant down to around -20°C

(Baust and Lee, 1987; Worland, 2010), which is considerably lower than the minimum 23

microhabitat temperatures experienced (Baust and Lee, 1981). In its native range, E. murphyi rarely experiences temperatures below -1.5°C, making this level of cold tolerance somewhat surprising, but this “pre-adaptation” to colder climates is what allowed this species to successfully establish in maritime Antarctica (Worland, 2010).

While larvae of both species undergo seasonal cold acclimation, it appears they remain freeze-tolerant year around (Baust and Edwards, 1979). Recently, RCH was described in

B. antarctica, which was the first report of RCH in a freeze-tolerant insect (Lee et al.,

2006b). Subsequently, RCH was characterized in E. murphyi (Everatt et al., 2012), suggesting that rapid enhancement of cold tolerance is a common adaption in Antarctic arthropods.

While low temperature is a significant stress limiting the range and fitness of

Antarctic arthropods, water availability is perhaps the biggest challenge for Antarctic arthropods. Water is frozen and therefore unavailable for much of the year, and inland areas receive very little annual precipitation (Kennedy, 1993). Also, due to their small body size and high surface area to volume ratio, arthropods as a whole are susceptible to water loss (Gibbs et al., 2003). Thus, not surprising, Antarctic arthropods are typically extremely tolerant of desiccating conditions, either with mechanisms to reduce water loss or mechanisms to tolerate a dehydrated state.

To survive desiccating conditions, arthropods either need to reduce water loss or be able to tolerate cellular dehydration. Among Antarctic mites, both strategies seem to be in play. For example, A. antarcticus relies on water conservation, with a thick coating of waterproofing hydrocarbons that limit evaporative water loss to the environment

(Benoit et al., 2008). In contrast, two predatory mites from the same habitat,

Hydrogamasellus antarcticus and Rhagidia gerlachei, have high transpiration rates and thus seek moist microhabitats to maintain water balance. However, how these two mites tolerate winter desiccation, when liquid water is unavailable, has not been examined.

Antarctic collembolans are typically found in moist environments and exhibit very little resistance to water loss. In a comparison of seven species of Antarctic mites and collembolans, Worland and Block (1986) found that collembolans had significantly higher rates of water loss than any of the mite species. At 5% RH and 0°C, collembolans lost anywhere from 9 to 25% of their body water per hour, depending on species.

Recently, the water balance of C. antarcticus was examined in more detail, revealing that this species is incapable of maintaining water balance at any relative humidity below saturation and therefore relies on liquid water to maintain water balance (Elnitsky et al.,

2008a). Since C. antarcticus is unable to prevent dehydration, it must be able to survive in severely dehydrated conditions. Indeed, more than 50% survived a 60% loss of body water at ecologically relevant conditions (Elnitsky et al., 2008a). Thus, in general, while

Antarctic collembolans have high water loss rates, they are capable of tolerating extreme desiccation, surviving in a near anhydrobiotic state during the Antarctic winter.

Like the collembolans, the Antarctic midges E. murphyi and B. antarctica lose water very rapidly but are extremely tolerant of dehydration. At ecologically relevant humidities, larvae of E. murphyi readily survive upwards of 50% water loss (Worland,

2010), while larvae of B. antarctica survive up to 70% loss of body water (Benoit et al.,

2007b; Hayward et al., 2007). In response to desiccation, larvae of B. antarctica combat

high water loss rates by behaviorally clustering and by increasing the waterproofing properties of their cuticle (Benoit et al., 2007b). Also, in B. antarctica, dehydration tolerance is restricted to the larval stage; adults, which are only active for a brief period during the summer, require moist habitats and succumb to mild dehydration (Benoit et al., 2007a). In addition to dehydration, Antarctic arthropods experience other forms of osmotic stress. For example, due to their proximity to sea, larvae of B. antarctica are also at risk of immersion in sea water, and as such larvae can survive up to 10 days in 1000 mOsm seawater (Elnitsky et al., 2009).

Recently, it has been discovered that many Antarctic arthropods are capable of a distinct form of dehydration termed cryoprotective dehydration. When arthropods with water-permeable cuticles are surrounded by environmental ice, a vapor pressure gradient draws water out of the body into the surrounding ice, thereby allowing the body fluid melting point to track environmental temperatures (Holmstrup et al., 2002). The prerequisites for cryoprotective dehydration are a permeable cuticle, extreme dehydration tolerance, and the ability to resist inoculative freezing. Among Antarctic species, cryoprotective dehydration has been demonstrated in the collembolan, C. antarcticus

(Elnitsky et al., 2008a; Worland and Block, 2003) and the midge, B. antarctica (Elnitsky et al., 2008b). Also, some non-arthropod soil dwelling organisms in the Antarctic are capable of cryoprotective dehydration, including nematodes (Wharton et al., 2005).

Cryoprotective dehydration is now considered by some to be a third overwintering strategy (along with freeze-tolerance and freeze-avoidance), since during cryoprotective dehydration arthropods are neither frozen nor supercooled, as their melting point matches the environmental temperature. 26

In the case of B. antarctica, the discovery of cryoprotective dehydration was somewhat surprising, since B. antarctica is freeze-tolerant. Also, B. antarctica was the first true insect in which cryoprotective dehydration was demonstrated. Cryoprotective dehydration was originally considered an adaptation for freeze-avoiding species, since it allows arthropods to remain unfrozen at subzero temperatures. Cryoprotective dehydration was described in a freeze-tolerant nematode (Wharton et al., 2003) and an enchytraeid worm (Pedersen and Holmstrup, 2003) but was considered a laboratory artifact, since neither of these species is likely able to avoid inoculative freezing in the field. While cryoprotective dehydration in B. antarctica has yet to be examined in a field setting, larvae are capable of avoiding inoculative freezing at ecologically relevant soil moisture levels (Elnitsky et al., 2008b). Thus, B. antarctica represents a unique case where both freeze tolerance and cryoprotective dehydration can be used. Furthermore, the supercooling point of B. antarctica larvae (-7°C) is lower than typical microhabitat temperatures, meaning that during brief subzero temperature exposures larvae are capable of remaining supercooled, provided larvae avoid inoculative freezing. However, it is uncertain which of the three strategies (freezing, supercooling, or cryoprotective dehydration) is employed in the field, and which is preferable from a survival and fitness standpoint.

Biochemical and molecular mechanisms of stress tolerance in Antarctic arthropods

In comparison to temperate insects, the molecular mechanisms of stress tolerance in Antarctic arthropods have received little attention. Nonetheless, early studies in the

1980’s characterized biochemical markers of environmental stress, and recent studies have capitalized on advances in molecular biology. Here, I will review the physiological 27

mechanisms of stress tolerance in Antarctic terrestrial arthropods and highlight some avenues for future research.

Like their temperate counterparts, seasonal and stress-induced accumulation of low-molecular-weight osmoprotectants is a hallmark of Antarctic arthropods. Every species profiled thus far accumulates some sort of osmoprotective compounds, although the type and amount vary from species to species. For example, the mite A. antarcticus primarily uses glycerol as an osmoprotectant, accumulating levels upwards of 0.5 M

(Block and Convey, 1995). Likewise, the collembolan C. antarcticus uses glycerol as the chief cryoprotectant and also accumulates glucose and trehalose in response to desiccation (Elnitsky et al., 2008a). In contrast, the midge B. antarctica accumulates very low levels of glycerol and instead relies primarily on glucose, erythritol, and trehalose as osmoprotectants (Baust and Lee, 1983; Lee and Baust, 1981). One theme that is emerging from studies of both polar and tropical desiccation-tolerant organisms is the importance of trehalose as an osmoprotectant during dehydration. Trehalose, the blood sugar of most arthropods, is the chief osmoprotectant in arthropods capable of anhydrobiosis (Clegg,

2001), and recent studies in C. antarcticus (Elnitsky et al., 2008a) and B. antarctica

(Benoit et al., 2007b; Elnitsky et al., 2008b) have implicated its importance during extreme dehydration in Antarctic arthropods as well.

In recent years, physiological studies in Antarctic arthropods has benefitted from advances in molecular biology and “omics” technology, although molecular experiments have only been conducted in two species, the collembolan C. antarcticus and the midge,

B. antarctica. Regardless, these studies are beginning to provide clues about the mechanisms of stress tolerance in some of the world’s most extreme arthropods.

Molecular experiments in C. antarcticus are restricted to two microarray studies of cold tolerance. In the first, animals with a “low” supercooling point (<-15°C) were compared to those with “high” supercooling points (>-15°C), to determine which genes are responsible for lowering supercooling points (Purac et al., 2008). This microarray contained a subset of 672 ESTs, and thus was not comprehensive. Nonetheless, expression patterns indicate upregulation of a number of cuticular proteins and other structural constituents in the “low” group, confirming the importance of the cuticle and molt cycle in regulating supercooling capacity in Antarctice collembolans (Worland and

Convey, 2008). Other genes upregulated in the “low” group relative to the “high” group include several mitochondrial genes involved in ATP synthesis, suggesting that boosting energy production may be a component of low temperature survival. A similar experiment was conducted later with a larger microarray (containing 5400 ESTs), and once again, a number of cuticular genes and genes involved in the molt cycle were upregulated in cold-acclimated collembolans (Burns et al., 2010).

Relative to other Antarctic arthropods, the Antarctic midge B. antarctica has by far the largest number of molecular studies associated with it. Like temperate insects, the heat shock proteins are important genes mediating stress tolerance in B. antarctica.

However, whereas most insects express heat shock proteins at very low levels until they are needed, larvae of B. antarctica constitutively express heat shock proteins at high levels all the time (Rinehart et al., 2006a). While adults of B. antarctica have a typical

heat shock protein response, neither heat nor cold increased expression of three different heat shock proteins (small hsp, hsp70, and hsp90) in larvae. This constant expression of heat shock proteins likely provides year-round protection against environmental stress, which can be frequent and unpredictable in maritime Antarctica. Whereas high expression of heat shock proteins typically hinders growth and development (Krebs and

Feder, 1997), larvae of B. antarctica are able to circumvent this and express heat shock proteins at high levels even while they are feeding and growing.

Constitutive defenses in B. antarctica are not restricted to heat shock proteins.

Likewise, larvae express the antioxidant enzyme superoxide dismutase at high levels even in the absence of oxidative stress (Lopez-Martinez et al., 2008). SOD mRNA levels modestly increase after exposure to sunlight, as well the antioxidant gene catalase and three heat shock proteins. Indeed, expression of these genes confers extremely high resistance to oxidative damage in B. antarctica, as the antioxidant capacity of B. antarctica larvae is 5X greater than that of a temperate freeze-tolerant insect, E. solidaginis (Lopez-Martinez et al., 2008). Adults of B. antarctica have even higher levels of antioxidant capacity, probably due to their near-constant exposure to sunlight as they walk on the surface in search for mates. Resistance to oxidative damage is critical for

Antarctic arthropods, as Antarctic sunlight contains very high levels of UV radiation

(Liao and Frederick, 2005), which is intensifying due to ozone damage (Weatherhead and

Andersen, 2006). Furthermore, repeated bouts of freeze-thaw exposure, which are very common in Antarctica, are known to cause oxidative damage in insects (Lalouette et al.,

2011).

Recently, aquaporins have been implicated as key regulators of water movement during stressful conditions such as dehydration (Liu et al., 2011) and freezing (Philip and

Lee, 2010; Philip et al., 2008). Aquaporins are pore-forming proteins that carry water, and sometimes other solutes, across the cell membrane (Borgnia et al., 1999). Since

Antarctic arthropods are challenged by numerous forms of osmotic stress, including freezing, dehydration, and immersion in sea water, aquaporins likely play an important role in mediating stress tolerance. Goto et al. (2011) cloned and characterized the first aquaporin from an Antarctic arthropod, an aquaporin-1 like gene from B. antarctica.

When expressed in Xenopus oocytes, this protein is capable of transmitting water, but not urea or glycerol, across the cell membrane. This specific aquaporin gene is expressed in several different tissues, indicating it may play a general role in water movement across cells. However, mRNA expression did not change in response to dehydration, so it is unclear what, if any, role this gene plays in mediating stress tolerance. A second study of

B. antarctica aquaporins found immunoreactivity to four different aquaporin antibodies from different species, and some of these were stress inducible (Yi et al., 2011).

However, the sequence identity of these aquaporin genes has not been established. In the same study, blocking aquaporins pharmacologically with mercuric chloride reduced the ex vivo freezing tolerance of fat body, midgut, and Malpighian tubules tissue, indicating that aquaporins are critical for water redistribution during freezing. Additionally, mercuric chloride reduced the water loss of midgut tissue, suggesting aquaporins also play a critical role in mediating dehydration stress.

Non-targeted, “omics” approaches have also benefitted our understanding of stress tolerance in B. antarctica. Using SSH, Lopez-Martinez et al. (2009) obtained a 31

number of dehydration-responsive clones, and Northern blots confirmed that 23 of these were indeed differentially expressed either during dehydration or rehydration.

Upregulated genes include three heat shock proteins (hsp26, hsp70, hsp90) and two antioxidant enzymes (superoxide dismutase and catalase), indicating that protein denaturation and oxidative damage are symptoms of dehydration stress. Other genes upregulated during dehydration include genes coding for cytoskeletal proteins and membrane restructuring, consistent with previous observations that dehydration causes cytoskeletal reorganization (Chen et al., 2005) and membrane lipid remodeling (Bayley et al., 2001). In addition, several genes are downregulated in response to dehydration, including two electron transport chain genes, suggesting a shutdown of metabolism during dehydration. At the protein level, a proteomics study of dehydration and rehydration in B. antarctica supported the idea that cytoskeletal and cell structural rearrangements are essential during dehydration (Li et al., 2009). Of the 18 proteins more abundant during dehydration, 13 are cell structural proteins, including several isoforms of actin and myosin. Interestingly, several contractile proteins are also less abundant during desiccation, indicating that certain contractile proteins are synthesized while others are degraded. Finally, Michaud et al. (2008) profiled metabolic adaptations to heat, freezing, and dehydration using non-targeted GC-MS metabolomics. Several metabolites show similar responses to cold and desiccation, including the osmoprotectants glycerol and erythritol, which may in part explain the cross-tolerance between these two stresses in B. antarctica (Hayward et al., 2007).

Conclusions and future directions

In contrast to the abundance of arthropods on other continents, the Antarctic arthropod community is depauperate and consists of only a handful of species. In the past

30 years, researchers have been intently studying the physiological ecology of these arthropods, and recent advances in molecular biology have fostered significant advances in our knowledge of the world’s most extreme arthropods. However, what is still lacking is an understanding of the unique adaptations that distinguish Antarctic arthropods from their tropical and temperate counterparts. Most adaptations described in Antarctic arthropods, such as accumulation of osmoprotectants (e.g. Baust and Lee, 1983) and the role of aquaporins during freeze tolerance (Yi et al., 2011) have been previously described in temperate species. That isn’t to say these studies aren’t important, only that unique Antarctic adaptations, if they exist, have been elusive to physiologists.

Discoveries such as the constitutively high expression of heat shock proteins in B. antarctica (Rinehart et al., 2006a) and the prevalence of cryoprotective dehydration in polar arthropods (Worland and Block, 2003) are promising starts, but there is still a ways to go.

One potential way to uncover unique Antarctic adaptations would be the use of comparative physiology. Basic ecophysiological studies, such as the cold tolerance of

Antarctic collembolans (Sinclair et al., 2003b; Sinclair et al., 2006) and the water balance of Antarctic mites (Benoit et al., 2008) have successfully used comparative approaches to elucidate similarities and differences between Antarctic species. However, to date there haven’t been any molecular studies in Antarctic arthropods that take advantage of a comparative design. For example, comparative physiological genomics of dehydration 33

responses across multiple Antarctic species could reveal critical, conserved molecular adaptations to water stress. Comparisons could be extended to closely related temperate species, to identify which molecular adaptations to stress are “Antarctic specific.” There are clear physiological and life history differences between Antarctic and temperate species (see Convey, 2010), and tools are now available to reveal the molecular underpinnings of these differences.

References

Baust, J.G., Edwards, J.S., 1979. Mechanisms of freezing tolerance in an antarctic midge, Belgica antarctica. Physiological Entomology, 4, 1-5.

Baust, J.G., Lee, R.E., 1981. Environmental "homeothermy" in an Antarctic insect. Antarctic Journal of the United States, 15, 170-172.

Baust, J.G., Lee, R.E., 1983. Population differences in antifreeze cryoprotectant accumulation patterns in an Antarctic insect. Oikos, 40, 120-124.

Baust, J.G., Lee, R.E., 1987. Multiple stress tolerance in an antarctic terrestrial arthropod: Belgica antarctica. Cryobiology, 24, 140-147.

Bayley, M., Petersen, S.O., Knigge, T., Kohler, H.R., Holmstrup, M., 2001. Drought acclimation confers cold tolerance in the soil collembolan Folsomia candida. Journal of Insect Physiology, 47, 1197-1204.

Bennett, V.A., Pruitt, N.L., Lee, R.E., 1997. Seasonal changes in fatty acid composition associated with cold-hardening in third instar larvae of Eurosta solidaginis. Journal of Comparative Physiology B-Biochemical Systemic and Environmental Physiology, 167, 249-255.

Benoit, J.B., Lopez-Martinez, G., Elnitsky, M.A., Lee, R.E., Denlinger, D.L., 2007a. Moist habitats are essential for adults of the Antarctic midge, Belgica antarctica (Diptera : Chironomidae), to avoid dehydration. European Journal of Entomology, 104, 9-14.

Benoit, J.B., Lopez-Martinez, G., Michaud, M.R., Elnitsky, M.A., Lee, R.E., Denlinger, D.L., 2007b. Mechanisms to reduce dehydration stress in larvae of the Antarctic midge, Belgica antarctica. Journal of Insect Physiology, 53, 656-667. 34

Benoit, J.B., Yoder, J.A., Lopez-Martinez, G., Elnitsky, M.A., Lee, R.E., Denlinger, D.L., 2008. Adaptations for the maintenance of water balance by three species of Antarctic mites. Polar Biology, 31, 539-547.

Bing, X., Zhang, J., Sinclair, B.J., 2012. A comparison of Frost expression among species and life stages of Drosophila. Insect Molecular Biology, 21, 31-39.

Block, W., Convey, P., 1995. The biology, life-cycle and ecophysiology of the Antarctic mite Alaskozetes antarcticus. Journal of Zoology, 236, 431-449.

Block, W., Worland, M.R., Bale, J., 1998. Respiratory responses to chilling and freezing in two sub-Antarctic insects. Cryobiology, 37, 163-166.

Borgnia, M., Nielsen, S., Engel, A., Agre, P., 1999. Cellular and molecular biology of the aquaporin water channels. Annual Review of Biochemistry, 68, 425-458.

Burn, A.J., 1981. Feeding and growth in the Antarctic collembolan Cryptopygus antarcticus. Oikos, 36, 59-64.

Burns, G., Thorne, M.A.S., Hillyard, G., Clark, M.S., Convey, P., Worland, M.R., 2010. Gene expression associated with changes in cold tolerance levels of the Antarctic springtail, Cryptopygus antarcticus. Insect Molecular Biology, 19, 113-120.

Cannon, R.J.C., Block, W., 1988. Cold tolerance of microarthropods. Biological Reviews of the Cambridge Philosophical Society, 63, 23-77.

Chen, C.P., Denlinger, D.L., Lee, R.E., 1987. Cold-shock injury and rapid cold hardening in the flesh fly Sarcophaga crassipalpis. Physiological Zoology, 60, 297-304.

Chen, S., Gollop, N., Glazer, I., 2005. Cross-stress tolerance and expression of stress- related proteins in osmotically desiccated entomopathogenic Steinernema feltiae IS-6. Parasitology, 131, 695-703.

Chou, C.L., Yip, K.P., Michea, L., Kador, K., Ferraris, J.D., Wade, J.B., Knepper, M.A., 2000. Regulation of aquaporin-2 trafficking by vasopressin in the renal collecting duct - Roles of ryanodine-sensitive Ca2+ stores and calmodulin. Journal of Biological Chemistry, 275, 36839-36846.

Clark, M.S., Thorne, M.A.S., Purac, J., Burns, G., Hillyard, G., Popovic, Z.D., Grubor- Lajsic, G., Worland, M.R., 2009. Surviving the cold: molecular analyses of insect cryoprotective dehydration in the Arctic springtail Megaphorura arctica (Tullberg). Bmc Genomics, 10, 328.

Clegg, J.S., 2001. Cryptobiosis - a peculiar state of biological organization. Comparative Biochemistry and Physiology B-Biochemistry & Molecular Biology, 128, 613-624.

Colinet, H., Larvor, V., Laparie, M., Renault, D., 2012a. Exploring the plastic response to cold acclimation through metabolomics. Functional Ecology, 26, 711-722.

Colinet, H., Lee, S.F., Hoffmann, A., 2010a. Functional characterization of the frost gene in Drosophila melanogaster: Importance for recovery from chill coma. Plos One, 5, 7.

Colinet, H., Lee, S.F., Hoffmann, A., 2010b. Knocking down expression of Hsp22 and Hsp23 by RNA interference affects recovery from chill coma in Drosophila melanogaster. Journal of Experimental Biology, 213, 4146-4150.

Colinet, H., Lee, S.F., Hoffmann, A., 2010c. Temporal expression of heat shock genes during cold stress and recovery from chill coma in adult Drosophila melanogaster. FEBS Journal, 277, 174-185.

Colinet, H., Renault, D., Charoy-Guevel, B., Com, E., 2012b. Metabolic and proteomic profiling of diapause in the aphid parasitoid Praon volucre. Plos One, 7, 12.

Convey, P., 1992. Aspects of the biology of the midge, Eretmoptera murphyi Schaeffer (Diptera, Chironomidae), introduced to Signy Island, maritime Antarctic. Polar Biology, 12, 653-657.

Convey, P., 1996a. The influence of environmental characteristics on life history attributes of Antarctic terrestrial biota. Biological Reviews of the Cambridge Philosophical Society, 71, 191-225.

Convey, P., 1996b. Overwintering strategies of terrestrial invertebrates in Antarctica - The significance of flexibility in extremely seasonal environments. European Journal of Entomology, 93, 489-505.

Convey, P., 1998. Latitudinal variation in allocation to reproduction by the Antarctic oribatid mite, Alaskozetes antarcticus. Applied Soil Ecology, 9, 93-99.

Convey, P., 2010. Life-history adaptations to polar and alpine environments, in: Denlinger, D.L., Lee, R.E. (Eds.), Low Temperature Biology of Insects. Cambridge University Press, Cambridge, UK, pp. 297-321.

Convey, P., Block, W., 1996. Antarctic Diptera: Ecology, physiology and distribution. European Journal of Entomology, 93, 1-13.

Coulson, S.J., Bale, J.S., 1990. Characterization and limitations of the rapid cold- hardening response in the housefly Musca domestica (Diptera, Muscidae). Journal of Insect Physiology, 36, 207-211.

Danks, H.V., 2000. Dehydration in dormant insects. Journal of Insect Physiology, 46, 837-852.

Davis, R.C., 1981. Structure and function of 2 Antarctic terrestiral moss communities. Ecological Monographs, 51, 125-143.

Denlinger, D.L., 1991. Relationship between cold hardiness and diapause, in: Lee, R.E., Denlinger, D.L. (Eds.), Insects at low temperature. Chapman and Hall, New York, pp. 174-198.

Denlinger, D.L., Lee, R.E., 1998. Physiology of cold sensitivity, in: Hallman, G.J., Denlinger, D.L. (Eds.), Temperature Sensitivity in Insects and Application in Integrated Pest Management. Westview Press, Boulder, CO, pp. 55-95.

Denlinger, D.L., Yocum, G.D., Rinehart, J.P., 2005. Hormonal control of diapause, in: Gilbert, L.I., Iatrou, K., Gill, S. (Eds.), Comprehensive insect molecular science. Elsevier, Amsterdam, pp. 615-650.

Denton, R.M., 2009. Regulation of mitochondrial dehydrogenases by calcium ions. Biochimica Et Biophysica Acta-Bioenergetics, 1787, 1309-1316.

Dollo, V.H., Yi, S.X., Lee, R.E., 2010. High temperature pulses decrease indirect chilling injury and elevate ATP levels in the flesh fly, Sarcophaga crassipalpis. Cryobiology, 60, 351-353.

Duman, J.G., Wu, D.W., Xu, L., Tursman, D., Olsen, T.M., 1991. Adaptations of insects to subzero temperatures. Quarterly Review of Biology, 66, 387-410.

Elnitsky, M.A., Benoit, J.B., Denlinger, D.L., Lee, R.E., 2008a. Desiccation tolerance and drought acclimation in the Antarctic collembolan Cryptopygus antarcticus. Journal of Insect Physiology, 54, 1432-1439.

Elnitsky, M.A., Benoit, J.B., Lopez-Martinez, G., Denlinger, D.L., Lee, R.E., 2009. Osmoregulation and salinity tolerance in the Antarctic midge, Belgica antarctica: seawater exposure confers enhanced tolerance to freezing and dehydration. Journal of Experimental Biology, 212, 2864-2871.

Elnitsky, M.A., Hayward, S.A.L., Rinehart, J.P., Denlinger, D.L., Lee, R.E., 2008b. Cryoprotective dehydration and the resistance to inoculative freezing in the Antarctic midge, Belgica antarctica. Journal of Experimental Biology, 211, 524-530.

Everatt, M.J., Worland, M.R., Bale, J.S., Convey, P., Hayward, S.A.L., 2012. Pre-adapted to the maritime Antarctic? - Rapid cold hardening of the midge, Eretmoptera murphyi. Journal of Insect Physiology, 58, 1104-1111.

Feder, M.E., Hofmann, G.E., 1999. Heat-shock proteins, molecular chaperones, and the stress response: Evolutionary and ecological physiology. Annual Review of Physiology, 61, 243-282.

Fujiwara, Y., Denlinger, D.L., 2007. p38 MAPK is a likely component of the signal transduction pathway triggering rapid cold hardening in the flesh fly Sarcophaga crassipalpis. Journal of Experimental Biology, 210, 3295-3300.

Gibbs, A.G., Fukuzato, F., Matzkin, L.M., 2003. Evolution of water conservation mechanisms in Drosophila. Journal of Experimental Biology, 206, 1183-1192.

Goto, S.G., 2000. Expression of Drosophila homologue of senescence marker protein-30 during cold acclimation. Journal of Insect Physiology, 46, 1111-1120.

Goto, S.G., 2001. A novel gene that is up-regulated during recovery from cold shock in Drosophila melanogaster. Gene, 270, 259-264.

Gressitt, L., 1971. Antarctic entomology with emphasis on biogeographical aspects. Pacific Insects Monograph, 25, 167-178.

Hawes, T.C., Bale, J.S., 2007. Plasticity in arthropod cryotypes. Journal of Experimental Biology, 210, 2585-2592.

Hayward, S.A.L., Rinehart, J.P., Sandro, L.H., Lee, R.E., Denlinger, D.L., 2007. Slow dehydration promotes desiccation and freeze tolerance in the Antarctic midge Belgica antarctica. Journal of Experimental Biology, 210, 836-844.

Hoffmann, A.A., 2010. Physiological climatic limits in Drosophila: patterns and implications. Journal of Experimental Biology, 213, 870-880.

Holmstrup, M., Bayley, M., Ramlov, H., 2002. Supercool or dehydrate? An experimental analysis of overwintering strategies in small permeable arctic invertebrates. Proceedings of the National Academy of Sciences of the United States of America, 99, 5716-5720.

Johnson, L.N., 1992. Glycogen-phosphorylase - Control by phosphorylation and allosteric effectors. FASEB Journal, 6, 2274-2282.

Joplin, K.H., Yocum, G.D., Denlinger, D.L., 1990. Cold shock elicits expression of heat- shock proteins in the flesh fly, Sarcophaga crassipalpis. Journal of Insect Physiology, 36, 825-834.

Kayukawa, T., Chen, B., Hoshizaki, S., Ishikawa, Y., 2007. Upregulation of a desaturase is associated with the enhancement of cold hardiness in the onion maggot, Delia antiqua. Insect Biochemistry and Molecular Biology, 37, 1160-1167.

Kelty, J., 2007. Rapid cold-hardening of Drosophila melanogaster in a field setting. Physiological Entomology, 32, 343-350.

Kelty, J.D., Lee, R.E., 1999. Induction of rapid cold hardening by cooling at ecologically relevant rates in Drosophila melanogaster. Journal of Insect Physiology, 45, 719-726.

Kelty, J.D., Lee, R.E., 2001. Rapid cold-hardening of Drosophila melanogaster (Diptera : Drosophilidae) during ecologically based thermoperiodic cycles. Journal of Experimental Biology, 204, 1659-1666.

Kennedy, A.D., 1993. Water as a limiting factor in the Antarctic terrestrial environment - A biogeographical synthesis. Arctic and Alpine Research, 25, 308-315.

Kim, M., Robich, R.M., Rinehart, J.P., Denlinger, D.L., 2006. Upregulation of two actin genes and redistribution of actin during diapause and cold stress in the northern house mosquito, Culex pipiens. Journal of Insect Physiology, 52, 1226-1233.

Knight, H., Trewavas, A.J., Knight, M.R., 1996. Cold calcium signaling in Arabidopsis involves two cellular pools and a change in calcium signature after acclimation. Plant Cell, 8, 489-503.

Kostal, V., 2010. Cell structural modifications in insects at low temperatures, in: Denlinger, D.L., Lee, R.E. (Eds.), Low Temperature Biology of Insects. Cambridge University Press, Cambridge, UK, pp. 116-140.

Kostal, V., Berkova, P., Simek, P., 2003. Remodelling of membrane phospholipids during transition to diapause and cold-acclimation in the larvae of Chymomyza costata (Drosophilidae). Comparative Biochemistry and Physiology B-Biochemistry & Molecular Biology, 135, 407-419.

Kostal, V., Renault, D., Mehrabianova, A., Bastl, J., 2007. Insect cold tolerance and repair of chill-injury at fluctuating thermal regimes: Role of ion homeostasis. Comparative Biochemistry and Physiology. Part A, Physiology, 147, 231-238.

Kostal, V., Simek, P., 1998. Changes in fatty acid composition of phospholipids and triacylglycerols after cold-acclimation of an aestivating insect prepupa. Journal of Comparative Physiology B-Biochemical Systemic and Environmental Physiology, 168, 453-460.

Kostal, V., Zahradnickova, H., Simek, P., 2011. Hyperprolinemic larvae of the drosophilid fly, Chymomyza costata, survive cryopreservation in liquid nitrogen. Proceedings of the National Academy of Sciences of the United States of America, 108, 13041-13046.

Koveos, D.S., 2001. Rapid cold hardening in the olive fruit fly Bactrocera oleae under laboratory and field conditions. Entomologia Experimentalis Et Applicata, 101, 257-263.

Krebs, R.A., Feder, M.E., 1997. Deleterious consequences of Hsp70 overexpression in Drosophila melanogaster larvae. Cell Stress and Chaperones, 2, 60-71.

Kroemer, G., Dallaporta, B., Resche-Rigon, M., 1998. The mitochondrial death/life regulator in apoptosis and necrosis. Annual Review of Physiology, 60, 619-642.

Lalouette, L., Williams, C.M., Hervant, F., Sinclair, B.J., Renault, D., 2011. Metabolic rate and oxidative stress in insects exposed to low temperature thermal fluctuations. Comparative Biochemistry and Physiology a-Molecular & Integrative Physiology, 158, 229-234.

Lee, R.E., 2010. A primer on insect cold tolerance, in: Denlinger, D.L., Richard E. Lee, J.r. (Eds.), Low Temperature Biology of Insects. Cambridge University Press, Cambridge, UK, pp. 3-34.

Lee, R.E., Baust, J.G., 1987. Cold-hardiness in the Antarctic tick, Ixodes uriae. Physiological Zoology, 60, 499-506.

Lee, R.E., Baust, J.G., 1981. Seasonal patterns of cold-hardiness in antarctic terrestrial arthropods. Comparative Biochemistry and Physiology. Part A, Physiology, 70, 579-582.

Lee, R.E., Chen, C.P., Denlinger, D.L., 1987. A rapid cold-hardening process in insects. Science, 238, 1415-1417.

Lee, R.E., Damodaran, K., Yi, S.X., Lorigan, G.A., 2006a. Rapid cold-hardening increases membrane fluidity and cold tolerance of insect cells. Cryobiology, 52, 459-463.

Lee, R.E., Denlinger, D.L., 1985. Cold tolerance in diapausing and non-diapausing stages of the flesh fly, Sarcophaga crassipalpis. Physiological Entomology, 10, 309-315.

Lee, R.E., Denlinger, D.L., 2010. Rapid cold-hardening: Ecological significance and underpinning mechanisms, in: Denlinger, D.L., Lee, R.E. (Eds.), Low Temperature Biology of Insects. Cambridge University Press, Cambridge, UK, pp. 35-58.

Lee, R.E., Elnitsky, M.A., Rinehart, J.P., Hayward, S.A.L., Sandro, L.H., Denlinger, D.L., 2006b. Rapid cold-hardening increases the freezing tolerance of the Antarctic midge Belgica antarctica. Journal of Experimental Biology, 209, 399-406.

Li, A., Denlinger, D.L., 2008. Rapid cold hardening elicits changes in brain protein profiles of the flesh fly, Sarcophaga crassipalpis. Insect Molecular Biology, 17, 565-572.

Li, A.Q., Benoit, J.B., Lopez-Martinez, G., Elnitsky, M.A., Lee, R.E., Denlinger, D.L., 2009. Distinct contractile and cytoskeletal protein patterns in the Antarctic midge are elicited by desiccation and rehydration. Proteomics, 9, 2788-2797.

Li, A.Q., Popova-Butler, A., Dean, D.H., Denlinger, D.L., 2007. Proteomics of the flesh fly brain reveals an abundance of upregulated heat shock proteins during pupal diapause. Journal of Insect Physiology, 53, 385-391.

Liao, Y.X., Frederick, J.E., 2005. The ultraviolet radiation environment of high southern latitudes: Springtime behavior over a decadal timescale. Photochemistry and Photobiology, 81, 320-324.

Liu, K., Tsujimoto, H., Cha, S.J., Agre, P., Rasgon, J.L., 2011. Aquaporin water channel AgAQP1 in the malaria vector mosquito Anopheles gambiae during blood feeding and humidity adaptation. Proceedings of the National Academy of Sciences of the United States of America, 108, 6062-6066.

Lopez-Martinez, G., Benoit, J.B., Rinehart, J.P., Elnitsky, M.A., Lee, R.E., Denlinger, D.L., 2009. Dehydration, rehydration, and overhydration alter patterns of gene expression in the Antarctic midge, Belgica antarctica. Journal of Comparative Physiology. B, Biochemical, Systemic, and Environmental Physiology, 179, 481-491.

Lopez-Martinez, G., Elnitsky, M.A., Benoit, J.B., Lee, R.E., Denlinger, D.L., 2008. High resistance to oxidative damage in the Antarctic midge Belgica antarctica, and developmentally linked expression of genes encoding superoxide dismutase, catalase and heat shock proteins. Insect Biochemistry and Molecular Biology, 38, 796-804.

MacMillan, H.A., Guglielmo, C.G., Sinclair, B.J., 2009. Membrane remodeling and glucose in Drosophila melanogaster: A test of rapid cold-hardening and chilling tolerance hypotheses. Journal of Insect Physiology, 55, 243-249.

MacMillan, H.A., Sinclair, B.J., 2011. Mechanisms underlying insect chill-coma. Journal of Insect Physiology, 57, 12-20.

Mahajan, S., Tuteja, N., 2005. Cold, salinity and drought stresses: An overview. Archives of Biochemistry and Biophysics, 444, 139-158.

Mayhew, P.J., 2007. Why are there so many insect species? Perspectives from fossils and phylogenies. Biological Reviews, 82, 425-454.

Michaud, M.R., Benoit, J.B., Lopez-Martinez, G., Elnitsky, M.A., Lee, R.E., Denlinger, D.L., 2008. Metabolomics reveals unique and shared metabolic changes in response to heat shock, freezing and desiccation in the Antarctic midge, Belgica antarctica. Journal of Insect Physiology, 54, 645-655.

Michaud, M.R., Denlinger, D.L., 2006. Oleic acid is elevated in cell membranes during rapid cold-hardening and pupal diapause in the flesh fly, Sarcophaga crassipalpis. Journal of Insect Physiology, 52, 1073-1082.

Michaud, M.R., Denlinger, D.L., 2007. Shifts in the carbohydrate, polyol, and amino acid pools during rapid cold-hardening and diapause-associated cold-hardening in flesh flies (Sarcophaga crassipalpis): a metabolomic comparison. Journal of Comparative Physiology. B, Biochemical, Systemic, and Environmental Physiology, 177, 753-763.

Mitsumasu, K., Kanamori, Y., Fujita, M., Iwata, K., Tanaka, D., Kikuta, S., Watanabe, M., Cornette, R., Okuda, T., Kikawada, T., 2010. Enzymatic control of anhydrobiosis- related accumulation of trehalose in the sleeping chironomid, Polypedilum vanderplanki. FEBS Journal, 277, 4215-4228.

Monroy, A.F., Dhindsa, R.S., 1995. Low-temperature signal-transduction - Induction of cold acclimation-specific genes of alfalfa by calcium at 25-degrees-C. Plant Cell, 7, 321- 331.

Monroy, A.F., Labbe, E., Dhindsa, R.S., 1997. Low temperature perception in plants: Effects of cold on protein phosphorylation in cell-free extracts. FEBS Letters, 410, 206- 209.

Monroy, A.F., Sarhan, F., Dhindsa, R.S., 1993. Cold-induced changes in freezing tolerance, protein-phosphorylation, and gene exression - Evidence for a role of calcium. Plant Physiology, 102, 1227-1235.

Moribe, Y., Oka, K., Niimi, T., Yamashita, O., Yaginuma, T., 2010. Expression of heat shock protein 70a mRNA in Bombyx mori diapause eggs. Journal of Insect Physiology, 56, 1246-1252.

Orvar, B.L., Sangwan, V., Omann, F., Dhindsa, R.S., 2000. Early steps in cold sensing by plant cells: the role of actin cytoskeleton and membrane fluidity. Plant Journal, 23, 785- 794.

Overgaard, J., Malmendal, A., Sorensen, J.G., Bundy, J.G., Loeschcke, V., Nielsen, N.C., Holmstrup, M., 2007. Metabolomic profiling of rapid cold hardening and cold shock in Drosophila melanogaster. Journal of Insect Physiology, 53, 1218-1232.

Overgaard, J., Sorensen, J.G., Petersen, S.O., Loeschcke, V., Holmstrup, M., 2005. Changes in membrane lipid composition following rapid cold hardening in Drosophila melanogaster. Journal of Insect Physiology, 51, 1173-1182.

Pearson, G., Robinson, F., Gibson, T.B., Xu, B.E., Karandikar, M., Berman, K., Cobb, M.H., 2001. Mitogen-activated protein (MAP) kinase pathways: Regulation and physiological functions. Endocrine Reviews, 22, 153-183.

Pedersen, P.G., Holmstrup, M., 2003. Freeze or dehydrate: only two options for the survival of subzero temperatures in the arctic enchytraeid Fridericia ratzeli. Journal of Comparative Physiology B-Biochemical Systemic and Environmental Physiology, 173, 601-609.

Philip, B.N., Lee, R.E., 2010. Changes in abundance of aquaporin-like proteins occurs concomitantly with seasonal acquisition of freeze tolerance in the goldenrod gall fly, Eurosta solidaginis. Journal of Insect Physiology, 56, 679-685.

Philip, B.N., Yi, S.X., Elnitsky, M.A., Lee, R.E., 2008. Aquaporins play a role in desiccation and freeze tolerance in larvae of the goldenrod gall fly, Eurosta solidaginis. Journal of Experimental Biology, 211, 1114-1119.

Pruitt, N.L., Lu, C., 2008. Seasonal changes in phospholipid class and class-specific fatty acid composition associated with the onset of freeze tolerance in third-instar larvae of Eurosta solidaginis. Physiological and Biochemical Zoology, 81, 226-234.

Purac, J., Burns, G., Thorne, M.A.S., Grubor-Lajsic, G., Worland, M.R., Clark, M.S., 2008. Cold hardening processes in the Antarctic springtail, Cryptopygus antarcticus: Clues from a microarray. Journal of Insect Physiology, 54, 1356-1362.

Qin, W., Neal, S.J., Robertson, R.M., Westwood, J.T., Walker, V.K., 2005. Cold hardening and transcriptional change in Drosophila melanogaster. Insect Molecular Biology, 14, 607-613.

Ragland, G.J., Denlinger, D.L., Hahn, D.A., 2010. Mechanisms of suspended animation are revealed by transcript profiling of diapause in the flesh fly. Proceedings of the National Academy of Sciences of the United States of America, 107, 14909-14914.

Ragland, G.J., Egan, S.P., Feder, J.L., Berlocher, S.H., Hahn, D.A., 2011. Developmental trajectories of gene expression reveal candidates for diapause termination: a key life- history transition in the apple maggot fly Rhagoletis pomonella. Journal of Experimental Biology, 214, 3948-3959.

Rako, L., Hoffmann, A.A., 2006. Complexity of the cold acclimation response in Drosophila melanogaster. Journal of Insect Physiology, 52, 94-104.

Rinehart, J.P., Denlinger, D.L., 2000. Heat-shock protein 90 is down-regulated during pupal diapause in the flesh fly, Sarcophaga crassipalpis, but remains responsive to thermal stress. Insect Molecular Biology, 9, 641-645.

Rinehart, J.P., Hayward, S.A.L., Elnitsky, M.A., Sandro, L.H., Lee, R.E., Denlinger, D.L., 2006a. Continuous up-regulation of heat shock proteins in larvae, but not adults, of a polar insect. Proceedings of the National Academy of Sciences of the United States of America, 103, 14223-14227.

Rinehart, J.P., Li, A., Yocum, G.D., Robich, R.M., Hayward, S.A.L., Denlinger, D.L., 2007. Up-regulation of heat shock proteins is essentail for cold survival during insect diapause. Proceedings of the National Academy of Sciences of the United States of America, 104, 11130-11137.

Rinehart, J.P., Robich, R.M., Denlinger, D.L., 2006b. Enhanced cold and desiccation tolerance in diapausing adults of Culex pipiens, and a role for hsp70 in response to cold shock but not as a component of the diapause program. Journal of Medical Entomology, 43, 713-722.

Rinehart, J.P., Yocum, G.D., Denlinger, D.L., 2000a. Developmental upregulation of inducible hsp70 transcripts, but not the cognate form, during pupal diapause in the flesh fly, Sarcophaga crassipalpis. Insect Biochemistry and Molecular Biology, 30, 515-521.

Rinehart, J.P., Yocum, G.D., Denlinger, D.L., 2000b. Thermotolerance and rapid cold hardening ameliorate the negative effects of brief exposures to high or low temperatures on fecundity in the flesh fly, Sarcophaga crassipalpis. Physiological Entomology, 25, 330-336.

Rosenzweig, M., Kang, K.J., Garrity, P.A., 2008. Distinct TRP channels are required for warm and cool avoidance in Drosophila melanogaster. Proceedings of the National Academy of Sciences of the United States of America, 105, 14668-14673.

Schulte, G.G., Elnitsky, M.A., Benoit, J.B., Denlinger, D.L., Lee, R.E., 2008. Extremely large aggregations of collembolan eggs on Humble Island, Antarctica: a response to early seasonal warming? Polar Biology, 31, 889-892.

Shimada, K., Ohyama, Y., Pan, C.X., 1991. Cold-hardiness of the Antarctic winged midge Parochlus steinenii during the active season at King George Island. Polar Biology, 11, 311-314.

Shreve, S.M., Kelty, J.D., Lee, R.E., 2004. Preservation of reproductive behaviors during modest cooling: rapid cold-hardening fine-tunes organismal response. Journal of Experimental Biology, 207, 1797-1802.

Shreve, S.M., Yi, S.X., Lee, R.E., 2007. Increased dietary cholesterol enhances cold tolerance in Drosophila melanogaster. Cryoletters, 28, 33-37.

Sinclair, B.J., Addo-Bediako, A., Chown, S.L., 2003a. Climatic variability and the evolution of insect freeze tolerance. Biological Reviews, 78, 181-195.

Sinclair, B.J., Chown, S.L., 2005. Climatic variability and hemispheric differences in insect cold tolerance: support from southern Africa. Functional Ecology, 19, 214-221.

Sinclair, B.J., Gibbs, A.G., Roberts, S.P., 2007. Gene transcription during exposure to, and recovery from, cold and desiccation stress in Drosophila melanogaster. Insect Molecular Biology, 16, 435-443.

Sinclair, B.J., Klok, C.J., Scott, M.B., Terblanche, J.S., Chown, S.L., 2003b. Diurnal variation in supercooling points of three species of Collembola from Cape Hallett, Antarctica. Journal of Insect Physiology, 49, 1049-1061.

Sinclair, B.J., Sjursen, H., 2001. Cold tolerance of the Antarctic springtail Gomphiocephalus hodgsoni (Collembola, Hypogastruridae). Antarctic Science, 13, 271- 279.

Sinclair, B.J., Terblanche, J.S., Scott, M.B., Blatch, G.L., Klok, C.J., Chown, S.L., 2006. Environmental physiology of three species of Collembola at Cape Hallett, North Victoria Land, Antarctica. Journal of Insect Physiology, 52, 29-50.

Sjursen, H., Sinclair, B.J., 2002. On the cold hardiness of Stereotydeus mollis (Acari : Prostigmata) from Ross Island, Antarctica. Pedobiologia, 46, 188-195.

Somme, L., 1981. Cold tolerance of alpine, Arctic, and Antarctic collembola and mites. Cryobiology, 18, 212-220.

Steponkus, P.L., 1984. Role of the plasma-membrane in freezing-injury and cold- acclimation. Annual Review of Plant Physiology and Plant Molecular Biology, 35, 543- 584.

Storey, K.B., Baust, J.G., Storey, J.M., 1981. Intermediary metabolism during low- temperature acclimation in the overwintering gall fly larva, Eurosta solidaginis. Journal of Comparative Physiology, 144, 183-190.

Takeda, K., Matsuzawa, A., Nishitoh, H., Tobiume, K., Kishida, S., Ninomiya-Tsuji, J., Matsumoto, K., Ichijo, H., 2004. Involvement of ASK1 in Ca2+-induced p38 MAP kinase activation. EMBO Reports, 5, 161-166.

Takeuchi, K., Nakano, Y., Kato, U., Kaneda, M., Aizu, M., Awano, W., Yonemura, S., Kiyonaka, S., Mori, Y., Yamamoto, D., Umeda, M., 2009. Changes in temperature preferences and energy homeostasis in dystroglycan mutants. Science, 323, 1740-1743.

Thomashow, M.F., 1999. Plant cold acclimation: Freezing tolerance genes and regulatory mechanisms. Annual Review of Plant Physiology and Plant Molecular Biology, 50, 571- 599.

Tursman, D., Duman, J.G., 1995. Cryoprotective effects of thermal hysteresis protein on survivorship of frozen gut cells from the freeze-tolerant centipede Lithobius forficatus. Journal of Experimental Zoology, 272, 249-257.

Vesala, L., Salminen, T.S., Kostal, V., Zahradnickova, H., Hoikkala, A., 2012a. Myo- inositol as a main metabolite in overwintering flies: seasonal metabolomic profiles and cold stress tolerance in a northern drosophilid fly. Journal of Experimental Biology, 215, 2891-2897.

Vesala, L., Salminen, T.S., Laiho, A., Hoikkala, A., Kankare, M., 2012b. Cold tolerance and cold-induced modulation of gene expression in two Drosophila virilis group species with different distributions. Insect Molecular Biology, 21, 107-118.

Weatherhead, E.C., Andersen, S.B., 2006. The search for signs of recovery of the ozone layer. Nature, 441, 39-45.

Wharton, D.A., Downes, M.F., Goodall, G., Marshall, C., 2005. Freezing and cryoprotective dehydration in an Antarctic nematode (Panagrolaimus davidi) visualised using a freeze substitution technique. Cryobiology, 50, 21-28.

Wharton, D.A., Goodall, G., Marshall, C.J., 2003. Freezing survival and cryoprotective dehydration as cold tolerance mechanisms in the Antarctic nematode Panagrolaimus davidi. Journal of Experimental Biology, 206, 215-221.

Williams, J.B., Ruehl, N.C., Lee, R.E., 2004. Partial link between the seasonal acquisition of cold-tolerance and desiccation resistance in the goldenrod gall fly Eurosta solidaginis (Diptera : Tephritidae). Journal of Experimental Biology, 207, 4407-4414.

Worland, M.R., 2010. Eretmoptera murphyi: pre-adapted to survive a colder climate. Physiological Entomology, 35, 140-147.

Worland, M.R., Block, W., 1986. Survival and water-loss in some Antarctic arthropods. Journal of Insect Physiology, 32, 579-584.

Worland, M.R., Block, W., 2003. Desiccation stress at sub-zero temperatures in polar terrestrial arthropods. Journal of Insect Physiology, 49, 193-203.

Worland, M.R., Convey, P., 2001. Rapid cold hardening in Antarctic microarthropods. Functional Ecology, 15, 515-524.

Worland, M.R., Convey, P., 2008. The significance of the moult cycle to cold tolerance in the Antarctic collembolan Cryptopygus antarcticus. Journal of Insect Physiology, 54, 1281-1285.

Worland, M.R., Convey, P., Lukesova, A., 2000. Rapid cold hardening: A gut feeling. Cryo-Letters, 21, 315-324.

Yancey, P.H., 2005. Organic osmolytes as compatible, metabolic and counteracting cytoprotectants in high osmolarity and other stresses. Journal of Experimental Biology, 208, 2819-2830.

Yano, S., Tokumitsu, H., Sodeling, T.R., 1998. Calcium promotes cell survival through CaM-K kinase activation of the protein-kinase-B pathway. Nature, 396, 584-587.

Yi, S.X., Benoit, J.B., Elnitsky, M.A., Kaufmann, N., Brodsky, J.L., Zeidel, M.L., Denlinger, D.L., Lee, R.E., 2011. Function and immuno-localization of aquaporins in the Antarctic midge Belgica antarctica. Journal of Insect Physiology, 57, 1096-1105.

Yi, S.X., Lee, R.E., 2004. In vivo and in vitro rapid cold-hardening protects cells from cold-shock injury in the flesh fly. Journal of Comparative Physiology B-Biochemical Systemic and Environmental Physiology, 174, 611-615.

Yi, S.X., Lee, R.E., 2011. Rapid cold-hardening blocks cold-induced apoptosis by inhibiting the activation of pro-caspases in the flesh fly Sarcophaga crassipalpis. Apoptosis, 16, 249-255.

Yi, S.X., Moore, C.W., Lee, R.E., 2007. Rapid cold-hardening protects Drosophila melanogaster from cold-induced apoptosis. Apoptosis, 12, 1183-1193.

Yocum, G.D., 2001. Differential expression of two HSP70 transcripts in response to cold shock, thermoperiod, and adult diapause in the Colorado potato beetle. Journal of Insect Physiology, 47, 1139-1145.

Yocum, G.D., Joplin, K.H., Denlinger, D.L., 1998. Upregulation of a 23 kDa small heat shock protein transcript during pupal diapause in the flesh fly, Sarcophaga crassipalpis. Insect Biochemistry and Molecular Biology, 28, 677-682.

Yoder, J.A., Benoit, J.B., Denlinger, D.L., Rivers, D.B., 2006. Stress-induced accumulation of glycerol in the flesh fly, Sarcophaga bullata: Evidence indicating anti- desiccant and cryoprotectant functions of this polyol and a role for the brain in coordinating the response. Journal of Insect Physiology, 52, 202-214.

Young, S.R., Block, W., 1980. Experimental studies on the cold tolerance of Alaskozetes antarcticus. Journal of Insect Physiology, 26, 189-200.

Zachariassen, K.E., Li, N.G., Laugsand, A.E., Kristiansen, E., Pedersen, S.A., 2008. Is the strategy for cold hardiness in insects determined by their water balance? A study on two closely related families of beetles: Cerambycidae and Chrysomelidae. Journal of Comparative Physiology. B, Biochemical, Systemic, and Environmental Physiology, 178, 977-984.

Zhang, J., Marshall, K.E., Westwood, J.T., Clark, M.S., Sinclair, B.J., 2011. Divergent transcriptomic responses to repeated and single cold exposures in Drosophila melanogaster. Journal of Experimental Biology, 214, 4021-4029.

Figures

Figure 1.1. Current model of the various cell signaling pathways involved in the cellular response to low temperature in insects. Figure drawn by Dr. Shu-Xia Yi.

Chapter 2: Combined Transcriptomic and Metabolomic Approach Uncovers Molecular Mechanisms of Cold Tolerance in a Temperate Flesh Fly

Abstract

The ability to rapidly respond to changes in temperature is critical for insects and other ectotherms living in variable environments. In a physiological process termed rapid cold-hardening (RCH), exposure to non-lethal low temperature allows many insects to significantly increase their cold tolerance in a matter of minutes to hours. Additionally, there are rapid changes in gene expression and cell physiology during recovery from cold injury, and we hypothesize that RCH may modulate some of these processes during recovery. In this study, we used a combination of transcriptomics and metabolomics to examine the molecular mechanisms of RCH and cold shock recovery in the flesh fly,

Sarcophaga bullata. Surprisingly, out of ~15,000 ESTs measured, no transcripts were upregulated during RCH, and likewise RCH had a minimal effect on the transcript signature during recovery from cold shock. However, during recovery from cold shock, we observed differential expression of ~1,400 ESTs, including a number of heat shock proteins, cytoskeletal components, and genes from several cell signaling pathways. In the metabolome, RCH had a slight yet significant effect on several metabolic pathways, while cold shock resulted in dramatic increases in gluconeogenesis, amino acid synthesis, and cryoprotective polyol synthesis. Several biochemical pathways showed congruence at both the transcript and metabolite levels, indicating that coordinated changes in gene expression and metabolism contribute to recovery from cold shock. Thus, while RCH had 50

very minor effects on gene expression, recovery from cold shock elicits sweeping changes in gene expression and metabolism along numerous cell signaling and biochemical pathways.

Introduction

Due to a small body size and ectothermic nature, adaptations for surviving low temperature are a critical component of an insect’s physiology. In many cases, the overwintering physiology of an insect limits its potential range, particularly in the face of a changing climate (Bale and Hayward, 2010). While insects undergo many gradual biochemical and physiological changes in anticipation of winter (Lee, 2010), they are also capable of responding to low temperature on much shorter time scales. In a process termed rapid cold-hardening (RCH), insects can significantly enhance their cold tolerance in response to brief (i.e. minutes to hours) ecologically relevant chilling exposure (Kelty and Lee, 1999; Lee et al., 1987). Furthermore, during recovery from a cold challenge, insects undergo a number of physiological changes to repair cellular damage, for example by synthesizing heat shock proteins (Hsps) to repair misfolded proteins (Colinet et al.,

2010).

In recent years, several studies have begun to unravel the cellular and molecular mechanisms associated with RCH. Evidence suggests that RCH is triggered at the cellular level by signaling events including p38 MAP kinase (Fujiwara and Denlinger,

2007) and calcium signaling (Teets et al., 2008). Downstream of these signaling events, several physiological changes contributing to RCH have been uncovered. In the flesh fly,

Sarcophaga crassipalpis, there is a slight elevation of the cryoprotectant glycerol in response to RCH, although the amount of glycerol appears to be too low to be a major driver of RCH (Lee et al., 1987). In both Drosophila melanogaster and S. crassipalpis,

RCH increases the proportion of unsaturated fatty acids in the cell membrane (Michaud and Denlinger, 2006; Overgaard et al., 2005), cf. (MacMillan et al., 2009), and inhibits apoptotic pathways to prevent cell death (Yi and Lee, 2011; Yi et al., 2007).

Additionally, a recent proteomics study indicated that three different proteins, including a small heat shock protein, are more abundant in the brains of flesh flies exposed to RCH

(Li and Denlinger, 2008). However, despite these recent advances, many of the underlying mechanisms of RCH remain unknown.

In D. melanogaster, three separate microarray studies have measured gene expression either in response to artificial selection for cold resistance (Telonis-Scott et al., 2009) or in direct response to various cold exposures (Qin et al., 2005; Zhang et al.,

2011). These experiments elucidated many key players in the molecular response to cold, including the multitude of Hsps involved in cold recovery (Qin et al., 2005) and possible cross-talk between environmental stress signals and immune pathways (Zhang et al.,

2011). While Qin et al. (Qin et al., 2005) intended to measure gene expression during

RCH, they allowed a 30 min recovery after hardening, so changes in gene expression could not be directly attributed to the hardening period. The only study to our knowledge to measure gene expression during cold hardening is Sinclair et al. (Sinclair et al., 2007), who measured the expression of five candidate genes in D. melanogaster during RCH. They failed to detect any expression differences in their candidate genes during hardening (although some were differentially expressed during recovery), and 52

hypothesized that RCH does not require the synthesis of new gene products. However, gene expression changes during RCH have yet to be examined on a genome-wide scale.

Also, since D. melanogaster is considerably less cold tolerant than its temperate counterparts (Hoffmann, 2010), similar studies in cold-adapted species are necessary to fully grasp molecular adaptations to low temperature. The completion of an EST library for S. crassipalpis (Hahn et al., 2009) has made such work possible for sarcophagid flies, which have long been a model for cold tolerance research (Adedokun and Denlinger,

1984; Chen et al., 1991). Recent work has described transcriptional changes associated with overwintering dormancy in S. crassipalpis (Ragland et al., 2010), but no large-scale transcriptomic studies of acute cold stress have been conducted in this group.

In this study, we explore the molecular mechanisms of RCH and cold shock recovery in Sarcophaga bullata using a combined transcriptomic and metabolomic approach. Our custom microarray platform allowed us to simultaneously measure the expression of ~15,000 transcripts in response to cold. Additionally, using a targeted GC-

MS approach, we tracked the levels of 35 metabolites in response to the same treatments.

Our experimental design permitted us to address the following hypotheses: 1) RCH causes changes in gene expression and/or metabolism during the hardening period; 2)

Recovery from cold shock elicits changes in gene expression and/or metabolism to repair cellular damage; and 3) RCH conditions alter gene expression and/or metabolite composition during recovery from cold shock. Our results indicate that while differential gene expression is not a major contributor to RCH, RCH does have a significant effect on specific metabolic pathways. Furthermore, our results identify a number of genes and metabolites that are rapidly elevated during recovery from cold shock. 53

Materials and Methods

Animals

Flesh flies, Sarcophaga bullata, were reared at 25°C and 16:8 L:D according to

Denlinger et al. (1972). Red-eye, pharate adults were used for all experiments.

Experimental conditions

For the microarray experiments, pharate adult flies were exposed to the following temperature conditions: control (maintained at 25°C; Figure 2.1A), RCH (exposed to 0°C for 2 h; Figure 2.1A), cold shock + 2 h recovery (CS+2R; -10°C for 2 h followed by 25°C for 2 h; Figure 2.1B), and RCH + cold shock + 2 h recovery (RCH+CS+2R; 0°C for 2 h, -

10°C for 2h, followed by 25°C for 2 h; Figure 2.1C). For metabolomics experiments, flies were exposed to the same 4 treatments as well as 2 additional treatments with 24 h recovery (CS+24R and RCH+CS+24R; Figure 2.1B, C). Immediately after treatment, flies were snap frozen in liquid nitrogen and stored at -70°C until RNA and metabolite extraction. For the microarray experiments, we collected 6 biological replicates for each treatment, while the metabolomics experiments consisted of 10 replicates for each treatment.

Microarray data acquisition

For each RNA sample, four flies were removed from storage at -70°C and immediately homogenized together in 4 ml of Tri Reagent (Ambion, Carlsbad, CA) using a ground glass homogenizer. From each homogenate, 1 ml was removed, and RNA was purified using the RiboPure Kit (Ambion, Carlsbad, CA) according to the manufacturer’s

protocol. Total RNA was quantified using a NanoDrop spectrophotometer (Thermo

Fisher Scientific, Waltham, MA), and the integrity was checked on an Agilent 2100

Bioanalyzer (Agilent, Santa Clara, CA). Starting with 500 ng of total RNA, Cy3 and Cy5 labeled cRNA was generated with the Agilent Low RNA Input Linear Amplification Kit

(Agilent, Santa Clara, CA). The labeled samples were hybridized to custom Agilent 4 x

44K arrays containing a previously designed probe set from a closely related species, S. crassipalpis (Ragland et al., 2010). Despite using a probeset designed for a different species, MA plots suggested a distribution of fold change and intensity values similar to those observed when similar arrays were hybridized to S. crassipalpis. Also, targeted analyses in our lab have revealed very little sequence difference between S. crassipalpis and S. bullata. Arrays were scanned on an Agilent G2505B scanner, and data were extracted using Feature Extraction 9.5 software (Agilent, Santa Clara, CA). For each treatment comparison, six biological replicates were conducted; however, several failed to pass quality control, hence we report the results from four independent arrays for each comparison. Because the 16 arrays did not contain the same 4 replicates of each treatment, our dataset includes 5 replicates of control flies, 4 replicates of RCH flies, 5 replicates of CS+2R flies, and 5 replicates of RCH+CS+2R flies. The hybridization scheme is depicted in Figure 2.1D.

Metabolomics data acquisition

Individual frozen flies were homogenized in 750µL of cold (-20°C) 2:1 methanol:chloroform using a tungsten bead-beating apparatus (RetschTM MM301, Retsch

GmbH, Haan, Germany) at 25 Hz for 1.5 min. After homogenization, 500 µl ice-cold water was added to each sample to separate an upper aqueous phase from a lower non- 55

polar phase. Two aliquots (60 and 180 µL) of the upper phase were transferred to chromatographic vials and vacuum dried using a Speed Vac Concentrator (MiVac,

Genevac Ltd., Ipswitch, England). The 60 µL aliquots were used for the quantification of the few metabolites from the larger volume sample that surpassed the upper detection limit of the equipment. The dried samples were doubly derivatized by first suspending the sample in 30 µL of 20 mg.mL-1 methoxyaminehydrochloride (Sigma-Aldrich, St. Louis,

MO, USA) and heating for 90 min at 40°C, followed by the addition of 30 μL of N- methyl-N-(trimethylsilyl) trifluoroacetamide (Sigma-Aldrich, St. Louis, MO, USA) and heating for 45 min at 40°C. This on-line derivatization process was conducted with a

CTC CombiPal autosampler (GERSTEL GmbH & Co.KG, Mülheim an der Ruhr,

Germany), which standardized the derivatization process and ensured that each sample was derivatized for an identical amount of time.

To identify and quantify metabolites, samples were injected into a GC-MS consisting of a Trace GC Ultra chromatograph and a Trace DSQII quadrupole mass spectrometer (Thermo Fischer Scientific Inc, Waltham, MA, USA). We injected 1 µl of each sample into the GC using the splitless mode (25:1), and the samples were gradually heated from 70 to 310°C as follows: the oven temperature ranged from 70 to 170°C at 5

°C.min-1, from 170 to 310 °C at 7 °C.min-1, and remained for 3 min at 310 °C. We used a fused silica column (TR5 MS, I.D. 25 mm, 95% dimethyl siloxane, 5% Phenyl

Polysilphenylene-siloxane), and helium at a rate of 1 mL.min-1 as the gas carrier. MS detection was achieved using electron impact. Ion source temperature was set to 250 °C, and the MS transfer line to 300 °C. The order of injection was randomized to prevent bias due to machine drift. Compounds were identified in the MS using a selective ion mode 56

(electron energy: -70 eV) to only search for ions that matched metabolites in our database of 60 pure reference compounds. The quantity of each metabolite was determined using the quadratic calibration curves drawn from pure compounds run at 11 different concentrations ranging from 10 to 3000 µM. Concentrations were also corrected relative to an internal standard, arabinose, to correct for any sample loss during extraction or injection. Finally, all concentrations were divided by the fresh mass of the individual.

Data analysis

Microarray data were processed and normalized using the limma package for R

(Smyth, 2004). The data were background corrected and normalized within arrays using a lowess approach. Additionally, to standardize intensities across arrays, we conducted a between array normalization with the “scale” method. After normalization, data integrity was checked using a combination of MA plots, box plots, and red-green intensity plots.

Replicate probes were collapsed by taking the average M value for each spot. To find differentially expressed probes, we used the limma pipeline to fit a linear model and compute empirical Bayes statistics from the linear contrasts. P-values were adjusted using the Benjamini and Hochberg method (Benjamini and Hochberg, 1995) to control the

FDR. For the top 150 differentially expressed genes, we constructed a heat map of the expression ratios using JMP 9 (SAS, Carry, NC). The microarray data are deposited in the NCBI Gene Expression Omnibus database, accession number GSE36483.

To conduct the multivariate analyses described below, the normalized probe intensities for each sample were averaged across replicate probes, and the resulting intensities were averaged across technical replicates of the same individual. The

intensities were log2 transformed prior to data analysis. Principal components analysis was conducted using the R package prcomp, while hierarchical clustering of the phenotypic classes was computed using the Ward method in JMP 9. To test for enriched functional categories in our dataset, Sarcophaga ESTs were mapped to their D. melanogaster homolog, and lists of significantly (FDR<0.05) differentially expressed genes were submitted to the DAVID functional annotation database

(http://david.abcc.ncifcrf.gov) (Huang et al., 2009). Specifically, we used the functional annotation clustering tool in DAVID, which finds overrepresented GO terms (Ashburner et al., 2000) and places these GO terms into non-redundant clusters. We also tested for enriched KEGG pathways using the R package GSA (Efron and Tibshirani, 2007).

Additionally, gene sets were generated from lists of differentially expressed genes from previous microarray studies, and these a priori lists were also tested for enrichment using

GSA. Whereas DAVID simply tests for enrichment in a defined list of genes, GSA takes into account expression values for each probe in calculating the enrichment score. For

GSA, we used full, non-redundant log2 transformed data and performed 1,000 permutations to estimate the FDR.

To analyze the metabolomics data, metabolite contents expressed as nmol/mg fresh mass were first log2 transformed. Metabolite quantities were compared across the six treatments using ANOVA and a pooled t-test in JMP 9, and the resulting p-values were adjusted using the Benjamini and Hochberg method (Benjamini and Hochberg,

1995) to control the FDR at 0.05. PCA and hierarchical clustering on the phenotypic classes were conducted as before. To identify metabolic pathways associated with our treatments, metabolite pathway enrichment analysis was conducted using MetaboAnalyst 58

(www. Metaboanalyst.ca), a web-based platform for metabolomics data analysis (Xia et al., 2009).

Results and Discussion

RCH significantly enhances cold tolerance

To verify that our strain of S. bullata exhibited the RCH phenotype, we measured the effects of RCH on the cold tolerance of pharate adult flies (i.e. flies that have completed ~75% of adult development but have yet to eclose from the puparium) (Lee et al., 1987). In this experiment, the criterion for survival was successful eclosion, which occurred approximately 5-6 days after the experiment. While ~80% of flies survived a 2 h cold shock at -8°C, only 20% survived at -9°C and none successfully emerged following 2 h at -10°C (Figure 2.2). However, when flies were exposed to 0°C for 2 h prior to being cold shocked, nearly all survived -9°C and ~50% survived -10°C. Thus, for these experiments, we selected -10°C as our cold shock temperature, since at this temperature RCH allowed significant survival at a temperature that is normally 100% lethal. While no flies emerge as adults after experiencing a -10°C cold shock, all were still able to continue adult morphogenesis, so flies sampled 2 h after cold shock were clearly still alive.

RCH has very little effect on the transcriptome

Our treatment design and microarray hybridization scheme (Figure 2.1) allowed us to test two separate hypotheses regarding the effects of RCH on gene expression: 1)

RCH induces or turns off specific genes during the hardening period (i.e. the 2 h at 0°C);

and 2) RCH alters the transcriptional signature during recovery from cold shock.

However, we found very little evidence in support of either hypothesis. In the Control vs.

RCH comparison, no ESTs were differentially expressed between the two groups (Table

2.1). In addition, while numerous ESTs were differentially expressed during recovery from cold shock (see below), RCH had little effect on gene expression during recovery; a direct comparison of the CS+2R and RCH+CS+2R comparisons showed that only 5

ESTs were differentially expressed between the two treatments, none of which differed by more than 33%. Of these 5 ESTs, 3 were downregulated cell signaling genes

(Arrestin-1, TNF-receptor-associated factor 4, and CG10737), perhaps indicating an

RCH-mediated shutdown of certain cell signaling events. In particular, TNF-receptor- associated factor 4 is a positive regulator of apoptosis (Cha et al., 2003); thus, significant downregulation of this transcript in the RCH+CS+2R group relative to the CS+2R group may reflect inhibition of apoptosis by RCH (Yi and Lee, 2011).

While RCH had little effect on a gene-by-gene comparison, we also conducted several multivariate analyses in an attempt to reveal subtle differences in gene expression caused by RCH. A heat map diagram of the top 150 differentially expressed ESTs, as determined by ranking the F-statistics for each probe, separated our four treatment groups into four distinct clades, indicating that RCH causes some differences in expression patterns among the most labile genes (Figure 2.3). However, when the entire dataset is considered, we were once again unable to detect differences attributable to RCH. Both

PCA and hierarchical clustering of the entire dataset produced similar results, in that the phenotypes form two distinct clusters: a cluster consisting of control and RCH samples, and a cluster consisting of CS+2R and RCH+CS+2R samples (Figure 2.4). Finally, we 60

attempted to discover subtle, coordinated changes in gene expression in response to RCH using gene set analysis (GSA) to 1) test for enrichment of specific KEGG pathways, and

2) test whether there were detectable similarities between our dataset and expression patterns from previous microarray studies of cold and other stressors using a priori gene lists from those published reports (Efron and Tibshirani, 2007). Because GSA tests for enrichment across entire pathways or lists of genes, it has the capability to detect differential expression of pathways even in the absence of major changes in individual genes. However, no gene sets were enriched when comparing control vs. RCH and

CS+2R vs. RCH+CS+2R. Thus, at both the individual gene and pathway level, RCH had very little effect on gene expression, both during the hardening period and during recovery from cold shock.

In plants, a number of cold-related genes are upregulated within minutes of transfer to low temperature (Thomashow, 1999), so we suspected some genes may be differentially expressed during RCH. However, in D. melanogaster, RCH does not appear to be associated with changes in gene expression (Sinclair et al., 2007) and can occur even when protein synthesis is blocked with cycloheximide (Misener et al., 2001).

Because RCH occurs so rapidly, there may not be ample time to synthesize new gene products during hardening, particularly at low temperatures. This idea is supported by a proteomics study of the wasp Aphidius colemani, where relatively few proteins were upregulated during cold exposure, while nearly 1/3 of the proteome changed during recovery (Colinet et al., 2007). Similarly, in the linden bug, Pyrrhocoris apterus, hsp70 is expressed at very low levels during cold exposure but is rapidly upregulated upon return to room temperature (Kostal and Tollarova-Borovanska, 2009). However, even more 61

puzzling is the failure of RCH to alter expression patterns during recovery from cold shock. RCH triggers a number of signaling pathways, including MAP kinase signaling

(Fujiwara and Denlinger, 2007) and apoptosis signaling (Yi and Lee, 2011), thus we hypothesized this would be reflected in the gene expression profiles upon return to ambient temperature. However, of the ~1,400 differentially expressed ESTs during recovery from cold shock (Table 2.1; see discussion below), only 5 ESTs were significantly changed by RCH. These results indicate that the existing cellular machinery is sufficient to carry out RCH, suggesting that second messenger systems and other post- translational processes are the likely drivers of RCH.

An abundance of ESTs are differentially expressed during recovery from cold shock

While RCH had little effect on gene expression in S. bullata, we identified 1,378

ESTs that were differentially expressed in the Control vs. CS+2R comparison (Table

2.1). Of these, 111 were either 1.5X up- or down-regulated. The results for the Control vs. RCH+CS+2R comparison were remarkably similar; 1,525 ESTs were differentially expressed, including 134 that were 1.5X up- or down-regulated (Table 2.1). This, in combination with the results discussed above, indicates that recovery from cold shock is the primary driver of differential gene expression in our treatments. Because the gene expression profiles of CS+2R and RCH+CS+2R flies were largely similar, we will discuss the results of the Control vs. CS+2R and Control vs. RCH+CS+2R comparisons together, but for simplicity will refer to specific results from only the Control vs. CS+2R comparison.

Using the DAVID functional annotation tool (Huang et al., 2009), we tested for enriched GO terms in our list of differentially expressed genes (Table 2.2, 2.3). Of note, we found several enriched GO terms related to cytoskeletal organization and cell shape.

Because the cell membrane is one of the major sites of cold shock damage (Cossins,

1983), recent evidence suggests that changes in the actin cytoskeleton are an essential component of cold-hardening and repair of cold damage. In mosquitoes, for example, two actin genes are upregulated in preparation for winter, and cold shock induces a reorganization of actin fibers in the midgut (Kim et al., 2006). Similar effects of cold on the cytoskeleton have been observed in plants (Abdrakhamanova et al., 2003), fish

(Detrich et al., 1989), and even mammals (Al-Fageeh and Smales, 2006), suggesting a central role for the actin cytoskeleton during cold stress. In our dataset, we found differential expression of 55 ESTs related to the GO biological process “actin cytoskeleton organization,” indicating that cytoskeletal reorganization is also an important component of cold shock recovery in S. bullata. Other noteworthy results from the DAVID analysis are 14 ESTs related to the heat shock response and 32 transcripts associated with programmed cell death, supporting the critical role of apoptotic cell death during cold shock injury (Yi and Lee, 2011; Yi et al., 2007).

We also identified several enriched KEGG pathways and a priori gene sets using

GSA (Table 2.4, 2.5). In particular, we identified several biochemical pathways

(discussed below) and cell signaling pathways that were enriched during recovery from cold shock. Some of these cell signaling pathways, including Jak/STAT signaling and insulin signaling, have yet to be implicated in the response to cold. Jak/STAT signaling does have known immune functions (Arbouzova and Zeidler, 2006), perhaps explaining 63

in part the overlap between cold stress and immune signaling (Zhang et al., 2011).

Others, such as Wnt signaling and dorso-ventral axis formation, are predominantly considered developmental pathways. However, recent research in Drosophila is revealing that many embryonic and developmental signaling pathways are co-opted for other functions during later stages of development. For example, Wnt signaling participates in hindgut regeneration in adults (Takashima et al., 2008), while the primary regulators of dorso-ventral axis formation are also key regulators of the fungal immune response later in life (Lemaitre et al., 1996). One enriched signaling pathway that is particularly promising is phosphatidylinositol signaling; components of this signaling system have been implicated as regulators of apoptosis during oxidative stress in Drosophila (Terhzaz et al. 2010). However, further experiments are needed to validate the exact function of these signaling pathways during cold shock recovery in S. bullata.

In addition to enrichment of several KEGG categories, GSA also revealed congruence between our dataset and several other transcriptomic datasets from the literature. In particular, during recovery from cold shock, there was significant enrichment of genes involved in the Drosophila response to cold (Qin et al., 2005), with several genes being significantly upregulated in both studies (Table 2.6). In addition, there was strong enrichment of genes involved in the Drosophila response to hypoxia

(Liu et al., 2006), hyperoxia (Landis et al., 2004), oxidative stress (Girardot et al., 2004), and heat (Sorensen et al., 2005), suggesting that these different environmental stresses have overlapping transcriptional signatures. Indeed, cold exposure directly causes oxidative stress in insects (Lalouette et al., 2011), so this likely explains the similarities in gene expression between cold and various forms of oxygen stress. The genes responsible 64

for much of this overlap were the heat shock proteins, which are known to be involved in a number of environmental stresses (Feder and Hofmann, 1999). Examples of other genes upregulated both in our cold shock recovery treatment and during other forms of environmental stress include: phosphoenolpyruvate carboxykinase (PEPCK), an important metabolic regulator (see below); hairy, a transcription factor that regulates metabolism during hypoxic stress (Zhou et al., 2008); and DGP-1, a translation elongation factor that functions during periods of oxidative stress (Girardot et al., 2004).

Interestingly, a gene set derived from a recent study on repeated cold exposure in D. melanogaster (Zhang et al., 2011), which included a single cold exposure treatment similar to our CS+2R treatment, showed no significant enrichment in our data. However, this study used a milder temperature regime and a longer recovery period (6 h), which could explain the lack of overlap.

RCH has a significant impact on several metabolic pathways

While RCH had no detectable effect on gene expression, we did observe several metabolic changes attributable to RCH. During the 2 h at 0°C, there was a significant increase in two glycolytic intermediates, glucose-6-phosphate (45% increase) and fructose-6-phosphate (9% increase; Figure 2.5, Table 2.7); these were the only two metabolites that changed during RCH. This suggests a rapid shift from aerobic metabolism to glycolysis/gluconeogenesis during RCH, perhaps to begin the process of diverting carbon flow toward cryoprotectant synthesis (Kostal et al., 2004). Similar observations were made by Michaud and Denlinger (2007) and Overgaard et al. (2007) in

S. crassipalpis and D. melanogaster, respectively, indicating that a shift towards glucose production may be a general feature of RCH. In addition, S. crassipalpis undergoes 65

several other changes during RCH, including elevation of the cryoprotectants glycerol and sorbitol. However, it is worth noting that Michaud and Denlinger (2007) used 8 h at

4°C as their RCH treatment; thus, these flies had a much longer time at a milder temperature to carry out these biochemical changes.

In addition to changes in metabolism during RCH, RCH also altered the metabolic signature of flies during recovery from cold shock. During recovery from cold shock, we observed increases in numerous compounds, including nearly every amino acid, sugar, and polyol that we measured (Figure 2.5, Table 2.7). While a number of these changes may be adaptive to help repair cold injury (such as synthesis of cryoprotectants to protect damaged membranes and proteins), some of these changes may simply reflect significant protein breakdown and an inability to maintain homeostasis due to cold shock damage. However, for several compounds, RCH dampened the increase, perhaps reflecting a homeostasis-preserving function for RCH. For example, after a 2 h recovery from cold shock, flies exposed to RCH had 47% less glucose, 25% less glycerol, and

46% less sorbitol. These differences resulted in RCH+CS+2R flies having significantly lower total levels of both sugars and polyols than their CS+2R counterparts (Figure 2.5,

Table 2.7). In Drosophila, long-term cold acclimation helps preserve metabolic homeostasis following cold shock (Colinet et al., 2012), and it appears that short-term

RCH has a similar effect in Sarcophaga. This effect was not as pronounced after 24 h of recovery, but nonetheless this evidence suggests that RCH helps preserve homeostasis immediately following cold shock.

To help put the above metabolic changes into context, we also conducted metabolic pathway enrichment analysis on our metabolomics data set. In a similar manner to GSA, metabolic pathway analysis looks for coordinated changes in metabolites that belong to the same pathway. In the control vs. RCH comparison, several metabolic pathways were enriched, including glycolysis/gluconeogenesis, amino sugar and nucleotide sugar metabolism, and starch and sucrose metabolism (Table 2.8).

Interestingly, in the CS+2R vs. RCH+CS+2R comparison, despite a number of differences in individual metabolites, no specific pathway enrichment was detected, meaning that the observed metabolite differences were not coordinated along an entire pathway. In contrast, after 24 h of recovery, there were several metabolic pathways that showed differences between CS+24R and RCH+CS+24R individuals, notably several pathways related to sugar and amino acid metabolism. One pathway that was enriched both during RCH and during recovery, the pentose phosphate pathway (also referred to as the hexose monophosphate shunt), was previously shown to provide some of the energy and reducing equivalents for cryoprotectant synthesis (Storey and Storey, 1990). Thus, the impact of RCH on this pathway may be particularly important for increased cold tolerance. While these changes need to be explored in more detail to determine their adaptive benefits, they do demonstrate that RCH exerts a direct effect on certain metabolic pathways, both during hardening and during recovery from cold shock.

Recovery from cold shock elicits sweeping changes in metabolite content

Although RCH had substantial effects on the metabolome, the effects of cold shock and recovery on the metabolome were even more dramatic (Figure 2.5, 2.6, Table

2.7). Using both PCA and hierarchical clustering (Figure 2.6), our metabolomics data 67

forms three distinct clusters: a cluster consisting of control and RCH flies, a cluster consisting of CS+2R and RCH+CS+2R flies, and a cluster consisting of CS+24R and

RCH+CS+24R flies. Thus, similar to the gene expression data, recovery from cold shock was the major driver of differences in metabolite content. As mentioned above, recovery from cold shock caused an increase in almost every metabolite we measured; out of 24 total amino acids, sugars, and polyols in our dataset, 22 were elevated by 24 h of recovery from cold shock (Figure 2.5, Table 2.7). Because nearly every compound changed, the results of metabolic pathway analysis were not particularly informative, since nearly every pathway showed evidence of enrichment relative to control samples.

One interesting finding from these data is the presence of a multiple-component cryoprotectant system in S. bullata. While the importance of glycerol as a cryoprotectant has been well-established in S. bullata (Yoder et al., 2006), sorbitol and inositol were both more abundant than glycerol in our samples. Of all the metabolites we measured, sorbitol showed the most dramatic changes; CS+24 R flies showed a 96-fold increase in sorbitol levels compared to control, an increase in sorbitol content from 6.2±0.6 to

608±66.9 pmol mg-1. However, despite the dramatic accumulation of sorbitol and other cryoprotectants, polyol levels in response to cold shock were much lower than those observed in overwintering individuals, both within S. bullata and in comparison with other species. For example, diapausing pupae of S. bullata accumulate ~450x more glycerol than CS+24 R flies in our study (Chen et al., 1991), while another temperate dipteran, Eurosta solidaginis, accumulates sorbitol at levels ~250x greater than CS+24R flies (Storey et al., 1981). Thus, while our data clearly demonstrate a multiple-component cryoprotectant system during recovery from cold shock, the actual concentrations of these

compounds are much lower than in some overwintering insects. However, evidence suggests that polyolcryoprotectants may have specific protective functions in the cell, in addition to the colligative effects most often attributed to these compounds (Yancey,

2005).

Coordinated changes in gene expression and metabolism

While the transcriptomics and metabolomics data have been discussed separately up to this point, there were several points of congruence between the two datasets. First, the shift to gluconeogenesis during recovery from cold shock correlates with a 3.9-fold upregulation of PEPCK. PEPCK catalyzes the conversion of oxaloacetate to phosphoenolpyruvate, the rate-limiting step of gluconeogenesis (Hanson and Reshef,

1997). Interestingly, we also found evidence in our GSA that genes related to insulin signaling are enriched during recovery from cold shock (Table 2.4, 2.5). This enrichment of insulin signaling appeared to be driven by significant upregulation of protein kinase

61c and downregulation of gigas, two regulatory proteins that are linked to insulin signaling. Normally, insulin signaling is thought to inhibit expression of PEPCK (Barthel and Schmoll, 2003), but this concept is primarily based on vertebrate research; in both S. crassiapalpis and the apple maggot, Rhagoletis pomonella, there is concurrent upregulation of both insulin signaling and PEPCK during overwintering diapause

(Ragland et al., 2010; Ragland et al., 2011). Also, due to the relative lack of information concerning invertebrate insulin signaling (Wu and Brown, 2006), our a priori list only contains 11 insulin-related genes, thus our list is likely not comprehensive.

Our combined GSA and metabolite pathway analysis also revealed four pathways that were enriched at both the transcript and metabolite level (Table 2.9). One such

KEGG pathway, valine, leucine, and isoleucine biosynthesis, was strongly enriched in both datasets. Also, the KEGG pathway urea cycle and metabolism of amine groups was significantly down-regulated during recovery from cold shock, thus contributing to the observed accumulation of amino acids (Table 2.4, 2.5). We hypothesize that increased amino acid biosynthesis either serves a cryoprotective role during recovery from cold shock, is necessary to support the burst of protein synthesis during recovery from cold shock (Joplin et al., 1990), or a combination of both. While previous studies have primarily focused on carbohydrates and low-molecular weight polyols in response to cold, recent research suggests that amino acids are indeed an important component of cold-hardiness in some insects (Kostal et al., 2011a; Kostal et al., 2011b). A second pathway that was enriched in both the transcriptome and metabolome was pyruvate metabolism, which serves as a key intersection point between carbohydrate and amino acid metabolism (Kanehisa and Goto, 2000). Finally, two pathways, inositol phosphate metabolism and pentose and glucuronate interconversions, reflect the biosynthesis of the second most abundant polyol, inositol, and the three 5-carbon polyols in our dataset, ribitol, xylitol, and arabitol. Overall, these results show reasonably good agreement between the transcriptomic and metabolomic data, although these pathways should be further explored with a targeted approach to verify their role in cold shock recovery.

While our data showed relatively good agreement at the transcript and metabolite level, there were many instances when changes in metabolism were not reflected by changes in gene expression. This is not surprising, as there are many levels of biological 70

organization between transcription and metabolite synthesis (Feder and Walser, 2005).

Thus, the transcriptomics data in particular should be taken with caution, because transcript levels may not be entirely representative of the physiological state of the organism. However, there are well-established cases where gene expression and metabolic endpoints are strongly correlated, as in the case of PEPCK discussed above

(Hanson and Reshef, 1997). Also, while transcriptomics data may not always reflect the physiological function of an organism, they can provide important clues as to which pathways are activated during times of stress, by identifying the downstream genes that are regulated by these pathways (Feder and Walser, 2005).

Conclusions

Our experiments allowed us to examine the effects of RCH on the transcriptome and metabolome both during the hardening period and during recovery from a subsequent cold shock. Despite its dramatic effect on cold tolerance, RCH had little effect on the transcriptome of S. bullata. Using several multivariate tools, we were unable to detect differences in gene expression attributable to RCH. At the very least, these results indicate that transcriptional regulation is not a major contributor to RCH. Instead, future research will focus on changes in protein phosphorylation and other signaling events that govern RCH. For example, p38 MAP kinase is rapidly phosphorylated during RCH

(Fujiwara and Denlinger, 2007), and there are likely numerous other signaling proteins, both upstream and downstream of p38, that have yet to be identified. In particular, we would like to identify the signaling mechanisms that drive the observed changes in metabolism attributable to RCH. 71

While the primary purpose of this study was to dissect the molecular mechanisms of RCH, a secondary objective was to discover pathways involved in recovery from cold shock. Indeed, while RCH had minor effects on gene expression and metabolism, an abundance of genes and metabolites were differentially expressed during recovery from cold shock. Our results have allowed us to generate/advance the following hypotheses regarding the mechanisms of cold injury repair in insects: 1) Cytoskeletal rearrangement is crucial for the repair of cold damage, as evidenced by the abundance of cytoskeletal genes upregulated during recovery from cold shock; 2) Coordination of numerous cell signaling pathways is a key component of cold-damage repair. Future experiments will seek to identify the relationships between these pathways and determine which are essential for cold repair; and 3) Many of the genes involved in cold shock repair are also essential for other forms of environmental stress. Evidence from both the present study and previous work suggests the presence of a common stress-signaling axis (or axes) that is activated by disparate forms of environmental stress.

Acknowledgements

We appreciate the assistance of Yanping Zhang of the Florida Genetics Core Facility for her assistance with microarray experiments. Also, we thank Vanessa Larvor at the

Unversity of Rennes for maintaining the GC-MS and running samples for the metabolomics experiments. Finally, we’d like to acknowledge the late Rob Michaud for his help in conceiving and designing this study.

Grants

This research was supported by NSF grant IOS-0840772 (www.nsf.gov). The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.

References

Abdrakhamanova, A., Wang, Q.Y., Khokhlova, L., Nick, P., 2003. Is microtubule disassembly a trigger for cold acclimation? Plant and Cell Physiology, 44, 676-686.

Adedokun, T.A., Denlinger, D.L., 1984. Cold-hardiness - A component of the diapause syndrome in pupae of the flesh flies, Sarcophaga crassipalpis and Sarcophaga bullata. Physiological Entomology, 9, 361-364.

Al-Fageeh, M.B., Smales, C.M., 2006. Control and regulation of the cellular responses to cold shock: the responses in yeast and mammalian systems. Biochemical Journal, 397, 247-259.

Arbouzova, N.I., Zeidler, M.P., 2006. JAK/STAT signalling in Drosophila: insights into conserved regulatory and cellular functions. Development, 133, 2605-2616.

Ashburner, M., Ball, C.A., Blake, J.A., Botstein, D., Butler, H., Cherry, J.M., Davis, A.P., Dolinski, K., Dwight, S.S., Eppig, J.T., Harris, M.A., Hill, D.P., Issel-Tarver, L., Kasarskis, A., Lewis, S., Matese, J.C., Richardson, J.E., Ringwald, M., Rubin, G.M., Sherlock, G., Gene Ontology, C., 2000. Gene Ontology: tool for the unification of biology. Nature Genetics, 25, 25-29.

Baker, D.A., Russell, S., 2009. Gene expression during Drosophila melanogaster egg development before and after reproductive diapause. BMC Genomics, 10, 242.

Bale, J.S., Hayward, S.A.L., 2010. Insect overwintering in a changing climate. Journal of Experimental Biology, 213, 980-994.

Barthel, A., Schmoll, D., 2003. Novel concepts in insulin regulation of hepatic gluconeogenesis. American Journal of Physiology. Endocrinology and Metabolism, 285, E685-E692.

Beckstead, R.B., Lam, G., Thummel, C.S., 2005. The genomic response to 20- hydroxyecdysone at the onset of Drosophila metamorphosis. Genome Biology, 6.

Benjamini, Y., Hochberg, Y., 1995. Controlling the false discovery rate - A practical and powerful approach to multiple testing. Journal of the Royal Statistical Society Series B- Methodological, 57, 289-300.

Cha, G.H., Cho, K.S., Lee, J.H., Kim, M., Kim, E., Park, J., Lee, S.B., Chung, J., 2003. Discrete functions of TRAF1 and TRAF2 in Drosophila melanogaster mediated by c-Jun N-terminal kinase and NF-kappa B-dependent signaling pathways. Molecular and Cellular Biology, 23, 7982-7991.

Chen, C.P., Denlinger, D.L., Lee, R.E., 1991. Seasonal variation in generation time, diapause and cold hardiness in a central Ohio population of the flesh fly, Sarcophaga bullata. Ecological Entomology, 16, 155-162.

Colinet, H., An Nguyen, T.T., Cloutier, C., Michaud, D., Hance, T., 2007. Proteomic profiling of a parasitic wasp exposed to constant and fluctuating cold exposure. Insect Biochemistry and Molecular Biology, 37, 1177-1188.

Colinet, H., Larvor, V., Laparie, M., Renault, D., 2012. Exploring the plastic response to cold acclimation through metabolomics. Functional Ecology, 26, 711-722.

Colinet, H., Lee, S.F., Hoffmann, A., 2010. Knocking down expression of Hsp22 and Hsp23 by RNA interference affects recovery from chill coma in Drosophila melanogaster. Journal of Experimental Biology, 213, 4146-4150.

Cossins, A.R., 1983. The adaptation of membrane structure and function to changes in temperature, in: Cossins, A.R., Sheterline, P. (Eds.), Cellular Acclimatisation to Environmental Change. Cambridge University Press, Cambridge, pp. 3-31.

Denlinger, D.L., 1972. Induction and termination of pupal diapause in Sarcophaga (Diptera - Sarcophagidae). Biological Bulletin, 142, 11-&.

Detrich, H.W., Johnson, K.A., Marcheseragona, S.P., 1989. Polymerization of Antarctic fish tubulins at low-temperatures - Energetic aspects. Biochemistry, 28, 10085-10093.

Efron, B., Tibshirani, R., 2007. On testing the significance of sets of genes. Annals of Applied Statistics, 1, 107-129.

Feder, M.E., Hofmann, G.E., 1999. Heat-shock proteins, molecular chaperones, and the stress response: Evolutionary and ecological physiology. Annual Review of Physiology, 61, 243-282.

Feder, M.E., Walser, J.C., 2005. The biological limitations of transcriptomics in elucidating stress and stress responses. Journal of Evolutionary Biology, 18, 901-910.

Girardot, F., Monnier, V., Tricoire, H., 2004. Genome wide analysis of common and specific stress responses in adult Drosophila melanogaster. BMC Genomics, 5, 16.

Hahn, D.A., Ragland, G.J., Shoemaker, D.D., Denlinger, D.L., 2009. Gene discovery using massively parallel pyrosequencing to develop ESTs for the flesh fly Sarcophaga crassipalpis. BMC Genomics, 10.

Hanson, R.W., Reshef, L., 1997. Regulation of phosphoenolpyruvate carboxykinase (GTP) gene. Annual Review of Biochemistry, 66, 581-611.

Hoffmann, A.A., 2010. Physiological climatic limits in Drosophila: patterns and implications. Journal of Experimental Biology, 213, 870-880.

Huang, D.W., Sherman, B.T., Lempicki, R.A., 2009. Bioinformatics enrichment tools: paths toward the comprehensive functional analysis of large gene lists. Nucleic Acids Research, 37, 1-13.

Joplin, K.H., Yocum, G.D., Denlinger, D.L., 1990. Cold shock elicits expression of heat- shock proteins in the flesh fly, Sarcophaga crassipalpis. Journal of Insect Physiology, 36, 825-834.

Kanehisa, M., Goto, S., 2000. KEGG: Kyoto Encyclopedia of Genes and Genomes. Nucleic Acids Research, 28, 27-30.

Kelty, J.D., Lee, R.E., 1999. Induction of rapid cold hardening by cooling at ecologically relevant rates in Drosophila melanogaster. Journal of Insect Physiology, 45, 719-726.

Kostal, V., Korbelova, J., Rozsypal, J., Zahradnickova, H., Cimlova, J., Tomcala, A., Simek, P., 2011a. Long-term cold acclimation extends survival time at 0 degrees C and modifies the metabolomic profiles of the larvae of the fruit fly Drosophila melanogaster. Plos One, 6, 10.

Kostal, V., Tollarova-Borovanska, M., 2009. The 70 kDa heat shock protein assists during the repair of chilling injury in the insect, Pyrrhocoris apterus. Plos One, 4, e4546.

Kostal, V., Tollarova, M., Sula, J., 2004. Adjustments of the enzymatic complement for polyol biosynthesis and accumulation in diapausing cold-acclimated adults of Pyrrhocoris apterus. Journal of Insect Physiology, 50, 303-313.

Kostal, V., Zahradnickova, H., Simek, P., 2011b. Hyperprolinemic larvae of the drosophilid fly, Chymomyza costata, survive cryopreservation in liquid nitrogen. Proceedings of the National Academy of Sciences of the United States of America, 108, 13041-13046.

Landis, G.N., Abdueva, D., Skvortsov, D., Yang, J.D., Rabin, B.E., Carrick, J., Tavare, S., Tower, J., 2004. Similar gene expression patterns characterize aging and oxidative stress in Drosophila melanogaster. Proceedings of the National Academy of Sciences of the United States of America, 101, 7663-7668.

Lee, R.E., 2010. A primer on insect cold tolerance, in: Denlinger, D.L., Lee, R.E. (Eds.), Low Temperature Biology of Insects. Cambridge University Press, Cambridge, pp. 3-34.

Lee, R.E., Chen, C.P., Denlinger, D.L., 1987. A rapid cold-hardening process in insects. Science, 238, 1415-1417.

Lemaitre, B., Nicolas, E., Michaut, L., Reichhart, J.M., Hoffmann, J.A., 1996. The dorsoventral regulatory gene cassette spatzle/Toll/cactus controls the potent antifungal response in Drosophila adults. Cell, 86, 973-983.

Li, A., Denlinger, D.L., 2008. Rapid cold hardening elicits changes in brain protein profiles of the flesh fly, Sarcophaga crassipalpis. Insect Molecular Biology, 17, 565-572.

Liu, G.W., Roy, J., Johnson, E.A., 2006. Identification and function of hypoxia-response genes in Drosophila melanogaster. Physiological Genomics, 25, 134-141.

Michaud, M.R., Denlinger, D.L., 2007. Shifts in the carbohydrate, polyol, and amino acid pools during rapid cold-hardening and diapause-associated cold-hardening in flesh flies

(Sarcophaga crassipalpis): a metabolomic comparison. Journal of Comparative Physiology. B, Biochemical, Systemic, and Environmental Physiology, 177, 753-763.

Misener, S.R., Chen, C.P., Walker, V.K., 2001. Cold tolerance and proline metabolic gene expression in Drosophila melanogaster. Journal of Insect Physiology, 47, 393-400.

Qin, W., Neal, S.J., Robertson, R.M., Westwood, J.T., Walker, V.K., 2005. Cold hardening and transcriptional change in Drosophila melanogaster. Insect Molecular Biology, 14, 607-613.

Smyth, G.K., 2004. Linear models and empirical Bayes methods for assessing differential expression in microarray experiments. Statistical Applications in Genetics and Molecular Biology, 3, Article 3.

Sorensen, J.G., Nielsen, M.M., Kruhoffer, M., Justesen, J., Loeschcke, V., 2005. Full genome gene expression analysis of the heat stress response, in Drosophila melanogaster. Cell Stress and Chaperones, 10, 312-328.

Storey, J.M., Storey, K.B., 1990. Carbon balance and energetics of cryoprotectant synthesis in a freeze-tolerant insect - responses to perturbation by anoxia. Journal of Comparative Physiology. B, Biochemical, Systemic, and Environmental Physiology, 160, 77-84.

Takashima, S., Mkrtchyan, M., Younossi-Hartenstein, A., Merriam, J.R., Hartenstein, V., 2008. The behaviour of Drosophila adult hindgut stem cells is controlled by Wnt and Hh signalling. Nature, 454, 651-U658.

Telonis-Scott, M., Hallas, R., McKechnie, S.W., Wee, C.W., Hoffmann, A.A., 2009. Selection for cold resistance alters gene transcript levels in Drosophila melanogaster. Journal of Insect Physiology, 55, 549-555.

Terhzaz, S., Finlayson, A.J., Stirrat, L., Yang, J.L., Tricoire, H., Woods, D.J., Dow, J.A.T., Davies, S.A., 2010. Cell-specific inositol 1,4,5 trisphosphate 3-kinase mediates epithelial cell apoptosis in response to oxidative stress in Drosophila. Cellular Signalling, 22, 737-748.

Thomashow, M.F., 1999. Plant cold acclimation: Freezing tolerance genes and regulatory mechanisms. Annual Review of Plant Physiology and Plant Molecular Biology, 50, 571- 599.

Wu, Q., Brown, M.R., 2006. Signaling and function of insulin-like peptides in insects. Annual Review of Entomology, 51, 1-24.

Xia, J.G., Psychogios, N., Young, N., Wishart, D.S., 2009. MetaboAnalyst: a web server for metabolomic data analysis and interpretation. Nucleic Acids Research, 37, W652- W660.

Yancey, P.H., 2005. Organic osmolytes as compatible, metabolic and counteracting cytoprotectants in high osmolarity and other stresses. Journal of Experimental Biology, 208, 2819-2830.

Yi, S.-X., Lee, R.E., 2011. Rapid cold-hardening blocks cold-induced apoptosis by inhibiting the activation of pro-caspases in the flesh fly Sarcophaga crassipalpis. Apoptosis, 16, 249-255.

Yi, S.-X., Moore, C.W., Lee, R.E., 2007. Rapid cold-hardening protects Drosophila melanogaster from cold-induced apoptosis. Apoptosis, 12, 1183-1193.

Zhou, D., Xue, J., Lai, J.C.K., Schork, N.J., White, K.P., Haddad, G.G., 2008. Mechanisms underlying hypoxia tolerance in Drosophila melanogaster: hairy as a metabolic switch. Plos Genetics, 4.

Tables

Table 2.1 – Number of differentially expressed probes in each pairwise comparison. Probes that were considered significant had FDR<0.05, while the 1.5X columns contain probes that were both significant (FDR<0.05) and 1.5-fold up- or down-regulated. For each comparison, we measured the expression of 15,558 distinct ESTs.

Comparison FDR <0.05 1.5X up 1.5X down Control vs. RCH 0 0 0 Control vs. CS+R 1,378 103 8 Control vs. RCH+CS+R 1,525 125 9 CS+R vs. RCH+CS+R 5 0 0

Table 2.2. DAVID enrichment analysis of the differentially expressed ESTs in the Control vs. CS+2R comparison. ESTs with FDR<0.05 and mapped to a Drosophila melanogaster protein database using blastx (E-value cutoff = 1E-4) were included in the enrichment analysis. The “clustered observations” were obtained using the DAVID functional annotation clustering tool to cluster similar GO terms into “clustered observations.” The “unclustered observations” were not placed into a functional cluster by the DAVID analysis. All clustered observations contained at least one GO term with FDR<0.05, while each of the unclustered observations had FDR<0.05.

Enrichment # genes Description Type of GO term score represented

Clustered Vesicle mediated transport, endocytosis, biological process 5.1 59 observations phagocytosis

Nucleotide binding molecular function 4.4 123

Cell adhesion biological process 3.9 31

Actin cytoskeleton organization biological process 3.4 55

Cell morphogenesis biological process 2.8 84

Tracheal system development biological process 2.5 28

Cytoskeleton-dependent intracellular biological process 2.4 17 transport Cytoskeleton organization biological process 2.4 61

Programmed cell death biological process 2.3 32

Unclustered Cytoskeletal protein binding molecular function 2.6 39 observations

Actin binding molecular function 3 26

Protein localization biological process 1.8 47

Unfolded protein binding molecular function 3 16

Calcium ion binding molecular function 1.9 33

Negative regulation of signal transduction biological process 2.7 18

Negative regulation of cell communication biological process 2.6 18

Response to heat biological process 3.1 14

Table 2.3. DAVID enrichment analysis of the differentially expressed ESTs in the Control vs. RCH+CS+2R comparison. ESTs with FDR<0.05 and mapped to a Drosophila melanogaster protein database using blastx (E-value cutoff = 1E-4) were included in the enrichment analysis. The “clustered observations” were obtained using the DAVID functional annotation clustering tool to cluster similar GO terms into “clustered observations.” The “unclustered observations” were not placed into a functional cluster by the DAVID analysis. All clustered observations contained at least one GO term with FDR<0.05, while each of the unclustered observations had FDR<0.05.

Enrichment # genes Description Type of GO term score represented Clustered Vesicle mediated transport, endocytosis, biological process 7.1 71 Observations phagocytosis Nucleotide binding molecular function 6.3 138

Response to temperature stress biological process 4 23

Actin cytoskeleton organization biological process 3.8 64

Epithelial development biological process 3.4 53

Cytoskeleton cellular component 3.2 86

Cytoskeleton organization biological process 2.9 97

Post-embryonic development biological process 2.9 63

RNAi mediated gene silencing biological process 2.8 46

Cell morphogenesis biological process 2.7 94

Cell adhesion biological process 2.6 21

Unclustered Cytoskeletal protein binding molecular function 2.8 46 Observations Actin binding molecular function 3.2 29

Cell adhesion biological process 2.8 34

Biological adhesion biological process 2.6 34

Regulation of cell shape biological process 3.3 23

Negative regulation of signal biological process 3.1 22 transduction Unfolded protein binding molecular function 3.4 19

Regulation of cell morphogenesis biological process 2.9 24

Negative regulation of cell biological process 3 22 communication Protein localization biological process 1.7 47

Protein folding biological process 2.5 20

Table 2.4. Gene set analysis of genes involved in recovery from cold shock. Log2 intensity values for each probe that mapped to a Drosophila melanogaster gene (E-value <1E-5) were included in the analysis. Gene sets were obtained from the KEGG database and from a priori lists generated from other microarray studies. Each gene set included in the table has FDR<0.1.

Category Geneset # of genes Score FDR measured KEGG pathway Valine, leucine, and isoleucine biosynthesis 8 2.57 <0.0001 KEGG pathway Dorso-ventral axis formation 11 0.75 <0.0001 KEGG pathway Jak/STAT signaling pathway 12 1.38 <0.0001 Drosophila hypoxia response (Liu et al., A priori 55 2.61 <0.0001 2006) Drosophila hyperoxia response (Landis et al., A priori 57 1.64 <0.0001 2004) A priori Drosophila cold stress (Qin et al., 2005) 14 3.05 <0.0001 Insulin receptor signaling pathway (Wu and A priori 9 0.95 <0.0001 Brown, 2006) KEGG pathway Urea cycle and metabolism of amine groups 18 -1.54 <0.0001 A priori Drosophila heat stress( Sorensen et al., 2005) 81 1.66 <0.0001 Drosophila oxidative stress response A priori 369 0.89 0.013 (Girardot et al., 2004) KEGG pathway Pentose and glucuronate interconversions 16 1.01 0.013 Drosophila ecdysone signaling( Beckstead et A priori 274 0.74 0.023 al., 2005) KEGG pathway Phosphatidylinositol signaling 25 1.06 0.042 KEGG pathway Wnt signaling 39 0.90 0.049 KEGG pathway VEGF signaling 28 0.83 0.059 KEGG pathway TGF-beta signaling 18 0.80 0.075 KEGG pathway Inositol phosphate metabolism 19 0.90 0.075 KEGG pathway Pyruvate metabolism 29 1.01 0.082 KEGG pathway p53 signaling 11 0.86 0.084 KEGG pathway Glycerolipid metabolism 34 0.83 0.093

Table 2.5. Gene set analysis of genes enriched in the C vs. RCH+CS+2R comparison. Log2 intensity values for each probe that mapped to a Drosophila melanogaster gene (E-value <1E-5) were included in the analysis. Gene sets were obtained from the KEGG database and from a priori lists generated from other microarray studies. Each gene set included in the table has FDR<0.1.

# of genes Category Geneset measured Score FDR Valine, leucine, and isoleucine KEGG pathway 8 2.57 <0.0001 biosynthesis Drosophila hypoxia response (Liu et al., A priori 55 2.57 <0.0001 2006) A priori Drosophila cold stress (Qin et al., 2005) 14 2.78 <0.0001 Drosophila heat stress (Sorensen et al., A priori 81 1.47 <0.0001 2005) Urea cycle and metabolism of amine KEGG pathway 18 -1.54 <0.0001 groups Drosophila hyperoxia response (Landis et A priori 57 1.61 0.011 al., 2004) KEGG pathway Jak/STAT signaling pathway 12 1.37 0.016 KEGG pathway Pyruvate metabolism 29 1.01 0.016 Drosophila ecdysone signaling (Beckstead A priori l 274 0.84 0.035 et al., 2005) Drosophila oxidative stress response A priori 369 0.87 0.044 (Girardot et al., 2004) KEGG pathway Pentose and glucuronate interconversions 16 0.88 0.044 KEGG pathway Inositol phosphate metabolism 19 0.94 0.044 KEGG pathway Phosphatidylinositol signaling 25 0.98 0.044 Drosophila reproductive diapause( Baker A priori 257 0.63 0.045 and Russell, 2009) KEGG pathway Dorso-ventral axis formation 11 0.96 0.055 Insulin receptor signaling pathway (Wu A priori 9 0.99 0.063 and Brown, 2006) KEGG pathway Wnt signaling 39 0.84 0.079

Table 2.6. Expression of genes involved in Drosophila cold stress during recovery from cold shock in S. bullata. This list of genes, identified as significantly enriched using gene set analysis (GSA), was obtained from a microarray study of Drosophila cold stress (Qin et al., 2005). The column “GSA gene score” is a modified t-statistic that reflects the relative importance of a particular transcript towards the overall enrichment of that gene set.

EST Accession Description Drosophila Blastx E- GSA Log2 FDR RefSeq value gene FC Homolog score CG8026, isoform B [Drosophila EZ605491 NP_610468.1 3.00E-72 15.48 0.90 5.56E-11 melanogaster] Sarcophaga crassipalpis 23kDa heat shock U96099.2 NA NA 11.41 1.74 3.74E-11 protein ScHSP23 mRNA, complete cds CG15745, isoform A [Drosophila EZ598021 NP_572873.1 6.98E-10 7.13 0.63 4.16E-09 melanogaster] EZ597482 Ubiquitin-5E [Drosophila melanogaster] NP_727078.1 4.90E-127 5.67 0.39 1.29E-05 SRY interacting protein 1 [Drosophila EZ601452 NP_524712.1 1.79E-08 4.30 0.12 2.45E-01 melanogaster] pinocchio, isoform A [Drosophila SRR006884.66098 NP_608568.1 8.51E-37 3.93 0.90 1.37E-05 melanogaster] CG3814, isoform A [Drosophila SRR006884.70084 NP_610824.1 1.20E-06 2.75 0.14 3.44E-01 melanogaster] EZ601126 CG3345 [Drosophila melanogaster] NP_608509.1 1.67E-06 1.66 0.21 5.46E-01 draper, isoform B [Drosophila EZ604149 NP_728660.2 1.97E-25 0.36 0.03 7.78E-01 melanogaster] CG15347, isoform A [Drosophila EZ610357 NP_572479.1 8.11E-63 -0.17 -0.02 8.83E-01 melanogaster] CG2118, isoform A [Drosophila SRR006884.112334 NP_651896.1 6.55E-08 -0.87 0.00 9.82E-01 melanogaster] mitochondrial acyl carrier protein 1, EZ599928 NP_477002.1 1.13E-13 -1.05 -0.04 6.88E-01 isoform B [Drosophila melanogaster] SRR006884.85625 Ect3 [Drosophila melanogaster] NP_650142.1 2.30E-13 -1.12 -0.07 6.29E-01 EZ600486 CG8778 [Drosophila melanogaster] NP_610805.1 1.45E-108 -1.14 -0.05 6.73E-01

Table 2.7. Metabolite content in response to RCH and cold shock. Metabolite contents are expressed as mean ± SE of the metabolite content in nmol metabolite mg-1 fresh mass. In each row, different letters represent significant differences (ANOVA, FDR<0.05) between groups for a particular metabolite.

Metabolite content (nmol/mg fresh mass) Metabolite Control RCH CS+2R RCH+CS+2R CS+24R RCH+CS+24R Valine 1.62±0.09a,b 1.60±0.13b 1.97±0.07c 1.87±0.09a,c 2.76±0.13d 2.73±0.09d Glycine 2.86±0.07a,b 2.73±0.10b 3.43±0.09c 3.16±0.12a,c 4.44±0.18d 4.66±0.18d Serine 1.62±0.08a 1.57±0.10a 2.13±0.09b 1.91±0.08b 3.03±0.17c 3.19±0.19c Glutamate 4.43±0.35a,b 4.59±0.28a 5.15±0.18a 4.69±0.31a 3.55±0.22c 3.63±0.13b,c Proline 5.66±0.22a 6.02±0.24a,b,c 6.27±0.42a,b,c 5.95±0.33a,c 7.03±0.36b 6.82±0.37b,c Leucine 1.08±0.06a 1.02±0.05a 1.29±0.06b 1.13±0.04a,b 1.86±0.07c 1.88±0.07c Isoleucine 0.65±0.02a 0.63±0.03a 0.84±0.04b 0.73±0.02c 1.35±0.04d 1.37±0.05d Threonine 0.62±0.03a 0.63±0.05a 0.87±0.06b 0.74±0.03b 1.48±0.07c 1.55±0.07c Alanine 0.64±0.02a 0.57±0.04a 0.63±0.06a 0.64±0.05a 1.4±0.06b 1.34±0.10b Phenylalanine 0.48±0.03a 0.50±0.02a,b 0.57±0.02b 0.53±0.03a,b 0.93±0.03c 0.88±0.03c Ribose 1.77E-02±1.2E-03a 1.60E-02±1.2E-03a 1.84E-02±5.1E-04a 1.66E-02±8.2E-04a 1.77E-02±8.1E-04a 1.70E-02±7.4E-04a Glucose 0.52±0.04a 0.59±0.04a 2.75±0.61b 1.47±0.11c 6.15±0.63d 4.12±0.47e Fructose 3.50E-03±2.0E-04a 2.88E-03±9.4E-05a 2.22E-02±2.8E-03b 1.51E-02±8.1E-04c 0.23±0.03d 0.13±0.01e Mannose 3.61E-02±9.1E-04a 3.82E-02±1.2E-03a 0.11±0.01b 7.81E-02±3.7E-03c 0.30±0.03d 0.22±0.02e Maltose 8.23E-02±7.6E-03a 8.26E-02±7.0E-03a 0.27±0.04b 0.18±0.02c 0.12±8.7E-3d 9.36E-02±6.6E-03a,d Trehalose 7.93±0.32a,b 7.38±0.33b 8.70±0.26a,c 8.33±0.30a,c 9.15±0.37c,d 9.92±0.32d Glycerol 2.69E-02±2.0E-03a 2.70E-02±2.2E-03a 4.11E-02±4.9E-03b 3.08E-02±1.4E-03a 0.18±6.5E-03c 0.18±8.5E-03c Erythritol 8.46E-02±2.8E-03a 8.50E-02±3.5E-03a 9.60E-02±3.9E-03a 8.98E-02±3.7E-03a 0.14±5.9E-03b 0.14±6.8E-03b Xylitol 6.62E-03±1.7E-04a 6.08E-03±2.5E-04a 1.21E-02±1.4E-03b 8.32E-03±3.53E-04c 2.32E-02±1.0E-03d 2.44E-02±2.1E-03d Arabitol 1.99E-02±6.5E-04a 1.98E-02±8.0E-04a 2.33E-02±8.7E-04b,c 2.11E-02±9.8E-04a,c 2.58E-02±1.1E-03d 2.80E-02±1.3E-03d Ribitol 3.10E-03±1.5E-04a 3.09E-03±1.4E-04a 4.20E-03±2.3E-04b 3.74E-03±1.8E-04b 1.47E-02±7.1E-04c 1.52E-02±1.2E-03c Galactitol 5.14E-02±3.2E-03a 5.21E-02±3.63E-03a 6.90E-02±4.0E-03b 5.74E-02±2.4E-03a,b 9.70E-02±6.2E-03c 9.52E-02±5.5E-03c Inositol 0.25±0.01a,b 0.24±0.01b 0.33±0.01c 0.28±0.01a 0.53±0.02d 0.57±0.03d Sorbitol 6.26E-03±5.8E-04a 5.65E-03±4.1E-04a 3.88E-02±7.4E-03b 2.10E-02±1.2E-03c 0.61±0.07d 0.46±0.06d Glucose-6-phosphate 0.25±7.4E-03a 0.37±0.01b 0.50±0.04c 0.41±0.01b,d 0.39±5.5E-03b 0.44±8.5E-03c,d Fructose-6-phosphate 0.17±2.8E-03a 0.19±4.3E-03b 0.21±5.8E-03c 0.20±2.3E-03b 0.20±2.2E-03b 0.20±3.4E-03b Citrate 3.87±0.15a 3.67±0.17a 4.17±0.18a 4.02±0.15a 1.20±0.12b 1.20±0.09b Succinate 1.19±0.04a,b 1.06±0.05b 1.43±0.05c 1.24±0.05a 1.22±0.08a,b 1.36±0.06a,c Fumarate 0.80±0.03a,b 0.75±0.03b 0.96±0.03c 0.89±0.03a,c 1.49±0.06d 1.58±0.05d Malate 3.22±0.07a 3.21±0.12a 3.70±0.12b 3.48±0.11a,b 7.51±0.28c 7.91±0.23c Glycerol-3-phosphate 6.00±0.18a,b 5.74±0.20b 7.11±0.21c 6.60±0.22a,c 8.05±0.28d 9.18±0.32d Phosphate 1.46±0.07a 1.46±0.08a 1.77±0.08b 1.52±0.08a 2.54±0.13c 2.75±0.14c Putrescine 2.21E-02±2.5E-03a,b 1.81E-02±3.0E-03b 2.63E-02±2.4E-03a 2.27E-02±3.3E-03a,b 2.83E-02±3.6E-03a 4.34E-02±4.1E-03c Cadaverine 2.55E-02±3.1E-03a,b 2.19E-02±4.0E-03b 2.97E-02±2.8E-03a 2.50E-02±3.2E-03a,b 3.16E-02±3.4E-03a 5.42E-02±5.2E-03c Glucono-delta-lactone 0.31±0.02a,b 0.28±0.02b 0.42±0.03c 0.36±0.03a,c 0.94±0.05d 0.78±0.06d Total amino acids 14.00±0.67a 13.84±0.64a 16.87±0.45b 15.39±0.62a,b 20.81±0.87c 21.25±0.76c Total sugars 8.59±0.32a 8.12±0.24a 11.87±0.71b 10.10±0.37c 15.96±0.99d 14.50±0.48d Total polyols 0.45±0.02a,b 0.43±0.02b 0.61±0.03c 0.51±0.02a 1.61±0.10d 1.52±0.10d

Table 2.8. Metabolic pathways modulated by RCH. Pairwise metabolite pathway enrichment analysis was conducted on the log2 metabolite contents for each comparison. Significantly enriched pathways with FDR<0.05 are included in the table.

# of represented Comparison Pathway metabolites Impact FDR

Control v. RCH Glycolysis or gluconeogenesis 3 1.43E-01 2.34E-03

Amino sugar and nucleotide sugar 4 1.53E-01 2.34E-03 metabolism Pentose phosphate pathway 5 8.99E-02 1.84E-02

Galactose metabolism 8 9.34E-02 2.21E-02

Starch and sucrose metabolism 6 8.77E-02 2.99E-02

CS+2R v. NONE RCH+CS+2R

CS+24R v. Starch and sucrose metabolism 6 8.77E-02 3.75E-02 RCH+CS+24R

Lysine degredation 2 1.46E-02 3.75E-02

Amino sugar and nucleotide sugar 4 1.53E-01 3.75E-02 metabolism Pentose phosphate pathway 5 8.99E-02 3.75E-02

Glycolysis or gluconeogenesis 3 1.43E-01 3.75E-02

Galactose metabolism 8 9.34E-02 3.75E-02

Fructose and Mannose metabolism 4 1.58E-01 3.75E-02

Glutathione metabolism 4 2.81E-02 4.46E-02

Table 2.9. Summary of biochemical pathways enriched in both the transcriptomic and metabolomic datasets during recovery from cold shock. KEGG pathways that were significantly enriched (FDR<0.1) in the Control vs. CS+2R comparison using both GSA for the gene expression data and metabolite pathway enrichment analysis for the metabolimics data are included.

# genes # metabolites Metabolite Biochemical Pathway measured GSA FDR measured pathway FDR Valine, leucine, and isoleucine biosynthesis 8 <1.00E-04 4 1.64E-03 Pentose and glucuronate interconversions 16 0.013 3 1.06E-04 Inositol phosphate metabolism 20 0.075 1 1.04E-03 Pyruvate metabolism 29 0.082 1 6.23E-03

Figures

Figure 2.1. Temperature treatments (A-C) and hybridization design (D) for microarray experiment. Temperature treatments are depicted in A-C, with an arrow depicting the time of sampling for each treatment. In (A), the solid line depicts control conditions while the dashed line indicates RCH conditions. In (D), treatments connected with a double arrow were hybridized on the same chip, N=4 biological replicates for each comparison.

Figure 2.2. Effect of rapid cold-hardening (RCH) on the cold tolerance of pharateadult flesh flies. Cold shocked flies were directly exposed to the indicated test temperature for 2 h, while flies in the RCH groups were exposed to 0°C for 2h prior to being transferred to the test temperature. Flies that successfully eclosed as adults were considered alive. Different letters represent significant differences between groups (ANOVA, Tukey, P<0.05).

Figure 2.3. Heat map showing expression patterns of the top 150 most differentially expressed ESTs. Expression values are given as the log2 ratio of each treatment relative to the control sample hybridized on the same chip. All control values are set to 0. The phenotypes (horizontal axis) and probes (vertical axis) are separated with hierarchical clustering, and each distinct cluster of samples is indicated by a colored bar.

Figure 2.4. Principal components analysis (A) and hierarchical clustering (B) of the entire microarray dataset. Input data were the log2 intensity values for each individual sample. In (B), the dendogram is scaled to represent the distance between each branch. The distinct cluster containing control and RCH samples is highlighted in blue, while the cluster containing CS+2R and RCH+CS+2R samples is highlighted in red.

Figure 2.5. Relative changes in metabolite contents in response to RCH and cold shock. Metabolite contents are expressed as the mean ±SE fold change of each metabolite relative to control. An “*”indicates a significant difference (ANOVA, FDR<0.05) between that treatment and the control within a particular metabolite.

Figure 2.6. Heat map diagram (A) and principal components analysis (B) of the entire metabolomics dataset. In (A), the colors represent the log2 fold change of each metabolite relative to the mean control level. Individual samples (horizontal axis) and compounds (vertical axis) are separated using hierarchical clustering, with the dendrogram scaled to represent the distance between each branch. The cluster containing control and RCH groups is highlighted in green, the cluster containing CS+2R and RCH+CS+2R groups is highlighted in orange, while the cluster containing the CS+24R and RCH+CS+24R groups is highlighted in red. In (B) the input data consisted of the log2 metabolite content for each compound measured in each sample.

Chapter 3: Calcium Signaling Mediates Insect Cold Sensing

Abstract

The ability to quickly respond to changes in temperature is critical for insects and other ectotherms in thermally variable environments. In a process called rapid cold-hardening

(RCH), insects dramatically enhance cold tolerance following brief (i.e. minutes to hours) exposure to non-lethal low temperature. RCH occurs during exposure to ecologically relevant conditions and preserves essential functions such as courtship and reproduction.

At the molecular level, RCH blocks apoptosis and activates MAP kinase signaling pathways, and recent experiments demonstrated that RCH occurs ex vivo in the absence of nervous or hormonal stimulation. However, the signaling mechanism responsible for cellular cold-sensing and the triggering of RCH is unknown. Here, we demonstrate that calcium signaling pathways govern cellular cold-sensing and RCH. In tracheal cells of the freeze-tolerant goldenrod gall fly larvae (Eurosta solidaginis), lowering the temperature from 25 to 0°C at 1°C/min evoked a 40% increase in intracellular calcium levels. This elevation of calcium activated calcium-calmodulin dependent protein kinase

II (CaMKII), and further experiments showed that pharmacological inhibition of both calcium entry and calmodulin/CaMKII activity prevented cellular-level RCH. Similar results were obtained for a freeze-intolerant species, adults of the flesh fly Sarcophaga bullata, suggesting activation of calcium signaling pathways is a general feature of RCH in insects. Our results provide an avenue by which insect tissues can instantly respond to

sudden changes in temperature and activate downstream cold-hardening mechanisms.

RCH is one of the fastest responses to environmental stress that has been characterized, and these experiments provide a key link to other signaling pathways and physiological processes that occur during RCH.

Introduction

Low temperature is one of the primary constraints for insects and other ectotherms living in temperate and polar regions (Denlinger and Lee, 1998). While seasonal adaptations to cold stress, including environmentally programmed periods of dormancy called diapause, have been well-studied (Hahn and Denlinger, 2011), physiological responses to sudden changes in temperature have received less attention. In a physiological process termed rapid cold-hardening (RCH), insects dramatically enhance their cold tolerance in a matter of minutes to hours (Lee et al., 1987). For example, in the flesh fly, Sarcophaga crassipalpis, the first species in which RCH was described, exposure to 0°C for as little as 30 minutes significantly enhances cold tolerance at -10°C.

RCH has since been described in dozens of insect species (Lee and Denlinger, 2010), including both freeze-intolerant (insects in which internal ice formation is lethal) and freeze-tolerant species (insects that tolerate internal ice formation) (Everatt et al., 2012;

Lee et al., 2006b). Naturally occurring thermoperiods can elicit RCH (Kelty, 2007), and

RCH preserves essential functions such as courtship and mating (Rinehart et al., 2000;

Shreve et al., 2004), supporting the relevance of this process to natural populations.

While the ecological relevance of RCH has been established, the physiological mechanisms are poorly understood. RCH causes a slight increase in the cryoprotectant glycerol (Lee et al., 1987), although the amount does not seem sufficient to significantly enhance cold tolerance. More recent experiments have demonstrated an increase in membrane fluidity in response to RCH (Lee et al., 2006a; Michaud and Denlinger, 2006;

Overgaard et al., 2005), as well as upregulation of a single heat shock protein in the brain

(Li and Denlinger, 2008). However, changes in gene expression do not appear to be a major driver of RCH; in the flesh fly, S. bullata, no transcripts (out of ~15,000 tested) were differentially expressed following two hours of RCH (Teets et al., 2012). Rather,

RCH appears to be a cell-autonomous process regulated and executed by second messenger signaling pathways. Isolated tissues retain the capacity for RCH ex vivo (Yi and Lee, 2004), and signaling pathways such as MAP kinase (Fujiwara and Denlinger,

2007) and apoptotic signaling (Yi and Lee, 2011; Yi et al., 2007) are activated by RCH.

However, the cellular cold-sensing mechanism that triggers RCH is unknown. Cold acclimation in plants, which occurs over the course of days to weeks (Guy, 1990), is regulated by cold-induced calcium influx, which activates downstream cold acclimation mechanisms (Knight et al., 1996). In addition, fruit flies (Drosophila melanogaster) with a mutant copy of the membrane protein dystroglycan have elevated levels of intracellular calcium, which correlates with improved performance and survival at suboptimal temperatures. We recently provided pharmacological evidence that calcium and calmodulin are required for RCH in the Antarctic midge, Belgica antarctica (Teets et al.,

2008), but cold-induced calcium signaling has not been examined in detail in insects.

Here, we provide several lines of evidence that RCH activates calcium signaling

pathways, and that these pathways are required for RCH to occur. RCH is the fastest cold hardening process that has been described, and we provide clear evidence of the underlying cold-sensing mechanisms.

Materials and Methods

Animals

Galls containing third instar larvae of E. solidaginis were collected from various goldenrod fields in central and southwest Ohio from September to January, 2007 to 2012.

Galls were stored at 18°C and larvae were removed just prior to experimentation. Flesh flies, S. bullata, were lab-reared according to Denlinger et al. (1972) under non- diapausing conditions (25°C, 16:8 L:D). Adults were fed sugar and liver ad lib, and males were used for experiments 4-8 d after eclosion.

Calcium imaging

Tissues from E. solidaginis and S. bullata were dissected in Coast’s solution

(Coast and Krasnoff, 1988) containing (in mM) 100 NaCl, 8.6 KCl, 4.0 NaHCO3, 4.0

NaH2PO4 -H2O, 1.5 CaCl2-2H2O, 8.5 MgCl2 -6H2O, 24 glucose, 25 HEPES, and 56 sucrose. For experiments with E. solidaginis the Coast’s solution was supplemented with

250 mM glycerol. After dissection, tissues were loaded with 10 µM fluo-3 AM (Life

Technologies, Grand Island, NY) for 1 h at room temperature. The loading solution contained 0.2% pluronic F-127 (Life Technologies, Grand Island, NY) to disperse the dye. After loading, cells were rinsed twice in Coast’s solution and imaged within the next

3 hours. Preliminary experiments established that tracheal cells from E. solidaginis and

salivary gland cells from S. bullata were most amenable to calcium imaging. Other tissues were difficult to maintain in the proper focal plane or failed to take up the calcium imaging dye. Tissues were imaged with a Visitech Infinity3 Hawk 2D Array live-cell confocal microscope at the Ohio State Campus Microscopy and Imaging Facility. Cells were excited at 488 nm, and fluorescence at 520 nM was measured every 10 s for the duration of the experiment. Temperature was controlled with an Instec TSA02i inverted microscope thermal stage (Instec, Boulder, CO). While images were acquired every 10 s, low temperature caused a slight drift in the focal plane in most samples, necessitating periodic refocusing. As a result, samples were refocused every 2.5- 5 min, so results are only presented in 2.5-5 min increments. Fluorescence was recorded in a region of interest that consisted of a single tracheal end cell in E. solidaginis and 4-5 salivary gland epithelial cells in S. bullata. Intracellular calcium concentration was calculated according

2+ 2+ the following equation [Ca ]i= Kd(F/Fo)/(Kd/[Ca ]i-rest+1-F/Fo), where F is the

2+ fluorescence, Fo is the initial fluorescence, and [Ca ]i-rest is the resting calcium concentration, which was estimated to be 100 nm, based on measurements of blow fly salivary glands conducted by Zimmermann and Walz (1999). The Kd for fluo-3 was corrected for changes in temperature according to Woodruff et al. (2002)

CaMKII assays

CaMKII activity was measured with the Signatect CaMKII Assay Kit (Promega,

Madison, WI). In this assay, CaMKII transfers phosphate from [γ-32P]ATP to a synthetic substrate for CaMKII, which is immobilized on a biotinylated membrane and measured with a scintillation counter. Protein was extracted from whole larvae of E. solidaginis and adults of S. bullata with RIPA buffer containing the HALT Protease and Phosphatase 99

Inhibitor cocktail (ThermoFisher Scientific, Waltham, MA), and 25 µg of protein was loaded into each reaction. For Western blotting, protein samples were generated in the same manner, and 35 (E. solidaginis) or 40 (S. bullata) µg protein from each sample was loaded into a 4-15% gradient SDS-PAGE gel. Western blotting was conducted according to Yi et al. (2007). Rabbit anti-CaMKII (total) was obtained from Santa Cruz

Biotechnology (Santa Cruz, CA), rabbit anti-CAMKII (pThr305) was obtained from

Millipore (Temecula, CA), while mouse anti-β-tubulin was obtained from the

Developmental Studies Hybridoma Bank, University of Iowa (Iowa City, IA).

Densitometry was conducted with AlphaView SA software (ProteinSimple, Santa Clara,

CA), with each sample normalized to β-tubulin. Each Western was repeated with 6 independent biological replicates. To verify whether mammalian-based assays and antibodies would work, we cloned and characterized full-length calmodulin and CaMKII cDNA sequences using the Clontech SMARTer RACE kit (Clontech Laboratories,

Mountain View, CA). Additionally, in E. solidaginis, we measured tissues specific expression in larval brains, midguts, fat body, salivary glands, Malpighian tubules, and epidermis using RT-PCR.

Cell viability assays

Tissues were dissected in Coast’s solution at room temperature. Tissues from E. solidaginis were exposed to the following temperature conditions: Control (4°C, 3 h), directly frozen (directly transferred from room temperature to -17.5 or -20°C, and held there for 2 h), and RCH (slowly ramped from 4°C to test temperature over 1 h, then held at test temperature for 2 h). For S. bullata, tissues were exposed to control (25°C, 4 h), cold shock (-14°C, 2 h), and RCH (0°C 2h, -14°C 2h) conditions; all tissues remained 100

supercooled at -14°C. Cell viability was assessed with the LIVE/DEAD sperm viability assay(Life Technologies, Grand Island, NY) (Yi and Lee, 2004), which is a membrane integrity assay consisting of SYBR green and propidium iodide. In this assay, living cells with intact membranes fluoresce green, while dead cells with damaged membranes fluoresce red. Viability is expressed as the percentage survival based on the counts of 300 cells per sample. To test whether calcium signaling is essential for RCH, tissues were exposed to the following solutions and drugs to manipulate calcium signaling: Nominally calcium free (NCF) medium, Coast’s solution prepared without calcium; 0.25 mM LaCl3, a general calcium channel blocker; 100 µM BAPTA-AM, an intracellular calcium chelator; 50 µM W-7, an inhibitor of calmodulin; 100 µM KN-93, an inhibitor of

CaMKII, and 10 µM 2-APB, an inhibitor of IP3-mediated calcium signaling. Tissues were loaded with drugs for 30-60 minutes prior to conducting the same temperature experiments described above.

Statistics

All data are expressed as mean ± SE. Data from calcium imaging experiments were analyzed with repeated measures ANOVA and a post-hoc Bonferroni test. All other means were compared with ANOVA and Tukey’s post-hoc multiple comparisons procedure on JMP9 (SAS Institute Inc., Cary, NC). For calcium imaging experiments,

N=2-5 for each experiment, while for CaMKII assays N=5-6 for each treatment. Cell viability assays were conducted with 4 biological replicates per treatment, and data were arcsin square-root transformed prior to analysis.

101

Results and Discussion

The goldenrod gall fly, Eurosta solidaginis, is one of the most cold-hardy temperate species known (Bennett and Lee, 1997) and has long been a model for studying freeze-tolerance. When midgut and salivary gland tissue from third instar larvae were removed and directly frozen at -20°C ex vivo, high mortality resulted; survival of midgut tissue dropped to ~50% while that of the salivary gland was ~60% (Fig. 3.1a).

However, when tissues were first given a 1 h RCH period, in which temperature was gradually lowered from 4°C to -20°C, survival increased by 25-30% in each tissue (Fig.

3.1a). This cold-hardening response occurred ex vivo in the absence of stimulation from nerves or hormones, indicating that RCH in isolated tissues of E. solidaginis is a cell- autonomous process.

As a first step in determining whether calcium might be involved in cellular cold- sensing, we measured intracellular calcium levels in response to changes in temperature.

Using live-cell confocal imaging, we observed a gradual, 40% increase in intracellular calcium in tracheal cells of E. solidaginis when temperature was lowered from 25 to 0°C at 1°C/min (Fig. 3.1b). Lowering temperature from 25 to 20°C raised intracellular calcium by 7%, and calcium rose continuously as temperature decreased. Maintaining temperature at 25°C for the duration of the 40 min experiment failed to elicit a calcium response (Fig. 3.1b), indicating the increase in calcium is attributable to low temperature.

Cold-induced increases in cytosolic calcium have also been demonstrated in plants

(Knight et al., 1996), although cold exposure in plants causes a single, sharp spike in intracellular calcium, suggesting a different mechanism is at play. Elevation of intracellular calcium at low temperature is also known in fish (Shiels et al., 2011) and 102

mammals (Haddad et al., 1999), although not in the context of cold-hardening. We hypothesize that changes in membrane properties (Lee et al., 2006a) and/or reduced activity of calcium-regulating transport mechanisms (e.g. Na+/Ca2+ exchanger and sarcoplasmic/endoplasmic reticulum Ca2+-ATPase) are responsible for calcium flux at low temperature. Cold-sensitive transient receptor potential channels have been described in D. melanogaster (Rosenzweig et al., 2008), although the function of these channels outside the CNS has not been examined.

Downstream of calcium entry into the cell, chilling activates calcium-dependent signaling pathways in E. solidaginis. RCH conditions induced a significant increase in the activity of calcium/calmodulin dependent protein kinase II (CaMKII), a multifunctional signaling protein that regulates numerous metabolic and gene expression processes (Colbran, 2004) from 0.65 to 0.88 pmol min-1 μg protein-1 (Fig. 3.1c). While this protein was activated during RCH, prolonged freezing at -15°C for 24 h returned activity to control levels, and activity remained at control levels following 2 h of recovery at 18°C (Fig. 3.1c). Thus, activation of this protein occurs specifically during the RCH period. Additionally, RCH decreased phosphorylation at the inhibitory Thr306 residue

(Fig. 3.1d, e), which is consistent with enzyme activation (Colbran, 2004). In contrast, chilling at 4°C for 24 h and 2 h recovery from freezing significantly increased the phosphorylation ratio (Fig. 3.1d, e), suggesting deactivation of this enzyme in response to prolonged chilling and recovery from freezing. We cloned and sequenced full-length calmodulin and CaMKII transcripts from E. solidaginis and measured their tissue- specific expression. Predicted amino acid sequences for both proteins were highly conserved and contained the expected domain structure (Fig. 3.2a, 3.3a), and transcripts 103

were expressed in all six larval tissues tested (Fig. 3.2b, 3.3b), indicating these genes could mediate cold-sensing throughout the body.

To determine the functional significance of calcium signaling during cold- sensing, we used pharmacological inhibitors to block various components of calcium signaling pathways during RCH. In midgut and salivary gland tissues of E. solidaginis larvae, removing calcium from the medium, blocking calcium channels with LaCl3, and chelating intracellular calcium with BAPTA-AM all inhibited ex vivo RCH (Fig. 3.4). In addition, blocking calmodulin with the drug W-7 and inhibiting CaMKII with KN-93 similarly reduced cell survival following RCH (Fig. 3.4). However, inhibition of IP3- mediated calcium signaling had no effect on survival, suggesting the observed increase in calcium during chilling is not a result of IP3-mediated calcium release. These drugs were not toxic to tissues held at 4°C and had little effect on the survival of tissues directly frozen at -20°C (Fig. 3.5a). Furthermore, at milder temperatures, inhibition of calcium signaling reduced baseline freezing tolerance, even in the absence of RCH. After direct freezing at -17.5°C, cell survival was significantly higher than at -20°C (Fig. 3.5b), but inhibition of calcium signaling reduced survival at -17.5°C (Fig. 3.5c). Thus, it appears calcium signaling operates not only during RCH but also during rapid responses to freezing at milder temperatures.

The above results clearly demonstrate a role for calcium signaling during RCH in

E. solidaginis, a freeze-tolerant organism. Subsequent experiments indicated that calcium signaling also governs cold-sensing and RCH in a freeze-intolerant species, the flesh fly

Sarcophaga bullata. Chilling to 0°C elicited a 70% increase in intracellular calcium in

104

adult salivary gland tissue (Fig. 3.6a), levels similar to those observed in tracheal cells from E. solidaginis (Fig. 3.1b). Once again, holding tissues at 25°C for the duration of the experiment failed to evoke a calcium response (Fig. 3.6b). Doubling the rate of temperature decrease to 2°C/min elicited a similar calcium response (Fig. 3.6c), although the magnitude was dampened, with calcium changes at a given temperature being 50-

70% of those during the slower ramp. Also, warming back to 25°C from 0°C caused an immediate reversal in calcium flux, nearly returning calcium levels to control levels by the end of the experiment (Fig. 3.6d). Thus, intracellular calcium levels track environmental temperature, which may in part explain the rapid attenuation of RCH upon warming (Coulson and Bale, 1990).

Like E. solidaginis, pharmacological experiments in S. bullata revealed that calcium signaling components are required for RCH to occur ex vivo. Inhibiting calcium entry and blocking activation of calmodulin both prevented RCH in adult midgut and fat body tissues (Fig. 3.7), while having no effect on control tissues and tissues cold shocked at -14°C (Fig. 3.5d). However, in S. bullata, blocking CaMKII with KN-93 failed to significantly inhibit RCH, suggesting activation of CaMKII is not required for RCH in this species, or that pharmacological inhibition was incomplete. Nonetheless, calmodulin and CaMKII were expressed in all post-embryonic developmental stages and in nearly every tissue (Fig. 3.8, 3.9), suggesting this signaling axis, or at the very least calmodulin, could mediate cold sensing throughout the body in every developmental stage.

The results presented here represent a significant advance in our understanding of the cell physiology of insect low temperature response. A hypothetical model for the

105

mode of calcium entry and the downstream targets of calcium signaling in response to low temperature is presented in Fig. 3.10. Calcium signaling provides a clear mechanism by which tissues can directly sense low temperatures, and while the downstream targets of calcium signaling during RCH are currently unknown, potential roles for calcium include cross-talk with apoptosis signaling (Yi and Lee, 2011; Yi et al., 2007) and MAP kinase signaling (Takeda et al., 2004), as well as regulation of carbohydrate mobilization

(Johnson, 1992). On a larger scale, mechanisms of cold-hardening also could be exploited to manipulate populations of insect pests. Our group has recently explored the possibility of disrupting overwintering diapause to control pest populations (Zhang et al.,

2011), and we envision disruption of acute cold-hardening being an equally effective strategy.

Acknowledgements

The authors thank Sarah Cole, Brian Kemmenoe, Richard Montione, and the rest of the Ohio State Campus Microscopy and Imaging Facility for their assistance with these experiments. We also thank Dr. Michael Ibba for providing lab space for experiments requiring radiolabeled substrates. We also acknowledge Josh Stapleton for help with densitometry and other members of the Laboratory for Ecophysiological

Cryobiology at Miami University for assistance with gall collecting.

Grants

This work was supported by NSF IOS-0840772.

106

References

Bennett, V.A., Lee, R.E., 1997. Modeling seasonal changes in intracellular freeze- tolerance of fat body cells of the gall fly Eurosta solidaginis (Diptera, Tephritidae). Journal of Experimental Biology, 200, 185-192.

Coast, G.M., Krasnoff, S.B., 1988. Fluid secretion by single isolated Malpighian tubules of the house cricket, Acheta domesticus, and their response to diuretic hormone. Physiological Entomology, 13, 381-391.

Colbran, R.J., 2004. Targeting of calcium/calmodulin-dependent protein kinase II. Biochemical Journal, 378, 1-16.

Coulson, S.J., Bale, J.S., 1990. Characterization and limitations of the rapid cold- hardening response in the housefly Musca domestica (Diptera, Muscidae). Journal of Insect Physiology, 36, 207-211.

Denlinger, D.L., 1972. Induction and termination of pupal diapause in Sarcophaga (Diptera - Sarcophagidae). Biological Bulletin, 142, 11-&.

Guy, C.L., 1990. Cold-acclimation and freezing stress tolerance - Role of protein- metabolism. Annual Review of Plant Physiology and Plant Molecular Biology, 41, 187- 223.

Haddad, P., Cabrillac, J.C., Piche, D., Musallam, L., Huet, P.M., 1999. Changes in intracellular calcium induced by acute hypothermia in parenchymal, endothelial, and Kupffer cells of the rat liver. Cryobiology, 39, 69-79.

Hahn, D.A., Denlinger, D.L., 2011. Energetics of insect diapause. Annual Review of Entomology, 56, 103-121.

107

Johnson, L.N., 1992. Glycogen-phosphorylase - Control by phosphorylation and allosteric effectors. FASEB Journal, 6, 2274-2282.

Kelty, J., 2007. Rapid cold-hardening of Drosophila melanogaster in a field setting. Physiological Entomology, 32, 343-350.

Knight, H., Trewavas, A.J., Knight, M.R., 1996. Cold calcium signaling in Arabidopsis involves two cellular pools and a change in calcium signature after acclimation. Plant Cell, 8, 489-503.

Lee, R.E., Chen, C.P., Denlinger, D.L., 1987. A rapid cold-hardening process in insects. Science, 238, 1415-1417.

Lee, R.E., Damodaran, K., Yi, S.-X., Lorigan, G.A., 2006a. Rapid cold-hardening increases membrane fluidity and cold tolerance of insect cells. Cryobiology, 52, 459-463.

Li, A., Denlinger, D.L., 2008. Rapid cold hardening elicits changes in brain protein profiles of the flesh fly, Sarcophaga crassipalpis. Insect Molecular Biology, 17, 565-572.

Rinehart, J.P., Yocum, G.D., Denlinger, D.L., 2000. Thermotolerance and rapid cold hardening ameliorate the negative effects of brief exposures to high or low temperatures on fecundity in the flesh fly, Sarcophaga crassipalpis. Physiological Entomology, 25, 330-336.

108

Shiels, H.A., Di Maio, A., Thompson, S., Block, B.A., 2011. Warm fish with cold hearts: thermal plasticity of excitation-contraction coupling in bluefin tuna. Proceedings of the Royal Society B-Biological Sciences, 278, 18-27.

Woodruff, M.L., Sampath, A.P., Matthews, H.R., Krasnoperova, N.V., Lem, J., Fain, G.L., 2002. Measurement of cytoplasmic calcium concentration in the rods of wild-type and transducin knock-out mice. Journal of Physiology-London, 542, 843-854.

Yi, S.-X., Lee, R.E., 2004. In vivo and in vitro rapid cold-hardening protects cells from cold-shock injury in the flesh fly. Journal of Comparative Physiology B-Biochemical Systemic and Environmental Physiology, 174, 611-615.

Yi, S.-X., Lee, R.E., 2011. Rapid cold-hardening blocks cold-induced apoptosis by inhibiting the activation of pro-caspases in the flesh fly Sarcophaga crassipalpis. Apoptosis, 16, 249-255.

Yi, S.-X., Moore, C.W., Lee, R.E., 2007. Rapid cold-hardening protects Drosophila melanogaster from cold-induced apoptosis. Apoptosis, 12, 1183-1193.

Zhang, Q., Nachman, R.J., Kaczmarek, K., Zabrocki, J., Denlinger, D.L., 2011. Disruption of insect diapause using agonists and an antagonist of diapause hormone. Proceedings of the National Academy of Sciences of the United States of America, 108, 16922-16926.

Zimmermann, B., Walz, B., 1999. The mechanism mediating regenerative intercellular Ca2+ waves in the blowfly salivary gland. Embo Journal, 18, 3222-3231.

109

Figures

Figure 3.1. Rapid cold-hardening reduces freezing injury in tissues of the goldenrod gall fly, E. solidaginis and activates calcium signaling pathways. (A) ex vivo RCH (slow ramping from 4 to -20°C over 1 h) prevents injury caused by direct freezing at - 20°C. (B) Chilling from 25 to 0°C elicits an increase in intracellular calcium in tracheal cells, while maintaining cells at 25°C causes no increase in calcium. (C) RCH activates calcium/calmodulin-dependent protein kinase II (CaMKII), with activity returning to control levels following prolonged freezing for 24 h, and freezing with 2 h recovery. (D) Low temperature increased levels of both total CaMKII and phospho-Thr306 CaMKII. (E) The ratio of phospho-Thr306 CaMKII to total CaMKII decreased during RCH, consistent with activation during RCH. All values are mean ± SEM, n ≥ 4, with different letters indicating significant differences (ANOVA, Tukey, p<0.05). In (B), an “*” indicates a significant difference between a particular time point and the initial calcium concentration (repeated measures ANOVA, post-hoc Bonferroni, p<0.05). 110

Figure 3.2. Predicted amino acid sequence and tissue-specific expression of E. solidaginis calmodulin. (A) Amino acid sequence alignement of calmodulin from E. solidaginis, S. bullata, Drosophila melanogaster, and Homo sapiens. (B) Tissue-specific expression of calmodulin measured using RT-PCR. Expression was detectable in every tissue tested. Negative control reactions lacking reverse transcriptase are at the left side of each gel. Tissue abbreviations are as follows: WB = whole body, Br = brain, SG = salivary gland, MT = Malpighian tubules, FB = fat body, Ep = epidermis, and MG = midgut. Midgut samples were run separately because they required more PCR cycles for detection.

111

Figure 3.3. Predicted amino acid sequence and tissue-specific expression of E. solidaginis calcium/calmodulin dependent protein kinase II. (A) Amino acid sequence alignment of calcium/calmodulin dependent protein kinase II (CaMKII) from E. solidaginis, S. bullata, Drosophila melanogaster and Homo sapiens. The asterisks (*) mark the locations of the two autophosphorylation residues, Thr287 (activation) and Thr 305 (inhibition). (B) Tissue-specific expression of CaMKII measured using RT-PCR. Transcripts were detected in every tissue. Negative control reactions lacking reverse transcriptase are at the left side of each gel. Tissue abbreviations are as follows: WB = whole body, Br = brain, SG = salivary gland, MT = Malpighian tubules, FB = fat body, Ep = epidermis, and MG = midgut. Midgut samples were run separately because they required more PCR cycles for detection. 112

Figure 3.4. Pharmacological inhibition of calcium signaling blocks rapid cold- hardening in tissues of E. solidaginis. (A) Fluorescent cell viability images of midgut tissue incubated with the indicated drug at the following temperature conditions: Control: 4°C/3 h; Directly Frozen: direct transfer to -20°C for 2 h; RCH: slow ramp from 4 to - 20°C over 1 h, then held at -20°C for 2 h. Green nuclei representing live cells and red nuclei representing dead cells. (B) Quantified viability results for both midgut and salivary gland tissue. All values are mean ± SEM, n = 4, and different letters indicate significant differences among the three temperature treatments, while an “*” indicates a significant reduction in survival in the RCH group (ANOVA, Tukey, p<0.05). NCF = nominally calcium free medium, LaCl3 is a general calcium channel blocker, BAPTA is an intracellular calcium chelator, W-7 is an antagonist of the calcium binding protein calmodulin, KN-93 inhibits activation of CaMKII, while 2-APB is an inhibitor of IP3- mediated calcium signaling.

113

Figure 3.5. Additional cell viability assays. (A) Effect of calcium-specific pharmacological agents on survival of control tissues and tissues directly frozen at -20°C. (B) Mild freezing at -17.5°C caused significantly less mortality than freezing at -20°C. (C) Inhibition of calcium signaling increases mortality of tissues frozen at -17.5°C. (D) Effect of calcium-signaling antagonists on cell viability following control (25°C, 4 h) and cold shock (-14°C, 2 h ) conditions in tissue of S. bullata. Bars represent mean ± SE, N=4. Different letters represent significant differences between two treatments, while an “*” indicates a group in which addition of a pharmacological agent significantly reduced survival relative to the calcium-rich group for the same temperature treatment (ANOVA, Tukey, p<0.05).

114

Figure 3.6. Chilling also evokes an increase in intracellular calcium in tissues of the flesh fly, S. bullata. (A) Decreasing temperature from 25 to 0°C at 1°C/min elicited an increase in intracellular calcium in salivary gland tissue of S. bullata. (B) Maintaining tissues at 25°C caused no change in intracellular calcium. (C) Increasing the rate of temperature decrease from 1 to 2°C/min causes an increase in calcium, although the response is dampened relative to (A). (D) Raising the temperature back to 25°C from 0°C at 1°C/min caused calcium levels to rapidly decrease towards control levels. Values are mean ± SEM, N=3 for A-C N=2 for D.

115

Figure 3.7. Pharmacological inhibition of calcium signaling also blocks rapid cold- hardening in tissues of S. bullata. (A) Fluorescent cell viability images of midgut tissue incubated with the indicated drug at the following temperature conditions: Control: 25°C/4 h; Cold Shock: -14°C/2 h; RCH: 0°C/2 h, -14°C/2 h. Green nuclei representing live and red nuclei representing dead cells. (B) Quantified viability results for both midgut and fat tissue. All values are mean ± SEM, n = 4. Different letters indicate significant differences in survival among the three temperature treatments, while an “*” indicates a significant reduction in survival in the RCH group (ANOVA, Tukey, p<0.05). NCF = nominally calcium free medium, LaCl3 is a general calcium channel blocker, BAPTA is an intracellular calcium chelator, W-7 is an antagonist of the calcium binding protein calmodulin, KN-93 inhibits activation of CaMKII, while 2-APB is an inhibitor of IP3-mediated calcium signaling.

116

Figure 3.8. Developmental and tissue-specific expression of calmodulin in the flesh fly S. bullata. (A) Developmental expression profile in all post-embryonic life stages. (B- D) Tissue-specific expression in larvae, pupae, and adult males, respectively. In B-D, fat body samples were run separately because they required more PCR cycles for detection. FL = feeding larvae, WL = wandering larvae, P = pupae, PA = pharate adult, Br = brain, MG = midgut, MT = Malpighian tubules, FB = fat body, Ep = epidermis, SG = salivary gland.

117

Figure 3.9. Developmental and tissue-specific expression of calcium/calmodulin- dependent protein kinase II (CaMKII) in the flesh fly, S. bullata. (A) Developmental expression profile in all post-embryonic life stages. (B-D) Tissue-specific expression in larvae, pupae, and adult males, respectively. In C-D, fat body samples were run separately because they required more PCR cycles for detection. FL = feeding larvae, WL = wandering larvae, P = pupae, PA = pharate adult, Br = brain, MG = midgut, MT = Malpighian tubules, FB = fat body, Ep = epidermis, SG = salivary gland.

118

Figure 3.10. Working model for the role of calcium signaling during cold-sensing and rapid cold-hardening. Low temperature causes an increase in intracellular calcium concentration, and we hypothesize this occurs by activation of calcium leak channels (CLC) in the cell membrane and/or by low temperature inhibition of ATP-dependent calcium export mechanisms. In this model, low temperature inhibits the activity of both sarcoplasmic endoplasmic reticulum calcium ATPase (SERCA) and sodium/potassium ATPase (Na/K ATPase) coupled to the sodium calcium exchanger (NCX). Inside the cell, calcium, via calcium/calmodulin-dependent protein kinase II (CaMKII) and other unknown mechanisms, triggers pathways involved in rapid cold-hardening, thereby enhancing the cell’s cold tolerance.

119

Chapter 4: Survival and Energetic Costs of Repeated Cold Exposure in the Antarctic Midge, Belgica antarctica: A Comparison Between Frozen and Supercooled Larvae

Abstract

In this study, we examined the effects of repeated cold exposure (RCE) on the survival, energy content, and stress protein expression in larvae of the Antarctic midge, Belgica antarctica (Diptera: Chironomidae). Additionally, we compared results between larvae that were frozen at -5°C in the presence of water during RCE and those that were supercooled at -5°C in a dry environment. While over 95% of larvae survived a single 12 h bout of freezing at -5°C, after 5 cycles of RCE survival of frozen larvae dropped below

70%. Meanwhile, the survival of control and supercooled larvae was unchanged, remaining around 90% for the duration of the study. At the tissue level, frozen larvae had higher rates of cell mortality in the midgut than control and supercooled larvae.

Furthermore, larvae that were frozen during RCE experienced a dramatic reduction in energy reserves; after 5 cycles, compared to supercooled larvae, frozen larvae had 25% less lipid, 30% less glycogen, and nearly 40% less trehalose. Finally, larvae that were frozen during RCE had higher expression of hsp70 than those that were supercooled, indicating a higher degree of protein damage in the frozen group. Results were similar between larvae that had accumulated 60 h of freezing at -5°C over 5 cycles of RCE and those that were frozen continuously for 60 h, suggesting that the total time spent frozen determines the physiological response. Our results suggest that it is preferable, both from

120

a survival and energetic standpoint, for larvae to seek dry microhabitats where they can avoid inoculative freezing and remain unfrozen during RCE.

Introduction

The Antarctic midge, Belgica antarctica Jacobs, is the southernmost insect (Sugg et al., 1983) and is found exclusively on the west coast of the Antarctic Peninsula and its islands. The midge has a two-year life cycle, in which larvae feed primarily on the algae

Prasiola crispa, mosses such as Drepanocladus uncinatus, microorganisms, and detritus in nutrient-enriched substrate near seal colonies and bird nesting sites (Convey and

Block, 1996). Adults emerge during a brief period in early summer, mate, oviposit, and die within 10 days. B. antarctica is apterous, a common adaption in wind-swept polar areas. This species has a sporadic, albeit locally abundant, distribution on the Antarctic

Peninsula.

Not surprisingly, given its habitat, larvae of B. antarctica are extremely tolerant of a number of environmental stresses; larvae can survive freezing to -15°C, loss of 70% of body water by desiccation, anoxia, long periods of immersion in fresh water and salt water, and a broad range of pH conditions (Baust and Lee, 1987). While air temperatures in Antarctica reach winter lows of -40°C on the Antarctic Peninsula, the microhabitat of

B. antarctica is thermally buffered by snow and ice, so that the microhabitat temperature rarely dips below -5°C (Baust and Lee, 1981; Elnitsky et al., 2008). Larvae remain freeze-tolerant year-around and can increase their freeze-tolerance by undergoing rapid cold-hardening (Lee et al., 2006). Furthermore, under laboratory conditions B. antarctica

121

larvae are capable of cryoprotective dehydration, in which gradual chilling to -3°C in the presence of ice causes larvae to lose 40% of their body water and remain unfrozen at subzero temperatures (Elnitsky et al., 2008)

In general, all insects can be classified as either freeze-tolerant or freeze- susceptible (Lee, 2010). Whereas freeze-tolerant insects can survive internal ice formation, freeze-avoiding insects must remain supercooled at subzero temperatures.

However, the evolutionary forces that drive insects to adopt a particular strategy remain unclear. Freeze-tolerance tends to predominate in the Southern Hemisphere, where higher climatic variability requires insects to be able to survive internal ice-nucleation when a sudden cold-snap arrives (Sinclair et al., 2003; Sinclair and Chown, 2005a). Also, a species’ water balance characteristics tend to drive its cold tolerance strategy; high trans- cuticular water permeability is often associated with freeze-tolerance, since freezing is a means of conserving body water in the winter (Zachariassen et al., 2008). In the case of

B. antarctica, larvae are clearly freeze-tolerant (Baust and Edwards, 1979), but during gradual chilling some larvae can avoid inoculative freezing and undergo cryoprotective dehydration, particularly when the soil moisture content is low (Elnitsky et al., 2008).

These larvae remain unfrozen at subzero temperatures for 14 days and nearly all survive, suggesting that cryoprotective dehydration is a viable means of long-term cold survival.

Additionally, provided they avoid contact with environmental ice, larvae could remain supercooled during a sudden, acute exposure to cold, since the minimum microhabitat temperature (i.e. approx. -5°C) rarely drops below the supercooling point (around -7°C).

However, it has not been empirically addressed whether freezing or supercooling is the preferred means of acute low temperature survival. 122

Moreover, while previous studies on stress tolerance in B. antarctica examined a single stress exposure, in nature larvae are exposed to repeated bouts of environmental stress. Particularly, during late summer and early winter, microhabitat temperatures fluctuate numerous times across the larvae’s freezing point (Elnitsky et al., 2008). In recent years, several studies have addressed the effects of repeated cold exposure (RCE) in insects. Typically, for freeze-avoiding insects, RCEs allow for better survival in comparison to a single long-term exposure, because insects can repair damage, such as disruption of ion homeostasis, accrued during the cold exposure (Kostal et al., 2007).

However, in the few studies that have looked at RCE in freeze-tolerant insects, the opposite appears to be true. In the freeze-tolerant hoverfly Syrphus ribesii, most larvae can only survive a single freezing event, and switch to a freeze-avoiding strategy thereafter (Brown et al., 2004). Similarly, in the freeze-tolerant caterpillar Pringleophaga marioni, repeated freeze-thaw cycles are deleterious, perhaps due to accumulated damage to sensitive tissues such as the midgut epithelium (Sinclair and Chown, 2005b). In addition to being ecologically relevant, studies of RCE can also reveal novel adaptations for cold survival. For example, in the parasitic wasp Aphidius colimani, the proteome of wasps exposed to RCE was distinct from those exposed to a single bout of low temperature (Colinet et al., 2007).

In this study we examined the effects of RCE in B. antarctica larvae.

Furthermore, results were compared between larvae that were frozen and those that were supercooled, by exposing them to cold in the presence and absence of water, respectively.

In particular, we focused on both the survival costs of RCE as well as other physiological consequences, including tissue damage, energy reserve depletion, and stress protein 123

expression. Insects have a suite of physiological adaptations to cold (Clark and Worland,

2008), but these adaptations are energetically expensive. Particularly, recovery from cold exposure can be costly, so measuring key energy reserves in response to cold is an indirect measure of both the amount of damage accumulated and the fitness costs of that damage, since energy is diverted from growth and reproduction (Marshall and Sinclair,

2010). Additionally, we measured heat shock protein 70 (hsp70) expression as a biomarker of cellular stress during RCE. Heat shock proteins, as well as other stress proteins, are essential for repairing misfolded proteins during the recovery phase (Kostal and Tollarova-Borovanska, 2009). This is the first study to examine multiple stress exposure and the consequences of freezing versus supercooling in B. antarctica.

Materials and Methods

Experimental animals

In January 2010, larvae were collected on Cormorant Island, Humble Island, and

Norsel Point near Palmer Station on the Antarctic Peninsula (64°46 S, 64°04 W). For experiments conducted at Palmer Station, larvae were immediately hand-picked from the substrate in ice water and placed on wet filter paper at 4°C overnight prior to the experiments. Additionally, samples of substrate were shipped chilled (approx. 0°C for 7 days) to Ohio State University and stored at 4°C for additional experiments.

Experimental conditions

For the RCE treatments, larvae were exposed to 1 to 5 diurnal cycles of 12 h at -

5°C followed by 12 h at 4°C, 100% RH. Additionally, to compare the effects of RCE to a

124

single prolonged exposure to cold, separate groups of larvae were exposed to -5°C for 60 h, since after 5 cycles of RCE larvae have accumulated 60 h at -5°C. These treatment groups were also given 12 h of recovery at 4°C, 100% RH. Meanwhile, control animals were held at 4°C, 100% RH for the duration of the study. Larvae in the “frozen” groups were submerged in approximately 50 μl of ddH20 prior to the -5°C exposure to promote inoculative freezing (Lee et al., 2006). After exposure to -5°C, tubes were thawed, and excess water was removed to prevent overhydration during recovery (Lopez-Martinez et al., 2009). Larvae in the “supercooled” groups were gently blotted dry on an absorbent tissue prior to the experiment and placed in dry centrifuge tubes. To assure that larvae could remain supercooled for 12 h at -5°C, a subset of larvae were placed in direct contact with a thermocouple at -5°C and monitored for 12 h for the presence of a freezing exotherm. Out of 12 larvae tested, only 2 froze, indicating that nearly all larvae remain supercooled at -5°C for 12 h (data not shown).

Whole animal survival

At the end of the 12 h recovery phase after each of 5 RCE cycles, larvae were assessed for survival. Concurrently, survival was measured each day for control larvae that were held continuously at 4°C. Groups of 15 larvae were placed in a drop of water under a microscope, and larvae that either moved spontaneously or in response to gentle prodding were considered to be alive. The mean survival for each treatment group is based on 5 replicates of 15 larvae each.

Cell survival of midgut tissue

To test for sublethal damage to tissues during RCE, cell viability of midgut tissue was assessed. The midgut tissue of B. antarctica is particularly sensitive to cold (Lee et 125

al., 2006; Teets et al., 2008), making it a suitable tissue for this experiment. After each of

5 cycles, midguts from control, frozen, and supercooled larvae were dissected in Coast’s solution (Coast and Krasnoff, 1988) and cell viability was assessed using the

LIVE/DEAD sperm viability kit (Invitrogen, Carlsbad, CA), as adapted by Yi and Lee

(2003). Only living, non-moribund larvae were dissected for cell viability determination.

Tissues were stained and images were taken via fluorescent microscopy at the Campus

Microscopy and Imagining Facility, Ohio State University. Living cells with intact membranes fluoresced green, while dead cells with damaged membranes fluoresced red.

Cell survival for each replicate was based on the mean count of 3 groups of 100 cells. For each treatment, 4 replicates were conducted.

Metabolite assays

To determine the effects of RCE on energy stores, colorimetric assays for lipid, glycogen, trehalose, and glucose were conducted. Also, glycerol, a common cryoprotectant and carbon source in cold-acclimated insects, was measured. After each of 5 RCE cycles, 5 replicates of 50 individuals from each treatment (i.e. control, frozen, and supercooled) were quickly frozen at -80°C and shipped on dry ice to Ohio State

University for biochemical analysis. Samples were thawed one at a time, and 5 individuals were set aside for lipid analysis. The remaining 45 individuals were blotted dry, weighed, and homogenized in 7% perchloric acid. The homogenates were centrifuged at 14,000g, and the supernatant was neutralized with 0.789 M KOH.

Neutralized extracts were stored at -70°C until analysis.

Lipid content was measured according to Sim and Denlinger (2009), using vanillin reagent. Briefly, groups of 5 larvae were homogenized in 1:1 126

chloroform:methanol, the solvent was evaporated, and vanillin reagent was added to the sample. The absorbance of the solution was read at 490 nm in a NanoDrop 2000C spectrophotometer and compared to known standards. Glycogen content was measured according to Kepler and Decker (1984). Amyloglucosidase from Aspergillus niger

(Sigma, A1602, St. Louis, MO) was used to liberate free glucose in the homogenate, and the glucose concentration was measured using a glucose assay kit (Sigma, GAGO20) by measuring the absorbance at 540 nm. The glycogen content was calculated by subtracting the free glucose content from the enzyme-digested glucose content, and comparing this value to standards of known glycogen concentration. Trehalose content was also determined enzymatically, according to Chen et al. (2002). Trehalose was hydrolyzed using trehalase from porcine kidney (Sigma, T8778), and the liberated glucose concentration was compared to trehalose standards, correcting for free glucose concentration. Finally, glycerol content was measured using Free Glycerol Reagent

(Sigma, F6428). The suggested ratio of homogenate to reagent was increased about 10X to get a signal, but otherwise the assay was conducted according to the manufacturer’s protocol. For each metabolite measured, the content is expressed as μg metabolite/mg dry mass.

Energy content calculations

After conducting assays for the major energy reserves, the total energy content

(TEC) of the major energy reserves (lipid, glycogen, trehalose, and glucose) was calculated according to Djawdan et al. (1998). The formula for the TEC is

TEC=0.393(lipid content) + 0.176(total carbohydrate content), where TEC is in J/mg dry mass, lipid content is in μg lipid/mg dry mass, and total carbohydrate content, in μg 127

carbohydrate/mg dry mass, is the sum of glycogen content, trehalose content, and glucose content. Additionally, since the TEC was dominated by lipids, the most abundant energy reserve, we also separately calculated the carbohydrate energy content (CEC), in J/mg dry mass, by CEC=0.176(total carbohydrate content).

Northern blot hybridization

To further assess the level of cellular stress caused by RCE, we measured hsp70 expression during RCE. Clones for hsp70 and 28s rRNA were obtained from previously reported sequences (Rinehart et al., 2006). For hsp70, the following primers were used: forward primer 5’-GATGCAGTCATCACAGTTCCAGC-3’, reverse primer

5’-AACAGAGATCCCTCGTCGATGGT-3’. For 28s, the following primers were used: forward primer 5’-ACTTGATTGATGTTGGCCTGGTGG-3’, reverse primer

5’-GCTAATTGCTTCGGCAGGTGAGTT-3’.

After each RCE cycle, groups of 25 larvae were immediately frozen at -80°C.

Larvae were homogenized in Trizol reagent, and RNA was extracted according to the manufacturer’s protocol. For hsp70 Northerns, 5 μg of RNA was separated on a 1.4% agarose, 0.41 M formaldehyde gel. Because of the high signal strength, only 1 μg of

RNA was used for the 28s control Northerns. After electrophoresis, RNA was transferred to a positively charged nylon membrane (Hybond-N+, Amersham Biosciences,

Piscataway, NJ) using the Turboblotter rapid downward transfer system (Schleicher and

Schuell, Inc., Keene, New Hampshire). DNA clones for hsp70 and 28s, obtained by PCR, were labeled with the DIG-High Prime labeling kit (Roche, Switzerland). Hybridization was performed using the DNA Labeling and Starter Kit II (Roche), and the membranes

128

were exposed on Blue Lite Autorad Film (ISC BioExpress, Kaysville, UT) for 5 min to

24 h, depending on signal strength. All Northerns were performed in technical triplicate.

Statistical analysis

Means were compared using ANOVA with a Bonferroni correction for multiple comparisons, using Minitab v. 13. The primary comparisons used the means at each time point for the control, frozen, and supercooled treatments. Survival data were arcsin- square root transformed prior to analysis, and statistical significance was set at P<0.05.

All data are reported as mean ± s.e.m.

Results

Whole animal survival

Nearly all control larvae held at 4°C survived the duration of the experiment;

97.3±1.6% survived 1 day at 4°C, while at the end of 5 days 90.8±1.7% were still alive

(Fig. 1A). Larvae in the RCE (1-5 cycles of 12 h at -5°C followed by 12 h at 4°C) group that were supercooled at -5°C also had high survival, as 96.0±2.2% survived a single cycle, while 88.0±2.5% survived 5 cycles of RCE state (Fig. 1A). At no point did the survival of supercooled individuals differ significantly from the control, and the lowest survival observed was 84±3.4% after 4 cycles. However, survival of larvae frozen at -5°C steadily decreased over the course of 5 cycles (Fig. 1A), going from 96.0±3.4% after a single cycle down to 68.0±3.9% after 5 cycles. This survival level was significantly lower than that of both control and supercooled larvae after 5 cycles (p<0.05). For larvae exposed to -5°C continuously for 60 h, the survival of supercooled larvae was

129

significantly higher than that of frozen larvae, 93.3±3.3% vs. 54.7±3.7%, respectively

(Fig. 1A, p<0.05).

Cell survival of midgut tissue

As with whole animal survival, RCE in the frozen state had a deleterious effect on cell survival of midgut tissue. After each of the 5 cycles, survival of midgut tissue from frozen larvae was significantly lower than that of either control or supercooled larvae

(Fig. 1B, p<0.05). After 1 cycle, survival of midgut tissue was 80.4±1.7% for frozen larvae, compared to 93.2±1.1% for control larvae and 91.8±2.4% for supercooled larvae.

From cycles 1-4, midgut cell survival of frozen larvae was approximately 80%, while that of control and supercooled larvae was above 90% (Fig. 1B). After the fifth cycle, cell survival of supercooled larvae (85.3±1.6%) dipped significantly below that of control larvae (94.3±0.3%), but was still significantly higher than that of frozen larvae, which decreased to 69.2±2.5% (Fig. 1B). For larvae continuously exposed to -5°C for 60 h, cell survival after 12 h of recovery was 80.7±3.6% for frozen larvae and 88.0±1.0% for supercooled larvae. However, this difference was not statistically significant (p>0.05).

Lipid content

For the first 4 cycles of RCE, lipid content did not change significantly, varying between 172.4.0±6.5 μg lipid/mg dry mass and 209.9±8.9μg lipid/mg dry mass for the treatment groups (Fig. 2A). After the fifth cycle, the lipid content of frozen individuals was 25% less than that of supercooled individuals, a difference that was statistically significant (p<0.05). However, the difference in lipid content between control and frozen individuals after 5 cycles was not significant. As with the larvae exposed to 5 cycles of

RCE, the lipid content of larvae supercooled for 60 h continuously (189.4±2.6 μg

130

lipid/mg dry mass) was significantly higher (p<0.05) than that of larvae frozen for 60 h

(141.2±5.4 μg lipid/mg dry mass).

Glycogen content

Similar to lipid content, glycogen content of larvae that were frozen during RCE was significantly lower than in controls and their supercooled counterparts (Fig. 2B).

After one cycle of 12 h at -5°C and 12 h at 4°C, the glycogen content of frozen individuals was 45.9±2.6 μg glycogen/mg dry mass, but this value steadily decreased every cycle, and by the end of 5 cycles glycogen content was only 21.7±1.6 μg glycogen/mg dry mass (Fig. 2B). By the third cycle, glycogen content was significantly lower than that of control larvae, and by the fifth it was lower than that of both control and supercooled larvae (p<0.05). After 5 cycles, glycogen content of frozen larvae was approximately 43% less than that of both control and supercooled individuals. Similarly, larvae frozen at -5°C for 60 h had significantly less glycogen than those supercooled at -

5°C for 60 h, with those being frozen having 26% less glycogen (Fig. 2B, p<0.05).

Trehalose content

The trehalose content of larvae exposed to RCE in the frozen state steadily decreased over the course of the experiment, starting at 52.9±2.3 μg trehalose/mg dry mass after 1 cycle and ending at 37.0±3.0 μg trehalose/mg dry mass (Fig. 2C). On the other hand, trehalose content of supercooled larvae increased modestly over the course of the experiment, starting at 52.0±3.5 μg trehalose/mg dry mass after 1 cycle, peaking at

62.4±2.8 μg trehalose/mg dry mass, and ending at 58.5±3.5 μg trehalose/mg dry mass

(Fig. 2C). After 3 cycles of RCE, trehalose content of frozen individuals was significantly lower than that of both control and supercooled larvae, while after 4 and 5 cycles it was

131

lower than that of supercooled larvae only (p<0.05). For larvae exposed to -5°C for 60 h, the trehalose content of frozen individuals was 43.1±1.4 μg trehalose/mg dry mass, which did not differ from that of supercooled individuals (53.1±3.5 μg).

Glucose content

The glucose content of control larvae did not change appreciably over the course of the experiment (Fig. 2D). However, larvae in both RCE groups experienced a dramatic increase in glucose content. After a single cycle of -5°C for 12 h and 4°C for 12 h, glucose content significantly increased nearly 3 fold in both frozen and supercooled larvae (p<0.05). For the duration of the experiment, glucose content of RCE individuals was between 2-4 times that of their control counterparts, maxing out at 2.10±0.09 μg glucose/mg dry mass for frozen larvae and 2.24±0.28 μg glucose/mg dry mass for supercooled larvae (Fig. 2D). Interestingly, for larvae continuously exposed to -5°C for

60 h, glucose content only increased in frozen individuals; the level for frozen larvae

(1.47±0.14 μg glucose/mg dry mass) was significantly higher than that of supercooled larvae (0.63±0.16 μg glucose/mg dry mass).

Glycerol content

For most of the experiment, glycerol content did not significantly differ between control, frozen, and supercooled larvae (Fig. 2E). There was a single peak of glycerol content after 3 cycles of RCE in supercooled larvae, in which glycerol content was more than double that of control larvae and more than triple that of frozen larvae (p<0.05).

However, after cycle 3, the glycerol content of supercooled larvae returned to that of the other groups for the remainder of the experiment. Likewise, there was no difference in

132

glycerol content between larvae frozen continuously for 60 h and larvae supercooled

(Fig. 2E).

Total energy content and carbohydrate energy content

To summarize the results of the above experiments, we calculated the TEC of larvae due to the major energy reserves, lipid, glycogen, trehalose, and glucose. The TEC was largely unchanged for the first three cycles of RCE, ranging between approximately

8.2 and 9.3 J/mg dry mass (Fig. 3A) for all treatments. However, after 4 cycles, the energy content of both frozen and control individuals were almost 20% lower than that of supercooled individuals, a difference that was significant (p<0.05). The energy content of frozen larvae further decreased during the fifth cycle, so that after 5 cycles the energy content of frozen individuals (7.08±0.37 J/mg dry mass) was significantly lower than that of both control larvae (8.36±0.23 J/mg dry mass) and supercooled larvae (9.69±0.26 J/mg dry mass). For larvae continuously exposed to -5°C for 60 h, the energy content of frozen individuals (7.13±0.0.30 J/mg dry mass) was significantly less than that of supercooled individuals (9.06±0.16 J/mg dry mass).

Additionally, since the total energy content was dominated by lipid energy reserves, and because changes in carbohydrate content were much more dramatic, we also calculated the CEC in response to RCE. The CEC of larvae frozen at -5°C during

RCE steadily declined over the course of the experiment, decreasing from 1.77±0.07

J/mg dry mass after 1 cycle to 1.07±0.07 J/mg dry mass after 5 cycles (Fig. 3B).

Meanwhile, the CEC of control individuals decreased slightly, while that of supercooled individuals increased slightly, although these changes were not statistically significant

(p>0.05). After 5 cycles of RCE, the CEC of frozen larvae was nearly 30% less than that

133

of control larvae and nearly 40% less than that of supercooled larvae (Fig. 3B). The CEC of larvae frozen for 60 h (1.37±0.05 J/mg dry mass) was also significantly lower than that of larvae supercooled for 60 h (1.77±0.07 J/mg dry mass; Fig. 3B).

Heat shock protein expression

Using Northern blot hybridization, we measured expression of hsp70, a common biomarker for environmental stress, in response to RCE. Over the 5 day experiment, expression of hsp70 mRNA increased slightly in control individuals and supercooled individuals (Fig. 4). However, this increase was not as dramatic as that observed in frozen larvae, where there was clear upregulation of hsp70 over the course of the experiment. We measured expression after 1,3, and 5 cycles of RCE, and after each of these cycles hsp70 transcript level was clearly higher in frozen larvae than in supercooled larvae. Likewise, for larvae exposed to -5°C continuously, expression of hsp70 was higher in the frozen group (Fig. 4). Expression of the control gene, 28s, did not change appreciably over the course of the experiment.

Discussion

While previous studies on environmental stress tolerance in B. antarctica consisted of a single stress exposure, in nature B. antarctica is exposed to repeated bouts of stress punctuated by periods of recovery. In this study, we used a thermal regime of 12 h at -5°C followed by 12 h at 4°C to examine the physiological consequences of RCE in

B. antarctica. Also, by holding them in the presence and absence of water at -5°C, we compared the effects of RCE in both the frozen and supercooled states. Overall, we observed deleterious effects of RCE in frozen, but not supercooled larvae. Larvae

134

exposed to multiple bouts of cold in the frozen state had higher mortality, more tissue damage, lower energy reserves, and higher expression of heat shock proteins than supercooled larvae.

RCE in frozen larvae negatively impacts survival

At our test temperature, -5°C, over 95% of larvae survived a single 12 h exposure, whether in the frozen or supercooled state (Fig. 1A). However, as the number of RCE cycles increased, survival of frozen larvae steadily decreased below that of control and supercooled larvae; after 3 cycles survival decreased to 77%, and by the conclusion of 5 cycles survival was only 68%. Meanwhile, survival of supercooled larvae remained around 90% for the duration of the study. To our knowledge, only one other study has directly addressed the question of freezing versus supercooling in the same species.

Layne and Kuharsky (2001) examined the effects of freezing and supercooling at -5°C in the goldenrod gall fly, Eurosta solidaginis, and found the opposite of our results; namely that larvae that were supercooled for 10 wks at -5°C had much lower adult emergence rates than those frozen at -5°C for 10 wks. Layne and Kuharsky (2001) postulated that the difference in survival could be attributed to higher water loss in the supercooled individuals, which was confirmed by Irwin and Lee (2002). In this study, there was no significant change in water content (all larvae had around 73% water content; data not shown), so differences in survival must be due to some other parameter. While it is likely that larvae would dehydrate when supercooled in nature, we maintained 100% RH in this study so that any observed changes could be attributed to freezing or supercooling, and not to changes in water content.

135

For chill-susceptible insects, intermittent periods of recovery during a cold exposure often permit higher survival than a single prolonged exposure to cold (Chen and

Denlinger, 1992; Colinet et al., 2006; Kostal et al., 2007; Leopold et al., 1998; Nedved et al., 1998; Renault et al., 2004). Periods of higher temperatures allow the insects to repair cold injury by restoring ion homeostasis (Kostal et al., 2007) and replenishing ATP levels

(Dollo et al., 2010), for example. However, this benefit of RCE only appears to hold true for freeze-avoiding species. While the literature on this topic is scarce, in both our study and in previous studies of the sub-Antarctic beetle Hydromedion sparsutum and the hoverfly Syrphus ribesii, RCE in the frozen state caused significant mortality (Bale et al.,

2001; Brown et al., 2004). In our study, it appears that mortality due to freezing is simply a function of the total time spent in the frozen state; survival after 5 cycles of RCE

(68.0%), in which 60 h of freezing are accumulated, was very similar to and not significantly different from the survival after 60 h of continuous freezing (54.7%). In the supercooled state, survival of larvae after 5 cycles of RCE (88.0%) was also very similar to survival of larvae exposed continuously to -5°C for 60 h (93.3%). What is not clear is whether a longer exposure to -5°C in the supercooled state would begin to produce mortality, and whether this mortality could be ameliorated by periodic recovery times, as is the case for previous studies of RCE.

Freezing, but not cold per se, causes significant tissue damage

As with whole animal survival, RCE in the frozen state had a negative impact on cell survival of midgut tissue. Cell survival of midgut tissue was consistently 10-20% higher in supercooled larvae compared to their frozen counterparts at each time point

(Fig. 1B). Similar results were obtained by Sinclair and Chown (2005b), who showed that

136

repeated freeze exposures in P. marioni cause considerable damage to gut tissue. In our study, while cell survival of frozen larvae was consistently lower in frozen larvae, the amount of mortality remained relatively constant until the fifth cycle. One possible explanation for this result is that insect midgut tissue has high cell turnover and rapid regeneration of dead cells (Okuda et al., 2007); thus the cumulative effects of multiple freeze cycles are not as dramatic since previously dead cells have already been removed.

Also, to reduce variability, we only selected live, non-moribund larvae for dissection, so we likely missed the true range in cell survival.

RCE in the frozen state significantly depletes energy reserves

Adaptations to survive low temperatures, such as cyroprotectant synthesis and stress protein production, are energetically costly. As a result, we predicted that RCE would result in significant depletion of key energy reserves. Our results indicate that while this is true when B. antarctica is frozen during RCE, this is not the case when supercooled. By the end of 5 cycles of RCE, frozen larvae had lower levels of lipid, glycogen, and trehalose compared to their supercooled counterparts (Fig. 2A-C). By pooling all the data, we observed that by the 5th cycle of RCE frozen larvae had 27% less

TEC and 39% less CEC than supercooled larvae (Fig. 3). Also, as with survival, the energetic effects of RCE in the frozen state seem to be a function of the cumulative time frozen at -5°C; the metabolite and total energy profiles of larvae frozen continuously for

60 h are remarkably similar to those exposed to 5 cycles of RCE (Figs 2, 3).

Our results are in agreement with previous studies looking at energetic consequences of RCE. In the freeze-tolerant caterpillar P. marioni, while RCE does not result in changes of body composition, damage to the gut hinders feeding, resulting in

137

decreased body size (Sinclair and Chown, 2005b). In Drosophila melanogaster, RCE causes a decrease in glycogen and lipid stores, which reduces fecundity (Marshall and

Sinclair, 2010). While we were unable to rear larvae to adulthood and look at the fitness consequences of RCE, it is likely that the severe energy depletion caused by freezing ultimately would have fitness consequences, particularly given the extremely short growing season in Antarctica. Furthermore, Renault et al. (2003) demonstrated that in the beetle Alphitobius diaperinus, body size positively correlates with the duration of cold survival, suggesting that higher energy reserves lead to prolonged cold survival. Thus, applying this to our results, it is possible that the reduced survival of frozen larvae was a direct consequence of the observed energy depletion.

While there was clearly a decrease in energy reserves in the frozen larvae, the exact fate of these energy reserves is unclear. One possibility is that lipid, glycogen, and/or trehalose serve as a carbon source for cryoprotectant production. In many cases, glycogen is the primary source of carbon for cryoprotectant production (Storey and

Storey, 1991), and often glycogen content positively correlates with an organism’s cold tolerance (Costanzo and Lee, 1993; Kostal et al., 2004; Overgaard et al., 2009). With this in mind, we measured the content of two known cryoprotectants in B. antarctica, glucose and glycerol (Baust and Edwards, 1979; Lee and Baust, 1981). While glucose levels increased nearly 5-fold (or roughly 1.26 μg/larva) in larvae exposed to RCE, this alone could only account for less than 10% of the reduction in glycogen in frozen larvae, which was roughly 19.3 μg per larva. At the same time, glycerol did not change during RCE, despite its known role as a protective compound during dehydration stress (Benoit et al.,

2007). Also, to our surprise, trehalose content decreased in larvae frozen during RCE,

138

despite its well-defined role as a cryoprotectant in B. antarctica (Elnitsky et al., 2008;

Lee and Baust, 1981). However, B. antarctica accumulates several other cryoprotectants, including erythritol, sucrose, and fructose, so it is possible that an increase in at least one of these compounds is at least partly responsible for the observed decrease in energy reserves during freezing.

A second, more likely explanation for the decreased energy reserves in frozen larvae is an elevated metabolic rate during recovery. Since freezing during RCE is clearly more damaging to the larvae (Figs 1, 4), the elevated repair cost could result in higher metabolism during recovery. Physiological responses to cold damage, including restoration of ion homeostasis (Kostal et al., 2007) and protein sysnthesis (Colinet et al.,

2007), are energetically expensive, suggesting that frozen larvae may need to elevate their metabolic rate during recovery. In support of this, Block, Worland and Bale (1998) found that the sub-Antarctic beetles H. sparsutum and Perimylops antarcticus both increase their metabolic rate during recovery from freezing, but not in response to chilling in a supercooled state. Similarly, the freeze-tolerant wood-frog Rana sylvatica elevates its metabolic rate several hours into recovery from freezing (Layne, 2000).

Future work on the metabolic rate of B. antarctica in response to freezing could elucidate whether an elevated post-freeze metabolic rate is responsible for the observed energy depletion during RCE.

RCE elevates heat shock protein expression in frozen larvae

As a final measure of the relative costs of RCE in the frozen versus supercooled state, we measured expression of hsp70. hsp70 is perhaps the best studied of the heat shock proteins, a group of molecular chaperones that are elevated in response to

139

numerous environmental stresses (Feder and Hofmann, 1999). Whereas most insects only express heat shock proteins transiently during stress, B. antarctica is unique in that heat shock proteins are expressed continuously (Rinehart et al., 2006). Despite the constitutive expression, heat shock protein transcript levels can still be elevated by a number of stresses, including UV exposure (Lopez-Martinez et al., 2008) and dehydration (Lopez-

Martinez et al., 2009). In this study, we observed that RCE in the frozen, but not supercooled, state enhances expression of hsp70 (Fig. 4). The same is true of a continuous freezing exposure; larvae frozen at -5°C for 60 h had noticeably higher expression of hsp70 compared to those supercooled at -5°C. While Rinehart et al. (2006) and Lopez-Martinez et al. (2008) did not observe an increase in hsp70 mRNA in response to freezing, their samples were only taken after a brief (i.e. 0-2h) recovery period.

Once again, our results for hsp70 point towards a higher cost of RCE in the frozen state. Since hsp70 is primarily responsive to protein damage, our results indicate that freezing during RCE likely leads to protein denaturation. Thus, hsp70 expression during

RCE is dependent on freezing, and not simply the drop in temperature. To our knowledge, this is the first study showing differential expression of a heat shock protein depending on whether the animal is frozen or supercooled.

Ecological relevance of this study

One caveat of this study is that our temperature fluctuations were likely of greater magnitude and periodicity than typical microhabitat conditions. Nonetheless, in the late summer and early winter, larvae are exposed to periodic freeze-thaw cycles in their natural habitat (see microclimate data from Elnitsky et al., 2008). Also, depending on the buffering capacity of its particular microhabitat, larvae could conceivably experience

140

greater fluctuations in temperature that more closely resemble the rapidly changing ambient temperatures. A second caveat is whether larvae can use supercooling for subzero exposures longer than 12 h. Out of 12 larvae tested, only 2 froze during a 12 h exposure to -5°C, and the 2 that froze did so during the initial rapid cooling phase. This suggests that the supercooled state is stable once the animal reaches -5°C. Finally, our results lead to the question if larvae can regulate whether they freeze or remain supercooled. The variation in moisture conditions within their environment (Elnitsky et al., 2008),suggests that larvae may avoid freezing by moving to a dry area where the risk of inoculative freezing is reduced. Thus, we feel that our experiments reasonably reflect natural conditions and can be used to make inferences about the physiological responses of larvae to acute fluctuating temperature exposures in their natural environment.

While this study and our previous ones have focused on acute exposures to cold, overwintering conditions, which feature several months of continuous subzero temperature, offer different challenges for larvae. Here, larvae can either remain frozen for the duration of the winter or avoid inoculative freezing by supercooling. While supercooling is a possibility for overwintering larvae, more likely overwintering larvae that avoid freezing would undergo cryoprotective dehydration due to the influence of surrounding ice and snow. Based on the results of the present study, we hypothesize that a freeze-avoiding overwintering strategy such as cryoprotective dehydration would be favorable compared to long-term freezing. Future studies by our group will address the overwintering physiological ecology of B. antarctica.

Finally, our results may have implications for global climate change. While increases in air temperature would seemingly benefit polar insects by extending the

141

growing season, this is likely not the case (Bale and Hayward, 2010). As the Earth warms, climate models predict more extreme variation in temperature (Easterling et al.,

2000). Also, while snowfall in the Antarctic Peninsula has increased over the past 150 years (Thomas et al., 2008), there are localized decreases in snow cover on the peninsula

(Fox and Cooper, 1998) due to increased melting. Consequently, the microhabitat of B. antarctica, which is normally thermally buffered by snow and ice cover, could potentially experience more extreme changes in temperature and a greater number of freeze-thaw cycles. Additionally, the increased precipitation and snow melt could enhance soil moisture, thereby increasing the likelihood of inoculative freezing. For example, a 40% increase in soil moisture doubles the likelihood of inoculative freezing during gradual chilling in B. antarctica (Elnitsky et al., 2008). Given the results of the present study, both of these potential consequences of climate warming (increased number of freeze- thaw cycles and increased risk of inoculative freezing) would be detrimental to survival of B. antarctica.

Conclusions

We found that RCE in the frozen state, but not supercooled state, was deleterious for larvae of B. antarctica. Larvae exposed to multiple bouts of cold in the frozen state had lower survival, greater tissue damage, lower energy reserves, and increased protein damage compared to supercooled individuals. Also, the accumulated damage from RCE appears to be a function of the total time spent in the frozen state; larvae frozen continuously for 60 h experienced similar levels of damage and energy depletion as those that accumulated 60 h of freezing over 5 cycles of RCE.

142

Thus, it appears that while B. antarctica is freeze-tolerant, it is preferable to remain supercooled during brief, repeated cold exposure. This begs the question as to why B. antarctica isn’t strictly freeze-avoiding. However, given the rapid changes in moisture content in its environment and the frequent changes in temperature, larvae are at continual risk of inoculative freezing. For this reason, freeze-tolerance is common among insects in the Southern Hemisphere, where oceanic influences can cause summer cold- snaps that put insects at higher risk of inoculative freezing (Sinclair et al., 2003; Sinclair and Chown, 2005a). Therefore, we assert that B. antarctica is obligately freeze-tolerant, but likely prefers dry microclimates so that it can remain unfrozen at subzero temperatures. Over the short term, this can be accomplished by supercooling (the present study), whereas cryoprotective dehydration provides a means to avoid freezing during long-term chilling (Elnitsky et al., 2008). In subsequent studies, we will track B. antarctica larvae in the field to determine which cold tolerance strategy (i.e. freeze- tolerance or freeze-avoidance) is used by B. antarctica larvae in their natural environment.

Acknowledgements

We are grateful for the hard work and assistance of the support staff at Palmer

Station. We also thank Brian Kemmenoe, Richard Montione, and the rest of the OSU

Campus Microscopy and Imaging Facility for their help.

Grants

This research was supported by NSF grants ANT-0837559 and 0837613.

143

References

Bale, J.S., Hayward, S.A.L., 2010. Insect overwintering in a changing climate. Journal of Experimental Biology, 213, 980-994.

Bale, J.S., Worland, M.R., Block, W., 2001. Effects of summer frost exposures on the cold tolerance strategy of a sub-Antarctic beetle. Journal of Insect Physiology, 47, 1161- 1167.

Baust, J.G., Edwards, J.S., 1979. Mechanisms of freezing tolerance in an antarctic midge, Belgica antarctica. Physiological Entomology, 4, 1-5.

Baust, J.G., Lee, R.E., 1981. Environmental "homeothermy" in an Antarctic insect. Antarctic Journal of the United States, 15, 170-172.

Baust, J.G., Lee, R.E., 1987. Multiple stress tolerance in an antarctic terrestrial arthropod: Belgica antarctica. Cryobiology, 24, 140-147.

Benoit, J.B., Lopez-Martinez, G., Michaud, M.R., Elnitsky, M.A., Lee, R.E., Denlinger, D.L., 2007. Mechanisms to reduce dehydration stress in larvae of the Antarctic midge, Belgica antarctica. Journal of Insect Physiology, 53, 656-667.

Block, W., Worland, M.R., Bale, J., 1998. Respiratory responses to chilling and freezing in two sub-Antarctic insects. Cryobiology, 37, 163-166.

Brown, C.L., Bale, J.S., Walters, K.F.A., 2004. Freezing induces a loss of freeze tolerance in an overwintering insect. Proceedings of the Royal Society of London. Series B: Biological Sciences, 271, 1507-1511.

Chen, C.P., Denlinger, D.L., 1992. Reduction of cold injury in flies using an intermittent pulse of high temperature. Cryobiology, 29, 138-143.

Chen, Q.F., Ma, E., Behar, K.L., Xu, T., Haddad, G.G., 2002. Role of trehalose phosphate synthase in anoxia tolerance and development in Drosophila melanogaster. Journal of Biological Chemistry, 277, 3274-3279.

Clark, M.S., Worland, M.R., 2008. How insects survive the cold: molecular mechanisms- a review. Journal of Comparative Physiology. B, Biochemical, Systemic, and Environmental Physiology, 178, 917-933.

144

Colinet, H., Renault, D., Hance, T., Vernon, P., 2006. The impact of fluctuating thermal regimes on the survival of a cold-exposed parasitic wasp, Aphidius colemani. Physiological Entomology, 31, 234-240.

Convey, P., Block, W., 1996. Antarctic Diptera: Ecology, physiology and distribution. European Journal of Entomology, 93, 1-13.

Costanzo, J.P., Lee, R.E., 1993. Cryoprotectant production capacity in the freeze-tolerant wood frog, Rana sylvatica. Canadian Journal of Zoology 71, 71-75.

Djawdan, M., Chippindale, A.K., Rose, M.R., Bradley, T.J., 1998. Metabolic reserves and evolved stress resistance in Drosophila melanogaster. Physiological Zoology, 71, 584-594.

Dollo, V.H., Yi, S.-X., Lee, R.E., 2010. High temperature pulses decrease indirect chilling injury and elevate ATP levels in the flesh fly, Sarcophaga crassipalpis. Cryobiology, 60, 351-353.

Easterling, D.R., Meehl, G.A., Parmesan, C., Changnon, S.A., Karl, T.R., Mearns, L.O., 2000. Climate extremes: Observations, modeling, and impacts. Science, 289, 2068-2074.

Elnitsky, M.A., Hayward, S.A.L., Rinehart, J.P., Denlinger, D.L., Lee, R.E., 2008. Cryoprotective dehydration and the resistance to inoculative freezing in the Antarctic midge, Belgica antarctica. Journal of Experimental Biology, 211, 524-530.

Feder, M.E., Hofmann, G.E., 1999. Heat-shock proteins, molecular chaperones, and the stress response: Evolutionary and ecological physiology. Annual Review of Physiology, 61, 243-282.

Fox, A.J., Cooper, A.P.R., 1998. Climate-change indicators from archival aerial photography of the Antarctic Peninsula. Annals of Glaciology, 27, 636-642.

Irwin, J.T., Lee, R.E., 2002. Energy and water conservation in frozen vs. supercooled larvae of the goldenrod gall fly, Eurosta solidaginis (Fitch) (Diptera : Tephritidae). Journal of Experimental Zoology, 292, 345-350.

Kepler, D., Decker, K., 1984. Glycogen, in: Bergmeyer, H.U. (Ed.), Methods of Enzymatic Analysis. Verlag Chemie, Weinheim, Germany, pp. 11-18.

145

Kostal, V., Tollarova-Borovanska, M., 2009. The 70 kDa heat shock protein assists during the repair of chilling injury in the insect, Pyrrhocoris apterus. Plos One, 4, e4546.

Layne, J.R., 2000. Postfreeze O-2 consumption in the wood frog (Rana sylvatica). Copeia, 879-882.

Layne, J.R., Kuharsky, D.K., 2001. Effects of prolonged freezing and supercooling on body composition, pupariation, and adult emergence of Eurosta solidaginis (Diptera : Tephritidae). Environmental Entomology, 30, 12-16.

Lee, R.E., 2010. A primer on insect cold tolerance, in: Denlinger, D.L., Lee, R.E. (Eds.), Low Temperature Biology of Insects. Cambridge University Press, Cambridge, pp. 3-34.

Lee, R.E., Baust, J.G., 1981. Seasonal patterns of cold-hardiness in antarctic terrestrial arthropods. Comparative Biochemistry and Physiology. Part A, Physiology, 70, 579-582.

Lee, R.E., Elnitsky, M.A., Rinehart, J.P., Hayward, S.A.L., Sandro, L.H., Denlinger, D.L., 2006. Rapid cold-hardening increases the freezing tolerance of the Antarctic midge Belgica antarctica. Journal of Experimental Biology, 209, 399-406.

Leopold, R.A., Rojas, R.R., Atkinson, P.W., 1998. Post pupariation cold storage of three species of flies: increasing chilling tolerance by acclimation and recurrent recovery periods. Cryobiology, 36, 213-224.

146

Nedved, O., Lavy, D., Verhoef, H.A., 1998. Modelling the time-temperature relationship in cold injury and effect of high-temperature interruptions on survival in a chill-sensitive collembolan. Functional Ecology, 12, 816-824.

Okuda, K., de Almeida, F., Mortara, R.A., Krieger, H., Marinotti, O., Bijovsky, A.T., 2007. Cell death and regeneration in the midgut of the mosquito, Culex quinquefasciatus. Journal of Insect Physiology, 53, 1307-1315.

Overgaard, J., Tollarova, M., Hedlund, K., Petersen, S.O., Holmstrup, M., 2009. Seasonal changes in lipid composition and glycogen storage associated with freeze-tolerance of the earthworm, Dendrobaena octaedra. Journal of Comparative Physiology. B, Biochemical, Systemic, and Environmental Physiology, 179, 569-577.

Renault, D., Hance, T., Vannier, G., Vernon, P., 2003. Is body size an influential parameter in determining the duration of survival at low temperatures in Alphitobius diaperinus Panzer (Coleoptera : Tenebrionidae)? Journal of Zoology, 259, 381-388.

Renault, D., Nedved, O., Hervant, F., Vernon, P., 2004. The importance of fluctuating thermal regimes for repairing chill injuries in the tropical beetle Alphitobius diaperinus (Coleoptera: Tenebrionidae) during exposure to low temperature. Physiological Entomology, 29, 139-145.

Rinehart, J.P., Hayward, S.A.L., Elnitsky, M.A., Sandro, L.H., Lee, R.E., Denlinger, D.L., 2006. Continuous up-regulation of heat shock proteins in larvae, but not adults, of a polar insect. Proceedings of the National Academy of Sciences of the United States of America, 103, 14223-14227.

Sim, C., Denlinger, D.L., 2009. Transcription profiling and regulation of fat metabolism genes in diapausing adults of the mosquito Culex pipiens. Physiological Genomics, 39, 202-209.

Sinclair, B.J., Addo-Bediako, A., Chown, S.L., 2003. Climatic variability and the evolution of insect freeze tolerance. Biological Reviews, 78, 181-195.

Sinclair, B.J., Chown, S.L., 2005a. Climatic variability and hemispheric differences in insect cold tolerance: support from southern Africa. Functional Ecology, 19, 214-221.

Sinclair, B.J., Chown, S.L., 2005b. Deleterious effects of repeated cold exposure in a freeze-tolerant sub-Antarctic caterpillar. Journal of Experimental Biology, 208, 869-879.

Storey, K.B., Storey, J.M., 1991. Biochemistry of cryoprotectants, in: Lee, R.E., Denlinger, D.L. (Eds.), Insects at Low Temperature. Champman and Hall, New York, pp. 64-93.

147

Sugg, P., Edwards, J.S., Baust, J., 1983. Phenology and life-history of Belgica antarctica, an Antarctic midge (Diptera, Chironomidae). Ecological Entomology, 8, 105-113.

Thomas, E.R., Marshall, G.J., McConnell, J.R., 2008. A doubling in snow accumulation in the western Antarctic Peninsula since 1850. Geophysical Research Letters, 35, L01706.

Yi, S.-X., Lee, R.E., 2003. Detecting freeze injury and seasonal cold-hardening of cells and tissues in the gall fly larvae, Eurosta solidaginis (Diptera : Tephritidae) using fluorescent vital dyes. Journal of Insect Physiology, 49, 999-1004.

148

Figures

Figure 4.1. Effect of RCE on larval survival (a) and cell survival (b) of B. antarctica larvae. Values are mean ± s.e.m. Larval survival values are based on percentage survival of 5 replicates of 15 larvae each. Cell survival of each replicate is based on the mean count of 3 groups of 100 cells, and the mean for each treatment is based on 4 replicates. Closed circles denote control larvae, open circles denote frozen larvae, and triangles denote supercooled larvae. At each time point, “A” indicates a significant difference between control and frozen, “B” represents a significant difference between control and supercooled, and “C” represents a significant difference between frozen and supercooled (ANOVA, Bonferroni, p<0.05).

149

Figure 4.2. Lipid (a), glycogen (b), trehalose (c), glucose (d), and glycerol (e) content of B. antarctica larvae during RCE. Values are mean ± s.e.m. The metabolite content is based on the mean of 5 replicates, 5 larvae per replicate. Closed circles denote control larvae, open circles denote frozen larvae, and triangles denote supercooled larvae. At each time point, “A” indicates a significant difference between control and frozen, “B” represents a significant difference between control and supercooled, and “C” represents a significant difference between frozen and supercooled (ANOVA, Bonferroni, p<0.05). 150

Figure 4.3. Total energy content (a) and carbohydrate energy content (b) due to the major energy stores of B. antarctica larvae during RCE. Values are mean ± s.e.m. The TEC, in J/mg dry mass, is calculated by the equation TEC=0.393(lipid content) + 0.176(total carbohydrate content), where lipid content and carbohydrate content are in μg/mg dry mass. The CEC, in J/mg dry mass, is calculated by the equation CEC= 0.176(total carbohydrate content) (Djawdan et al., 1998). Values are based on the mean of 5 individuals. Closed circles denote control larvae, open circles denote frozen larvae, and triangles denote supercooled larvae. At each time point, “A” indicates a significant difference between control and frozen, “B” represents a significant difference between control and supercooled, and “C” represents a significant difference between frozen and supercooled (ANOVA, Bonferroni, p<0.05).

151

Figure. 4.4. mRNA expression of hsp70 during RCE in B. antarctica larvae. 28s rRNA gene is used as a control.

152

Chapter 5: Energetic Consequences of Repeated and Prolonged Dehydration in the Antarctic Midge, Belgica antarctica

Abstract

Larvae of the Antarctic midge, Belgica antarctica, routinely face periods of limited water availability in their natural environments on the Antarctic Peninsula. As a result, B. antarctica is one of the most dehydration-tolerant insects studied, surviving up to 70% loss of its body water. While previous studies have characterized the physiological effects of a single bout of dehydration, in nature larvae are likely to experience multiple bouts of dehydration throughout their lifetime. Thus, we examined the physiological consequences of repeated dehydration and compared results to larvae exposed to a single, prolonged period of dehydration. For the repeated dehydration experiment, larvae were exposed to 1-5 cycles of 24 h dehydration at 75% RH followed by 24 h rehydration. Each bout of dehydration resulted in 30-40% loss of body water, with a concomitant 2-3 fold increase in body fluid osmolality. While nearly 100% of larvae survived a single bout of dehydration, <65% of larvae survived five such cycles.

Larvae subjected to multiple bouts of dehydration also experienced severe depletion of carbohydrate energy reserves; glycogen and trehalose content decreased with each successive cycle, with larvae losing 89 and 48% of their glycogen and trehalose, respectively, after five cycles of dehydration/rehydration. Larvae exposed to prolonged dehydration (99% RH for 10 d) had 26% less water, 43% less glycogen, and 27% less

153

lipid content than controls, but did not experience any mortality. Thus, both repeated and prolonged dehydration results in substantial energetic costs that are likely to negatively impact fitness.

Introduction

The Antarctic midge, Belgica antarctica, is the world’s southernmost insect and the only insect endemic to Antarctica (Sugg et al., 1983). Found exclusively on the west coast of the Antarctica Peninsula and its surrounding islands, B. antarctica has a patchy distribution and is typically found in nutrient-rich areas associated with penguin and seal rookeries. The midge has a 2-year life cycle in which larvae feed during the brief austral summer and overwinter as any of 4 larval instars (Convey and Block, 1996). Apterous adults synchronously emerge for a brief period in the summer, during which they mate, oviposit, and die within 10 days.

Not surprisingly, larvae are extremely tolerant of a number of environmental stresses, including low temperature (Lee et al., 2006), dehydration (Benoit et al., 2007;

Hayward et al., 2007), oxidative stress (Lopez-Martinez et al., 2008), and osmotic perturbations (Elnitsky et al., 2009). While Antarctic terrestrial organisms cope with a number of environmental insults, water availability is the primary factor governing the distribution of terrestrial organisms on the continent (Kennedy, 1993). For much of the year, water is frozen and therefore biologically unavailable. Even in the summer months, when liquid water is abundant, there can be spatial and temporal differences in microhabitat water availability (Benoit et al., 2007; Kennedy, 1993).

154

Thus, throughout its life cycle, B. antarctica is exposed to numerous periods of desiccating conditions. At ecologically relevant relative humidities, larvae of B. antarctica tolerate up to 70% loss of their body water, the highest level of dehydration tolerance measured in a polar arthropod (Hayward et al., 2007). Additionally, the high cuticular water permeability and dehydration tolerance of larvae allow them to use cryoprotective dehydration as a means to survive prolonged periods of subzero exposure

(Elnitsky et al., 2008). In response to dehydration, larvae accumulate osmoprotectants such as glycerol and trehalose, alter their cuticular hydrocarbon profile, and decrease their respiration rate as a means to conserve water and enhance dehydration tolerance

(Benoit et al., 2007). Several other metabolic changes are elicited by dehydration, including elevation of specific polyols and amino acids, although the adaptive benefits of these changes remain unexplored (Michaud et al., 2008). Furthermore, a number of stress related genes, including several heat shock proteins and genes involved in oxidative stress, are upregulated in response to dehydration (Lopez-Martinez et al., 2009). At the protein level, dramatic changes in the abundance of cytoskeletal proteins occur during dehydration, presumably to counteract the change in body size accompanying dehydration (Li et al., 2009).

Despite obvious benefits to the larvae, many of these physiological changes during dehydration would appear to be energetically costly. Also, additional energy input is required during rehydration to repair the accrued damage. For example, the cellular machinery responsible for restoring perturbed ion gradients during desiccation consumes a large amount of energy (Harvey et al., 1998). In Drosophila melanogaster, the link between energy reserves and the ability to survive dehydration is well established; 155

populations of D. melanogaster selected for desiccation resistance have increased carbohydrate stores, which provide additional energy to acclimate and recover from dehydration and bind bulk water to increase the pool of available water (Chippindale et al., 1998; Djawdan et al., 1998). However, the energetic consequences of dehydration in

B. antarctica, or any other polar species, have not been examined.

Previous studies on dehydration in B. antarctica only assessed physiological changes in response to a single bout of dehydration, whereas in nature larvae are likely to be exposed to numerous rounds of dehydration and rehydration. While most studies of stress tolerance have only used a single bout of stress, several recent studies have demonstrated the importance of using repeated stress exposures that better reflect natural conditions. Previously, we demonstrated that repeated freeze/thaw cycles in larvae of B. antarctica cause significant mortality, damage to midgut tissue, expression of stress- related genes, and a severe reduction in energy reserves, and that these effects worsen with each freeze/thaw cycle (Teets et al., 2011). While several other studies have addressed the question of repeated cold exposure (Colinet et al., 2007; Marshall and

Sinclair, 2010, 2011), few have addressed the consequences of repeated dehydration in insects. In Drosophila, prior exposure to dehydration can rapidly enhance dehydration- tolerance, primarily by altering cuticular water loss rates (Bazinet et al., 2010; Hoffmann,

1991). Meanwhile, only one study has addressed the consequences of repeated, cyclic bouts of dehydration; in the mosquito, Culex pipiens, repeated bouts of dehydration cause a significant reduction in energy reserves and fecundity, effects that are not measurable after a single bout of dehydration (Benoit et al., 2010). In all cases, the use of repeated

156

stress exposures reveals novel physiological changes that cannot be predicted from a single exposure to stress.

Thus, our primary objectives in this study were to 1) measure the effects of repeated dehydration and rehydration on the water balance and osmolality of B. antarctica larvae, 2) determine the ability of larvae to survive multiple bouts of severe dehydration and 3) compare the energetic consequences of repeated and prolonged dehydration on the energy balance and osmoprotectant levels. Together, these experiments provide the first overview of the dynamics and energetic consequences of dehydration that confront a polar insect.

Materials and Methods

Insects

Larvae of B. antarctica were collected from Humble, Cormorant, and Christine

Islands near Palmer Station on the Antarctic Peninsula (64°46’ S, 64°04’ W) in January and February of 2011. Larvae were extracted from substrate (containing soil, rocks, and detritus) into ice water using a modified Berlese apparatus, and concentrated samples of larvae were placed back in their natural substrate and held at 2°C until the time of experiments. On the day prior to beginning an experiment, larvae were hand-sorted from substrate in ice water and placed on moist filter paper overnight to standardize body water content. Only fourth instar larvae were used in this study.

157

Experimental Conditions

Temperature was maintained at 2°C for the duration of the experiment. Different relative humidities were generated in sealed desiccators using saturated salt solutions according to Winston and Bates (1960); 75% RH was obtained with a saturated solution of NaCl, while 99% humidity was obtained with a saturated solution of K2SO4. Larvae were placed in groups of ten in microcentrifuge tubes with mesh tops for dehydration exposures, while control animals were held in groups of ten in vials with moist paper towel for the duration of the experiment. For the repeated dehydration treatment, larvae were placed at 75% RH for 24 h, followed by 24 h of rehydration by transferring the larvae to a moist paper towel. This cycle was repeated for a total of five dehydration and rehydration cycles. Water content and osmolality were measured each day for 10 d, while samples for survival, cell survival, and metabolite analyses were taken every 2 d after the rehydration phase of the treatment. For the prolonged dehydration treatment, larvae were held for 10 d at 99% RH, followed by 24 h of rehydration on a moist paper towel. Water content and osmolality were determined after both the 10 d dehydration and rehydration, while larval survival, cell survival, and metabolite contents were assessed after rehydration.

Water content and osmolality

On each day of the experiment, water content was determined gravimetrically.

From each group, 15 individuals were weighed to the nearest 0.2 μg on an electrobalance and immediately transferred to a drying oven at 65°C. Individuals were weighed to

-1 constant dryness, and body water content, expressed as mg H2O mg dry mass (DM), was calculated for each individual. 158

Body fluid osmolality was determined using a vapor pressure depression technique according to Elnitsky et al. (2008). Briefly, five individuals from each treatment were crushed with a teflon pestle, sealed in a C-52 sample chamber (Wescor

Inc., Logan, UT), and allowed to equilibrate for 30 min at room temperature. The osmolality was then measured using a Wescor HR-33T Dew Point Microvoltmeter

(Wescor Inc., Logan, UT), with values being compared to standard curves produced from solutions of known salt concentration (Opti-Mole, Wescor Inc., Logan, UT). For each treatment, osmolality was based on the mean of five biological replicates of five individuals.

Survival and cell survival

At the end of each dehydration/rehydration cycle, survival was assessed by checking for larval movement. Concurrently, survival was checked for control animals that were held on moist paper towel. Groups of ten larvae were placed in a drop of water under a microscope, and those that either moved spontaneously or in response to gentle prodding were considered alive. For each treatment, survival was based on five replicates of ten larvae each.

To test for sublethal tissue damage caused by dehydration, cell survival of midgut tissue was measured according to Yi and Lee (2003). The midgut was selected because it is highly susceptibile to damage from freezing (Teets et al., 2008) and because it is amenable to the staining procedures of this assay. After each rehydration cycle, the midgut was dissected in Coast’s solution (Coast and Krasnoff, 1988) and stained with the

LIVE/DEAD sperm viability kit (Invitrogen, Carlsbad, CA). Tissue samples were stained

159

for 15 min in SYBR followed by an additional 15 min in propidium iodide. After staining, images were obtained using fluorescent microscopy at the Ohio State Campus

Microscopy and Imaging Facility. The nuclei of living cells with intact cell membranes appear green, while dead cells with damaged membranes are red. Cell viability for each sample was determined by counting three groups of 100 cells and calculating the mean proportion of green cells from these three counts. For each treatment, viability was based on the mean of six biological replicates.

Metabolite assays

To assess the effects of dehydration on energy stores and osmoprotectant levels, we conducted assays for total lipid, triglycerides, glycogen, trehalose, glucose, and glycerol. After each dehydration/rehydration cycle, ten groups of five individuals were weighed and quickly frozen at -70°C for lipid and triglyceride assays (five groups for lipid, five groups for triglycerides). For water-soluble metabolites, five groups of 20 individuals were weighed and frozen from each treatment. Samples were shipped frozen on dry ice from Palmer Station, Antarctica to Ohio State University, where the assays were conducted. Samples for the water-soluble metabolites were homogenized in 1M perchloric acid, centrifuged at 14,000Xg, and the supernatant was neutralized with 1 M

KOH. Neutralized extracts were stored at -70°C until analysis.

Total lipid content was measured using vanillin-phosphoric acid reagent according to Sim and Denlinger (2009). Groups of five individuals were homogenized in

500 μl 1:1 chloroform:methanol and centrifuged, and the solvent was evaporated by heating at 90°C. Samples were then re-suspended in H2SO4, followed by the addition of

160

vanillin-phosphoric acid reagent. Absorbance was read at 490 nm in a NanoDrop 2000C spectrophotometer (Thermo Scientific, Waltham, MA) and compared to absorbance of known standards containing canola oil dissolved in chloroform.

Triglycerides were measured according to Marshall and Sinclair (2010), with the following modifications: Samples of five larvae were homogenized in 1:1 chloroform:methanol and centrifuged to remove pigments. The supernatant was evaporated by heating at 90°C, and the sample was re-suspended in 0.05% Tween-20.

Triglyceride content was measured using triglyceride reagent (T2449, Sigma, St. Louis,

MO) to liberate glycerol from triglyerides. Glycerol content was then measured using free glycerol reagent (F6428, Sigma, St. Louis, MO) by measuring absorbance at 540 nm and comparing to known glycerol standards.

Glycogen content was measured in perchloric acid extracts by first using amyloglucosidase from Aspergillus niger (A1602, Sigma, St. Louis, MO) to liberate free glucose (Kepler and Decker, 1984). Free glucose was then assayed using a glucose assay kit (GAGO20, Sigma, St. Louis, MO), and the absorbance at 540 nm was compared to standards of known glycogen concentration. Trehalose content was measured according to Chen et al. (2002), by using trehalase from porcine kidney (T8778, Sigma, St. Louis,

MO) to hydrolyze trehalose into glucose monomers. The resulting glucose content was then measured as described previously. In each sample, glucose content was also measured in undigested perchloric acid extracts and used to correct the values of glycogen and trehalose. Finally, glycerol content in perchloric acid extracts was measured with free glycerol reagent according to Teets et al. (2011). For each metabolite,

161

values were based on the mean of five replicates and are reported as μg metabolite mg-1

DM.

Total energy and carbohydrate energy content

To summarize the results of the energy reserve assays, the total energy content

(TEC) and carbohydrate energy content (CEC) values were calculated according to

Djawdan et al. (1998). Using thermodynamic values for lipid and carbohydrate, the formula for TEC is TEC=0.0393(lipid content) + 0.0176TCC, where TEC is in J mg-1

DM, lipid content is in μg lipid mg-1DM, and TCC is total carbohydrate content in μg carbohydrate mg-1 DM and is the sum of glycogen, trehalose, and glucose content.

Because TEC predominantly reflects lipid content, we also calculated CEC, in J mg-1

DM, using the formula CEC=0.0176CCC.

Statistical analyses

All statistical analyses were conducted with JMP 9 (SAS Institute Inc., Cary, NC).

Means were compared using ANOVA with a post-hoc pooled T-test. A false discovery rate correction (Benjamini and Hochberg, 1995) with α=0.05 was applied to each hypothesis test to control for the probability of a type I error. Based on the number of hypothesis tests conducted, the approximate false discovery rate for the entire study was

0.025. Whole-animal and cell survival data were arcsin-square-root-transformed prior to analyses. Statistical significance was set at P<0.05, and all data are reported as mean±SE.

162

Results

Water content and osmolality

To facilitate discussing the results, odd numbered days (i.e. 1, 3, 5, 7 and 9) represent the dehydration phase of the dehydration/rehydration cycle, while even numbered days (i.e. 2, 4, 6, 8 and 10) represent the rehydration phase. Also, for convenience, larvae subjected to repeated dehydration/rehydration will be referred to RD larvae, while those subjected to prolonged dehydration (i.e. 99% for 10 d) will be referred to as PD larvae. After each dehydration cycle, the water content of RD larvae was between 26 and 48% less than control larvae, differences that were statistically significant in all cases (P<0.05). After the first rehydration event, the water content of RD

-1 larvae (2.41±0.11 μg H2O mg DM) remained significantly lower than the water content

-1 of control larvae (2.92±0.09 μg H2O mg DM; P<0.05). However, after each subsequent rehydration cycle the water content of RD larvae was statistically indistinguishable from that of control larvae (P>0.05). Over the course of the experiment, the water content of

RD animals after both dehydration and rehydration increased slightly (Fig. 1A). The

-1 water content of RD larvae on days 7 and 9 (1.88±0.12 and 2.15±0.06 mg H2O mg DM, respectively) was significantly greater than in RD larvae on days 3 and 5 (1.46±0.07 and

-1 1.47±0.14 mg H2O mg DM, respectively; P<0.05). Likewise, the water content of RD larvae on days 6, 8, and 10 was significantly higher than that of RD larvae on day 2

(P<0.05). In the prolonged dehydration experiment, water content of PD larvae was 26% less than that of control larvae after day 10 (P<0.05) but returned to control levels following rehydration (P>0.05; Table 1).

163

In the repeated dehydration experiment, changes in body fluid osmolality tracked the observed changes in water content (Fig. 1B). Following each of the five dehydration events, body fluid osmolality of RD larvae was significantly higher than that of control larvae (P<0.05), while there was no difference between RD and control larvae after each cycle of rehydration (P>0.05). The greatest difference in osmolality was observed on day

7, when the osmolality of control larvae was 466.0±45.4 mOsm kg-1 while that of RD larvae was 1519.4±290.9 mOsm kg-1 (P<0.05). Following prolonged dehydration, the osmolality of PD larvae was 889.9±83.0 mOsm kg-1, compared to 616.9±59.9 mOsm kg-1 in control larvae, but this difference was not statistically significant (P>0.05; Table 1).

Likewise, following rehydration, there was no difference in osmolality between control and PD larvae (P>0.05).

Whole-animal and cell survival

Throughout much of the repeated dehydration experiment, there were no significant differences in survival between control and RD larvae (Table 2). For the first four dehydration/rehydration cycles, survival ranged from 94.4±4.0 to 100.0±0.0% for control larvae and from 92.8±4.9 to 98.0±2.0% for RD larvae. However, after the fifth cycle, survival of RD larvae (63.9±12.3%) was significantly lower than that of corresponding control larvae (96.2±2.3%; P<0.05). In the prolonged dehydration experiment, there was no difference in survival between PD larvae (100.0±0.0%) and control larvae (98.0±2.0%; P>0.05; Table 1).

In the repeated dehydration experiment, cell survival of midgut tissue in RD larvae was consistently lower than that of control larvae (Table 2). After each of the five

164

dehydration/rehydration cycles, cell survival of RD larvae was significantly lower than that of control larvae, ranging between 9.4 and 18.5% lower than their control counterparts (P<0.05). In the prolonged dehydration experiment, there was no difference in cell survival between control and PD larvae (P>0.05; Table 1).

Energy reserve and osmoprotectant levels

For the entire repeated dehydration experiment, total lipid levels did not vary across treatments, ranging from 171.0±10.2 to 201.8±11.2 μg lipid mg-1 DM (Figure 2A;

P>0.9827). Likewise, triglyceride levels did not significantly differ between control and

RD larvae after any of the dehydration/rehydration cycles (Figure 2B; P>0.05), ranging from 79.2±6.2 to 128.8±10.3 μg triglyceride mg-1 DM. Following prolonged dehydration and rehydration, PD larvae had 27% less lipid than their control counterparts, a difference that was significant (P<0.05; Fig. 3A). However, the observed difference in triglyceride content between control (88.5±11.8 μg triglyceride mg-1 DM) and PD (73.8±7.1 μg triglyceride mg-1 DM) larvae was not significant (P>0.05).

While repeated dehydration had little effect on lipid energy reserves, carbohydrate energy reserves were dramatically altered by repeated dehydration. After each of the five cycles of dehydration/rehydration, the glycogen content of RD larvae was significantly lower than that of control larvae (P<0.05; Fig. 2C). The glycogen content of RD larvae continued to decrease with each successive cycle, decreasing from 29.5±2.4 μg glycogen mg-1 DM after one dehydration/rehydration cycle to 4.9±0.9 μg glycogen mg-1 DM after five dehydration/rehydration cycles. By the fifth cycle of dehydration/rehydration, the glycogen content of RD larvae was 89% lower than their control counterparts. Prolonged

165

dehydration likewise caused a significant reduction in glycogen, although not to the levels seen in the RD larvae. After rehydration, the glycogen content of PD larvae

(31.9±3.9 μg glycogen mg-1 DM) was 43% less than that of control larvae (55.9±1.6 μg glycogen mg-1 DM; P<0.05; Fig. 3A).

The other major carbohydrate energy store, trehalose, was also significantly depleted by repeated bouts of dehydration/rehydration (Fig. 2D). After one dehydration/rehydration cycle, the trehalose content of RD larvae (42.8±0.8 μg trehalose mg-1 DM) did not differ from the control value (48.3±1.5 μg trehalose mg-1 DM).

However, trehalose levels of RD larvae were gradually diminished with each cycle and were significantly lower than control larvae for cycles 2-5 (P<0.05). After the fifth cycle of dehydration/rehydration, trehalose levels were 48% lower in RD larvae compared to controls. In the prolonged dehydration experiment, there was no difference in trehalose content between control (52.6±0.5 μg trehalose mg-1 DM) and RD (46.4±1.4 μg trehalose mg-1 DM) larvae (Fig. 3A).

In addition to the major energy reserves, we also measured levels of two osmoprotectants in B. antarctica, glucose and glycerol. Glucose levels did not change appreciably throughout the first four cycles of repeated dehydration/rehydration (Fig 2D); however, after the fifth cycle, the glucose content of RD larvae sharply increased to

0.46±0.06 μg glucose mg-1 DM, which was significantly higher than control larvae

(0.10±0.02 μg glucose mg-1 DM; P<0.05). In the prolonged dehydration experiment, there was no difference in glucose content between PD and control larvae (P>0.05; Fig.

3A). Meanwhile, glycerol levels followed a similar pattern to glucose (Fig. 2E). There

166

were no observed differences in glycerol content until after the fifth dehydration/rehydration cycle, where glycerol levels were 2.6-fold greater in RD larvae

(P<0.05). As with glucose, there were no differences in glycerol between control and PD larvae (P>0.05; Fig. 3A).

Total energy and carbohydrate energy content

Using thermodynamic values for lipid and carbohydrate energy content, we calculated the change in energy content of larvae in response to dehydration and rehydration. In the repeated dehydration experiment, TEC was nominally lower in RD larvae than in control larvae after each cycle (Fig. 4A), ranging from 8.5±0.3 to 9.6±0.5 J mg-1 DM in control larvae and 7.3±0.4 and 8.5±0.3 J mg-1 DM in RD larvae. However, these differences were not statistically significant (P>0.05). Because lipid levels, which did not change significantly, predominantly influence TEC we also calculated the CEC.

Unlike TEC, there were dramatic changes in CEC (Fig. 4B); after each of the five cycles of dehydration and rehydration, the CEC was significantly lower in RD larvae (P<0.05).

By the fifth cycle, the CEC was 67% lower in RD larvae than in their control counterparts. In contrast to the repeated dehydration treatment, prolonged dehydration caused a significant decrease in both TEC and CEC (Fig. 3B). Following prolonged dehydration and rehydration, the TEC was 27% lower in PD larvae, and the CEC was

28% lower.

167

Discussion

While previous studies have elucidated a number of physiological responses to dehydration in larvae of B. antarctica, the present study is the first to address the physiological consequences of repeated dehydration in a polar insect. During the repeated dehydration trials, larvae lost 30-40% of their total body water during each 24 h exposure to 75% RH (Fig. 1A). However, the dynamics of dehydration and rehydration changed noticeably over the course of the experiment. After the first bout of dehydration/rehydration, larvae failed to fully rehydrate within 24 h, with water levels being 17% lower in RD larvae on day 2 of the experiment. With each successive cycle, larvae took in more water during rehydration, with water content in RD larvae peaking on

-1 day 8 at 3.02±0.12 mg H2O mg DM. Also, over time, there was a trend towards lower rates of water loss; for example, between days 8 and 9, larvae lost 29% of their water, compared to a 39% loss of body water between days 2 and 3. While we did not probe the mechanisms of this reduced water loss, we suspect that changes in cuticular hydrocarbons may be responsible (Benoit et al., 2007). Ultimately, these changes resulted in larvae having significantly higher water content following dehydration on days 7 and 9 compared to days 3 and 5, when the lowest water levels were observed.

Thus, it appears these changes are adaptive as they allow larvae to maintain higher water content during each successive bout of dehydration. In general, insects can increase their desiccation resistance by either 1) increasing water content, 2) reducing water loss rates, or 3) increasing the amount of water that can be lost before death (Gibbs et al., 1997). Here, we observed evidence for both 1 and 2 in response to repeated dehydration, although a thorough study of water balance parameters is needed to 168

substantiate these claims. Because these results were somewhat unexpected, our study was not designed to fully test the mechanisms by which the dynamics of dehydration/rehydration change in RD larvae. Also, it is worth noting that because water content is expressed as µg H2O/mg DM, a decrease in DM over the course of the experiment due to respiration could skew the results slightly. However, because of the large variation in size in field-collected larvae, expressing water content per individual was not practical.

To test how repeated dehydration affected the osmotic balance of larvae, we measured changes in body fluid osmolality during repeated dehydration and rehydration.

On the whole, osmolality of larvae closely tracked observed changes in water content

(Fig 1B). After each dehydration cycle, there was a significant increase in osmolality, which is not surprising due to the concentration of solutes following evaporation of water. However, the reduction in water content alone cannot account entirely for the observed changes in osmolality. For example, between days 2 and 3, the osmolality increased by a factor of 3, while the change in water content could account for only a 1.7- fold increase in osmolality. Thus, it appears additional osmolytes are produced during dehydration, although the identity of these osmolytes is unknown. Elnitsky et al. (2008) observed similar results in response to cryoprotective dehydration, with some of the increased osmolality coming from glucose and trehalose synthesis. In the present study, we did not observe any accumulation of osmoprotectants (i.e. trehalose, glucose, and glycerol), although metabolite composition was only measured after recovery. Thus, it’s possible that additional osmolytes were accumulated during dehydration but recycled into insoluble energy stores during recovery. 169

As expected, the experimental perturbations in water content had substantial costs. While larvae survived the first four cycles of dehydration/rehydration very well, by the fifth cycle there was significant mortality in RD larvae (Table 2). The exact cause of mortality is unclear, but it could be related to damage to midgut tissue (Table 2).

Compared to control larvae, midgut cell mortality was between 2.5 and 6 times greater in

RD larvae. However, midgut mortality did not increase with each successive cycle, suggesting there is some repair of dead cells during rehydration. Because of the substantial volume changes that accompany dehydration stress in B. antarctica (Li et al.,

2009), the midgut may be particularly sensitive to dehydration since it occupies such a large proportion of the animal’s total volume. On the whole, the observed differences in midgut cell survival were very similar to our previous study of repeated cold exposure in

B. antarctica (Teets et al., 2011), suggesting that both repeated freezing and repeated dehydration cause considerable damage to the midgut.

In addition to the above effects on larval mortality and tissue damage, repeated dehydration had severe energetic consequences for larvae. While total lipid and triglyceride levels were unaffected by repeated dehydration (Fig. 2A,B), we observed a severe depletion of carbohydrate energy reserves in response to dehydration. After five cycles of dehydration/rehydration, RD larvae had 89% less glycogen and 48% less trehalose than their control counterparts (Fig. 2C,D), corresponding to a 67% loss of carbohydrate energy content (Fig. 4B). These results were analogous to our previous study on repeated cold exposure, where frozen larvae lost ~40% of their glycogen and

~25% of their trehalose (Teets et al., 2011).

170

The depletion of glycogen during repeated dehydration is in line with hypotheses of dehydration adaptations in other species. Glycogen is often accumulated in desiccation-resistant Drosophila strains, both to serve as an energy source and as a means of sequestering additional water, since glycogen can bind 3-5x its weight in water (Gibbs et al., 1997). Using these estimates, the observed breakdown of glycogen would provide

-1 an additional 120-200 μg H2O mg DM over the course of the study, which is roughly 4-

7% of the larvae’s total body water after five cycles of dehydration/rehydration. Also, since glycogen is the primary source of glucose for osmoprotectant production in insects

(Storey, 1997), glycogen depletion may be in part due to osmoprotectant production during dehydration. The observed increase in glucose and glycerol after cycle 5 is one possible fate of glycogen, although these modest increases only accounted for <1% of the total glycogen depletion.

One surprising result of this study was the observed depletion of trehalose, the other major carbohydrate energy store. Trehalose is accumulated during dehydration in B. antarctica (Benoit et al., 2007; Elnitsky et al., 2008), and is perhaps the most commonly used solute to protect against dehydration in insects (Danks, 2000). Thus, using trehalose as an energy source for recovery would seem to be maladaptive if it generates a deficit of osmoprotectants in the next dehydration cycle. Also, as with glycogen, the accumulation of glucose and glycerol after cycle 5 represented only a fraction of the amount of trehalose lost, so conversion of trehalose to other carbohydrates does not appear to be a major factor here. Perhaps, the inability to maintain adequate trehalose content is partially responsible for mortality during the last cycle of repeated dehydration.

171

In addition to quantifying the energetic consequences of repeated dehydration, a secondary goal of this study was to measure the energetic costs of gradual, prolonged dehydration on larvae. Thus, the prolonged dehydration experiment was designed to assess how the duration and severity of dehydration exposure affected the energy balance of larvae. The 10 d prolonged dehydration treatment encompassed the same amount of time as five cycles of repeated dehydration (i.e. 2 d per cycle), allowing us to make energetic comparisons between the two treatments. We anticipated that prolonged, gradual dehydration would be less energetically costly than rapid dehydration, because during gradual dehydration larvae 1) become drought-acclimated, thereby reducing the damage caused by dehydration (Hayward et al., 2007), and 2) decrease their respiration rates, presumably to conserve energy while they are dehydrated and unable to feed

(Benoit et al., 2007). However, somewhat surprisingly, while prolonged dehydration at

99% RH did not deplete carbohydrate energy to the extent of repeated dehydration, prolonged dehydration did cause a significant reduction in lipid content. As a result, PD larvae had 27% less TEC than their control counterparts, while after five cycles RD larvae only had a 15% deficit in TEC, a difference that was not statistically significant.

This suggests that larvae at 99% RH maintain a higher average meatabolic rate than larvae at both 75 and 100% RH. However, further experiments using identical relative humidities are needed to directly compare the energetic consequences of repeated and prolonged dehydration.

All told, the energy depletion observed in response to dehydration caused by both repeated and prolonged dehydration likely has fitness consequences for B. antarctica. In the mosquito Culex pipiens, repeated bouts of dehydration cause a similar reduction in 172

energy stores that lead to reduced egg production by adult females (Benoit et al., 2010).

Similarly, in Drosophila, repeated bouts of cold cause a concurrent reduction in both glycogen stores and fecundity (Marshall and Sinclair, 2010). While we are unable to raise larvae to adulthood and directly measure fitness, the depletion of energy stores likely reflects a diversion of energy from growth and reproduction. This may be especially problematic for polar insects, which have an extremely short growing season during which to replenish energy stores. Also, our results have implications related to climate change on the Antarctic Peninsula. While total precipitation has increased in response to climate warming, climate change has caused localized changes in snow melt and evaporation that could result in greater fluctuations in moisture regimes in the B. antarctica habitat (Fox and Cooper, 1998). Thus, even though B. antarctica is the most dehydration-tolerant polar insect known (Hayward et al., 2007), there are dramatic energetic costs to dehydration and rehydration that likely restrict the potential habitat of

B. antarctica in nature.

Acknowledgements

We are especially grateful for the hard work and assistance of the Palmer Station support staff during our field season. We also thank Mary Hahn for assistance in collecting data during a pilot study for these experiments. Microscopic images were generated using the instruments and services at the Campus Microscopy and Imaging

Facility, The Ohio State University.

173

Grants

This work was supported by NSF grants ANT-0837559 and ANT-0837613.

References

Bazinet, A.L., Marshall, K.E., MacMillan, H.A., Williams, C.M., Sinclair, B.J., 2010. Rapid changes in desiccation resistance in Drosophila melanogaster are facilitated by changes in cuticular permeability. Journal of Insect Physiology, 56, 2006-2012.

Benoit, J.B., Patrick, K.R., Desai, K., Hardesty, J.J., Krause, T.B., Denlinger, D.L., 2010. Repeated bouts of dehydration deplete nutrient reserves and reduce egg production in the mosquito Culex pipiens. Journal of Experimental Biology, 213, 2763-2769.

Chippindale, A.K., Gibbs, A.G., Sheik, M., Yee, K.J., Djawdan, M., Bradley, T.J., Rose, M.R., 1998. Resource acquisition and the evolution of stress resistance in Drosophila melanogaster. Evolution, 52, 1342-1352.

Convey, P., Block, W., 1996. Antarctic Diptera: Ecology, physiology and distribution. European Journal of Entomology, 93, 1-13.

174

Danks, H.V., 2000. Dehydration in dormant insects. Journal of Insect Physiology, 46, 837-852.

Djawdan, M., Chippindale, A.K., Rose, M.R., Bradley, T.J., 1998. Metabolic reserves and evolved stress resistance in Drosophila melanogaster. Physiological Zoology, 71, 584-594.

Fox, A.J., Cooper, A.P.R., 1998. Climate-change indicators from archival aerial photography of the Antarctic Peninsula. Annals of Glaciology, 27, 636-642.

Gibbs, A.G., Chippindale, A.K., Rose, M.R., 1997. Physiological mechanisms of evolved desiccation resistance in Drosophila melanogaster. Journal of Experimental Biology, 200, 1821-1832.

Harvey, W.R., Maddrell, S.H.P., Telfer, W.H., Wieczorek, H., 1998. H+ V-ATPases energize animal plasma membranes for secretion and absorption of ions and fluids. American Zoologist, 38, 426-441.

Hoffmann, A.A., 1991. Acclimation for desiccation resistance in Drosophila - Species and population comparisons. Journal of Insect Physiology, 37, 757-762.

Kennedy, A.D., 1993. Water as a limiting factor in the Antarctic terrestrial environment - A biogeographical synthesis. Arctic and Alpine Research, 25, 308-315.

Kepler, D., Decker, K., 1984. Glycogen, in: Bergmeyer, H.U. (Ed.), Methods of Enzymatic Analysis. Verlag Chemie, Weinheim, Germany, pp. 11-18.

175

Marshall, K.E., Sinclair, B.J., 2010. Repeated stress exposure results in a survival- reproduction trade-off in Drosophila melanogaster. Proceedings of the Royal Society of London. Series B: Biological Sciences, 277, 963-969. Marshall, K.E., Sinclair, B.J., 2011. The sub-lethal effects of repeated freezing in the woolly bear caterpillar Pyrrharctia isabella. Journal of Experimental Biology, 214, 1205- 1212.

Sim, C., Denlinger, D.L., 2009. Transcription profiling and regulation of fat metabolism genes in diapausing adults of the mosquito Culex pipiens. Physiological Genomics, 39, 202-209.

Storey, K.B., 1997. Organic solutes in freezing tolerance. Comparative Biochemistry and Physiology A-Physiology, 117, 319-326.

Sugg, P., Edwards, J.S., Baust, J., 1983. Phenology and life-history of Belgica antarctica, an Antarctic midge (Diptera, Chironomidae). Ecological Entomology, 8, 105-113.

Teets, N.M., Elnitsky, M.A., Benoit, J.B., Lopez-Martinez, G., Denlinger, D.L., Lee, R.E., 2008. Rapid cold-hardening in larvae of the Antarctic midge Belgica antarctica: cellular cold-sensing and a role for calcium. American Journal of Physiology-Regulatory Integrative and Comparative Physiology, 294, R1938-R1946.

176

Winston, P.W., Bates, D.S., 1960. Saturated salt solutions for the control of humidity in biological research. Ecology, 41, 232-237.

Yi, S.X., Lee, R.E., 2003. Detecting freeze injury and seasonal cold-hardening of cells and tissues in the gall fly larvae, Eurosta solidaginis (Diptera : Tephritidae) using fluorescent vital dyes. Journal of Insect Physiology, 49, 999-1004.

177

Tables

Table 5.1. Effect of prolonged dehydration (10 d at 99% RH) and rehydration (1 d at 100% RH) on water content, osmolality, survival, and midgut cell survival in larvae of B. antarctica. Values are presented as mean±SE, N=15 for water content, N=5 for osmolality and survival, N=6 for cell survival. Different letters within the same row represent significant differences between treatment groups (ANOVA, FDR, P<0.05). DM, dry mass.

Control, 10 Prolonged Control, 11 Prolonged days dehydration days dehydration + rehydration

Water content 2.48±0.06a 1.83±0.07b 2.54±0.07a 2.75±0.12a (mg H2O/mg DM)

Body fluid osmolality 616.9±59.9a 889.9±83.0a 552.0±21.3a 673.9±15.6a (mOsm/kg)

Survival (%) NA NA 98.0±2.0a 100.0±0.0a

Cell Survival (%) NA NA 92.0±2.8a 90.5±2.7a

178

Table 5.2. Effects of repeated bouts of dehydration and rehydration on whole- animal and midgut cell survival in larvae of B. antarctica. Each cycle of dehydration and rehydration consisted of 24 h at 75% RH followed by 24 h on a moist paper towel. Values are given as mean±SE, N=5 for organismal survival and N=6 for cell survival. An “*” indicates a significant difference between control and repeated dehydration larvae at a particular time point (ANOVA, FDR, P<0.05).

Number of Cycles of Dehydration/Rehydration

1 2 3 4 5

Control survival (%) 100.0±0.0 100.0±0.0 94.0±4.0 98.0±2.0 96.2±2.3

Dehydration/rehydration 98.0±2.0 94.4±3.7 92.8±4.9 94.0±4.0 63.9±12.3* survival (%)

Control cell survival (%) 93.7±2.2 96.5±1.6 90.9±3.0 96.3±1.1 92.0±2.8

Dehydration/rehydration 84.3±1.2* 78.0±1.0* 78.3±2.8* 79.5±3.8* 79.5±4.7* cell survival

179

Figures

Figure 5.1. Water content (A) and osmolality (B) of larvae of B. antarctica during repeated bouts of dehydration and rehydration. Each bout of dehydration and rehydration consisted of dehydration for 24 h at 75% RH followed by rehydration for 24 h on a moist paper towel. Odd numbered days represent days following dehydration while even numbered days represent days following rehydration. Each data point represents the mean±SE of 15 individuals for water content and for five samples for osmolality. An “*” indicates a significant difference between control and repeated dehydration larvae at a particular time point (ANOVA, FDR, P<0.05). DM, dry mass.

180

Figure 5.2. Lipid (A), triglyceride (B), glycogen (C), trehalose (D), glucose (E), and glycerol (F) content in larvae of B. antarctica during repeated dehydration/rehydration cycles. Metabolites were sampled at the end of each of five dehydration/rehydration cycles, consisting of 24 h dehydration at 75% RH followed by 24 h rehydration on a moist paper towel. For lipid and triglyceride, values are the mean±SE of five replicates of five individuals each, while for the other metabolites values are mean±SE of five replicates of 20 individuals. An “*” indicates a significant difference between control and repeated dehydration larvae at a particular time point (ANOVA, FDR, P<0.05). DM, dry mass. 181

Figure 5.3. Effect of prolonged dehydration on metabolite (A) and energy content (B) of larvae of B. antarctica. Prolonged dehydration consisted of 10 d at 99% RH followed by rehydration for 24 h on a moist paper towel. The bars represent the relative value of each parameter relative to the control; i.e. the control level is set at 1 for each metabolite or energy content. The actual measured values (mean) for each bar are included in parentheses. Error bars represent the SE. For lipid and triglyceride, values are the mean of five replicates of five individuals, while for the other metabolites values are the mean of five replicates of 20 individuals. An “*” indicates a significant difference between control and prolonged dehydration larvae for a particular pair of values (ANOVA, FDR, P<0.05).

182

Figure 5.4. Effect of repeated dehydration on the total energy content (TEC) (A) and carbohydrate energy content (CEC) (B) of larvae of B. antarctica. Larvae were exposed to five cycles of 24 h dehydration at 75% RH followed by 24 h rehydration on a moist paper towel. TEC and CEC were calculated according to Djawdan et al. (1998) using the following equation: TEC=0.0393(lipid content) + 0.0176(carbohydrate content), CEC=0.0176(carbohydrate content), where lipid and carbohydrate content are in μg mg-1 DM. Values are the mean±SE of five replicates. An “*” indicates a significant difference between control and repeated dehydration larvae at a particular time point (ANOVA, FDR, P<0.05).

183

Chapter 6: Expression of Genes Involved in Energy Mobilization and Osmoprotectant Synthesis During Thermal and Dehydration Stress in the Antarctic Midge, Belgica antarctica

Abstract

The Antarctic midge, Belgica antarctica, experiences sub-zero temperatures and desiccating conditions for much of the year, and in response to these environmental insults, larvae undergo rapid shifts in metabolism, mobilizing carbohydrate energy reserves to promote synthesis of low-molecular-mass osmoprotectants. In this study, we measured the expression of 11 metabolic genes in response to thermal and dehydration stress. During both heat and cold stress, we observed upregulation of phosphoenolpyruvate carboxykinase (pepck) and glycogen phosphorylase (gp) to support rapid glucose mobilization. In contrast, there was a general downregulation of pathways related to polyol, trehalose, and proline synthesis during both high and low temperature stress. Pepck was likewise upregulated in response to different types of dehydration stress, however, for many of the other genes, expression patterns depended on the nature of dehydration stress. Following fast dehydration, expression patterns were similar to those observed during thermal stress, i.e., upregulation of gp accompanied by downregulation of trehalose and proline synthetic genes. In contrast, gradual, prolonged dehydration (both at a constant temperature and in conjunction with chilling), promoted marked upregulation of genes responsible for trehalose and proline synthesis. On the whole, our data agree with known metabolic adaptations to stress in B. antarctica, 184

although a few discrepancies between gene expression patterns and downstream metabolite contents point to fluxes that are not controlled at the level of transcription.

Introduction

The ability to survive extremes in temperature and water availability is critical for insects living in polar environments, such as the Antarctic midge, Belgica antarctica, which inhabits offshore islands and ice-free areas along the Antarctic Peninsula (Sugg et al., 1983). While the microhabitat of B. antarctica is buffered from extreme variations in temperature (Baust and Lee, 1981), B. antarctica is nonetheless exposed annually to sub- freezing temperatures for more than 8 months (Elnitsky et al., 2008). Additionally, because liquid water is unavailable much of the year (Kennedy, 1993), larvae face a significant risk of dehydration. As such, larvae are extremely tolerant of both cold and dehydration; larvae can survive freezing to -20°C (Lee et al., 2006) and can tolerate water losses of up to 70% (Benoit et al., 2007).

Our recent studies have begun to elucidate the physiological and molecular mechanisms used by B. antarctica to tolerate environmental extremes. Larvae constitutively express both heat shock proteins (Rinehart et al., 2006) and antioxidant enzymes (Lopez-Martinez et al., 2008) in anticipation of adverse conditions. In response to dehydration, larvae rapidly upregulate a number of genes, including genes encoding several chaperone proteins, membrane restructuring enzymes, and structural components of the cytoskeleton (Lopez-Martinez et al., 2009). Recent evidence also demonstrates the

185

importance of aquaporins in mediating dehydration stress (Goto et al., 2011; Yi et al.,

2011).

For animals exposed to adverse environmental conditions, rapid changes in metabolism are essential to maintain energy balance and protect structural components of the cell (e.g. lipids and proteins). The accumulation of low-molecular-mass cryoprotectants, such as glycerol and glucose, has been documented in numerous organisms, including both freeze-tolerant insects and frogs (Storey, 1997). Trehalose has also been implicated as a key protective solute in a range of environmental stress responses, including low temperature (Lee, 2011) and hypoxia (Chen and Haddad, 2004), presumably by stabilizing cell membranes and proteins during periods of osmotic imbalance (Elbein et al., 2003). Trehalose appears to be particularly important during periods of extreme dehydration; for example, most invertebrates that are capable of anhydrobiosis use trehalose as the primary osmoprotectant (Clegg, 2001). Finally, the amino acid proline has been identified as a potent cryoprotectant in insects. Proline is accumulated during cold acclimation (Kostal et al., 2011a), and diet supplementation with proline can substantially enhance the cold tolerance, and even confer freezing tolerance, in two species of drosophilid flies (Kostal et al., 2011b, 2012).

Metabolic adaptations also appear to be essential components of the stress response in B. antarctica. Using metabolomics, Michaud et al. (2008) identified several metabolites, including two osmoprotective polyols, glycerol and erythritol, that are responsive to cold and dehydration. During both freezing and dehydration, larvae significantly deplete glycogen reserves, presumably converting them to glucose and other

186

osmoprotectants (Teets et al., 2011, 2012a). Indeed, glucose levels increase during both recovery from cold stress (Teets et al., 2011) and in response to cryoproective dehydration (Elnitsky et al., 2008), although it is unknown whether glucose serves an osmoprotective role or simply functions as a precursor for other metabolites. In addition, several studies have described the importance of trehalose as an osmoprotectant in B. antarctica, particularly during dehydration stress. In response to dehydration, larvae rapidly accumulate trehalose (Benoit et al., 2007), which in turn seems to facilitate cross- tolerance between dehydration and cold tolerance (Benoit et al., 2009). Larvae also synthesize large amounts of trehalose during cryoprotective dehydration, an overwintering strategy by which larvae remain unfrozen by allowing body fluids to remain at vapor pressure equilibrium with surrounding ice (Elnitsky et al., 2008).

Despite mounting evidence that larvae of B. antarctica undergo significant metabolic adjustments in response to stress, the molecular mechanisms of these biochemical changes have not been investigated. In this study, we measured larval expression of 11 metabolic gene transcripts in response to both thermal and dehydration stress. For the thermal stress experiment, we exposed larvae to heat shock, as well as cold shock in both the frozen and supercooled states. It appears to be beneficial for larvae to be supercooled during periods of sub-zero temperature (Teets et al., 2011), thus we measured whether the type of cold exposure (i.e. freezing or supercooling) influences metabolic gene expression. In the dehydration experiments, we exposed larvae to three types of dehydration regimens: fast dehydration, gradual, slow dehydration, and cryoprotective dehydration. All three resulted in similar levels of water loss, thus

187

enabling us to test whether there is a common molecular response to dehydration, or whether the severity and duration of dehydration treatment impacts gene expression.

Specifically, we isolated genes involved in 1) the synthesis of glucose via glycogenolysis and gluconeogenesis, 2) the synthesis and recycling of polyol cryoprotectants, 3) the synthesis, breakdown, and transport of trehalose, and 4) the synthesis of proline. The biochemical pathways for which we monitored gene expression are illustrated in Fig. 1. We report that expression profiles of metabolic genes are highly labile during periods of stress, and that the magnitude and direction of expression changes are highly dependent on the type and severity of the stress experienced.

Materials and Methods

Experimental animals

Larvae of B. antarctica were collected on various off-shore islands in the vicinity of Palmer Station (64°46’S, 64°04’W) on the Antarctic Peninsula in January and

February, 2011. Larvae were returned to the station and extracted from their natural substrate into ice water using a modified Berlese apparatus. Concentrated samples of larvae were returned to natural substrate and stored at 2°C until used for experiments.

This temperature is near the average air temperature at Palmer Station during January and

February. Prior to beginning an experiment, larvae were sorted from substrate in ice water and held on moist filter paper overnight to standardize body water content. Only fourth instar larvae were used for experiments.

188

Experimental conditions

For each treatment group described below, we collected 5 biological replicates, each replicate containing 20 larvae. The treatment conditions for the thermal stress experiments were as follows: Control (4°C/24 h), heat shock (25°C/24 h), supercooled (-

5°C/24 h in a dry centrifuge tube), supercooled with recovery (-5°C/24 h in a dry centrifuge tube, followed by 12 h at 4°C), frozen (-5°C/24 h in a centrifuge tube with ~50

μl water), and frozen with recovery (-5°C/24h in a centrifuge tube with ~50 μl water, followed by 12 h at 4°C). In the frozen with recovery treatment, water was removed immediately upon thawing.

For the dehydration experiments, treatments were as follows: Control (100%

RH/24 h at 4°C), fast dehydration (75%RH/24h at 4°C), fast dehydration + rehydration

(75%RH/24 h followed by 24 h rehydration on moist paper towel at 4°C), slow dehydration (99% RH/10 d at 4°C), slow dehydration + rehydration (99%RH/10 d followed by 24 h rehydration on moist paper towel at 4°C), and cryoprotective dehydration (gradually chilled from -0.6°C to -3°C at vapor pressure equilibrium with ice, then held at -3°C for 10 d; see Elnitsky et al. (2008)). For these experiments, 75 and

99% RH were generated with saturated solutions of NaCl and K2SO4, respectively

(Winston and Bates, 1960). The % water loss was ~35% for each dehydration treatment, while 24 h of rehydration was sufficient to return water content to the control level (data not shown). For all the above conditions, survival was at or near 100%. Immediately following exposure to the experimental conditions, larvae were frozen at -70°C, and frozen samples were shipped back to Ohio State University on dry ice, where they were held at -70°C until the time of RNA extraction. 189

Bioinformatics

Candidate genes for this study were identified from the following sources: 1) enzymes previously linked to cryoprotectant synthesis (Joanisse and Storey, 1994), 2) genes involved in trehalose synthesis in the desiccation tolerant midge, Polypedilum vanderplanki (Mitsumasu et al., 2010), and 3) genes involved in proline synthesis in cold- selected Drosophila melanogaster lines (Misener et al., 2001). After assembling a candidate list of genes, homologs were identified from a B. antarctica genome and transcriptome that is in preparation. Protein sequences from Aedes aegypti, Anopheles gambiae, and D. melanogaster were obtained for the genes in Fig. 1. Using the tblastn algorithm, we searched the B. antarctica transcriptome for putative homologs. The B. antarctica homologs were reciprocally blasted against the Refseq protein database using blastx to provide further evidence of the transcripts’ putative identity. Finally, the B. antarctica sequences were translated into protein sequences and searched against the

Pfam database (http://pfam.sanger.ac.uk/) to confirm that the predicted protein had the expected domain structure. Accession numbers and results of the bioinformatics analysis are summarized in Table 1.

Isolating fragments for qPCR

PCR primers for the 11 metabolic genes plus two nuclear-encoded ribosomal protein reference genes, rp49 [GenBank: JX462669] and rpl19 [GenBank: JX462670], were manufactured by Integrated DNA Technologies (IDT, Coralville, IA). Using IDT’s primer design software (http://www.idtdna.com), primers were designed to be 24 nt in length, have a product size between 100-180 bp, and an annealing temperature of 60°C.

Primers were tested against B. antarctica cDNA using conventional PCR, and products 190

were visualized on an agarose gel for the presence of a single band at the expected size.

Following PCR, products were purified using the Invitrogen PureLink PCR Purification

Kit (Life Technologies, Grand Island, NY) and sequenced by Sanger sequencing at the

Ohio State Plant Microbe Genomics Facility. Finally, primer linearity and efficiency were measured by running an 8-point standard curve, with purified PCR product as the template. For all primers tested, the R2 value was >0.99 and the efficiencies were between 85.1 and 94.7% (Table 2).

RNA extraction and cDNA synthesis

To measure gene expression in biological samples, we first extracted total RNA from frozen samples, each containing 20 larvae, using the Ambion RiboPure Kit (Life

Technologies, Grand Island, NY). Samples were removed from -70°C one at a time to prevent RNA degradation during processing. RNA quantity and purity were assessed on a

NanoDrop 2000 spectrophotometer (Thermo Fisher Scientific, Waltham, MA) by measuring absorbance at 260, 280, and 230 nm. Starting with 1 μg total RNA for each sample, we synthesized cDNA using the Invitrogen SuperScript VILO cDNA Synthesis

Kit (Life Technologies, Grand Island, NY). The resulting cDNA samples were diluted

10X in water prior to measuring gene expression and stored at -20°C. qPCR

For qPCR analysis, each 20 μl reaction contained 2 μl cDNA template, 2 μl of each primer at 250 nM concentration, 4 μl water, and 10 μl 2X iQ SYBR Green

Supermix (Bio-Rad, Hercules, CA). Reactions were run on a Bio-Rad iCycler iQ Real-

Time PCR Detection System (Bio-Rad, Hercules, CA) with the following parameters: 3

191

min at 94°C followed by 40 cycles of 10 s at 94°C, 30 s at 58°C, and 30 s at 72°C.

Following each reaction, a melt-curve analysis was conducted to verify that only one product was synthesized with no primer dimer.

Data Analysis

Following qPCR reactions, background-corrected fluorescent intensities were obtained from the Bio-Rad analysis software. Background correction, amplitude normalization, and threshold cycles (Ct) for each reaction were calculated according to

Larionov et al. (2005), using a custom MatLab script. Relative gene expression was calculated using the 2-ΔCt method. In short, for each sample, the Ct for the reference gene

(either rp49 or rpl19) was subtracted from the Ct of the gene of interest to obtain the ΔCt value. The relative expression was then calculated using the formula 2-ΔCt, and the mean and SE were calculated for each treatment group (N = 5). Finally, relative fold change

(FC) for each gene was calculated by dividing the mean relative expression value for each treatment group by the mean expression value of the control group. Preliminary experiments determined that rp49 was the most stable reference gene for the temperature series, while rpl19 was more appropriate for the dehydration series.

To compare the relative mRNA abundance in each group, one-way ANOVA with a post-hoc pooled t-test was conducted on the ΔCt values using JMP 9 (SAS Institute

Inc., Cary, NC). To control the false discovery rate (FDR), p-values were adjusted using the Benjamini and Hochberg method (Benjamini and Hochberg, 1995). Statistical significance was set at FDR <0.05. Since the thermal stress and dehydration stress samples were run on different qPCR plates (and were designed as separate experiments),

192

separate hypothesis tests were conducted for the temperature and dehydration samples.

Principal components analysis (PCA) was conducted for each experiment using the R package prcomp. The first two principal component scores for each sample were then plotted to determine whether the treatment groups formed distinct clusters.

Results

Bioinformatics

From the B. antarctica genome and transcriptome, we isolated transcripts for the following genes: glycogen phosphorylase (gp), phosphoenolpyruvate carboxykinase

(pepck), glucose-6-phosphatase (g6pase), aldo-keto reductase (akr), sorbitol dehydrogenase (sordh), glycerol-3-phosphate dehydrogenase (g3pdh), trehalose-6- phosphate synthase (tps), trehalose-6-phosphate phosphatase (t6pp), trehalase (treh), trehalose transporter 1 (tret1), and pyrroline-5-carboxylate reductase (p5cr, Table 1, Fig.

1). With the exception of g6pase, full-length coding sequences were obtained for all transcripts. All 11 sequences showed high similarity to annotated D. melanogaster and mosquito sequences at the amino acid level, with blastx E-values ≤1.00E-45 for all transcripts (Table 1). Furthermore, all translated protein sequences had the same predicted domain configuration as their closest homolog, as determined by a Pfam search

(Table 1).

Gene expression during thermal stress

Using qPCR, we measured expression of the above 11 transcripts in response to both heat and cold. Following both heat and cold stress, there was significant

193

upregulation of two genes involved in glucose synthesis, gp and pepck (Fig. 2). Of all the genes measured, pepck was the most labile, as mRNA levels were 2.22±0.04-fold higher in supercooled individuals relative to control. Also, pepck was the only transcript that differed between supercooled and frozen individuals; while pepck was significantly elevated in frozen individuals relative to control (FC = 1.72±0.06), this level was significantly lower than that of supercooled individuals (ANOVA, pooled-t, FDR<0.05).

During recovery from cold, the transcript levels of gp and pepck remained elevated relative to control. The only other transcript significantly elevated in response to thermal stress was treh, which was slightly elevated in supercooled individuals (Fig. 2).

There were also several transcripts that were downregulated in response to thermal stress. Two genes involved in polyol metabolism, sordh and g3pdh, were significantly downregulated in at least some of the temperature treatments (Fig. 2).

Relative to control, sordh was downregulated in response to heat shock and recovery from supercooling, while g3pdh was significantly downregulated in all treatments, with transcript levels being 80-90% of the control levels. The two genes involved in trehalose synthesis, tps and t6pp, were strongly down-regulated in response to thermal stress. Both tps and t6pp were downregulated with FC ≈ 0.55 in response to both heat and cold (Fig.

2), and remained lower than control values during recovery from cold. Likewise, tret1, a trehalose transport protein, was significantly downregulated during recovery from both supercooling and freezing. Finally, p5cr, a key enzyme in proline synthesis, was significantly downregulated in response to both supercooling and freezing, and remained downregulated during recovery. A simplified summary of expression changes during thermal stress is presented in Fig. 3. 194

Gene expression during dehydration stress

In comparison to thermal stress, changes in gene expression were much more pronounced in response to dehydration stress. During fast dehydration, we observed upregulation of both gp and pepck, with levels returning to control levels during a 24 h recovery period (Fig. 4). In contrast, gp was significantly downregulated during both slow dehydration and cryoprotective dehydration, with levels at ~80% of control.

However, pepck was strongly upregulated in response to both slow dehydration and cryoprotective dehydration, with levels peaking in the cryoprotective dehydration group

(FC = 5.21±0.25). While g6pase levels did not differ from control in all five dehydration treatments, levels were significantly higher during fast dehydration than they were during slow dehydration (ANOVA, pooled-t, FDR<0.05).

The three genes involved in polyol synthesis and breakdown, akr, sordh, and g3pdh, were all unresponsive to fast dehydration (Fig. 4). However, both akr and sordh were significantly downregulated in response to slow dehydration, while akr and g3pdh were upregulated during cryoprotective dehydration. In fact, akr was the only transcript that showed opposite expression patterns in response to slow and cryoprotective dehydration.

With regards to trehalose synthesis genes, we observed opposite results in response to fast versus slow dehydration. During fast dehydration, levels of tps were unchanged while t6pp was downregulated (FC = 0.56±0.04); however, during slow dehydration, both tps and t6pp were significantly upregulated (FC = 1.80±0.17 and

1.69±0.18, respectively). Similarly, tps and t6pp were upregulated during cryoprotective

195

dehydration. Treh, which codes for an enzyme that breaks down trehalose into glucose monomers, was upregulated in response to all three dehydration regimes, and remained elevated during rehydration (Fig. 4). Finally, tret1 was likewise upregulated in response to all three dehydration treatments.

In our gene set, the lone gene involved in proline synthesis, p5cr, was upregulated in response to both slow and cryoprotective dehydration (FC = 1.61±0.07 and 1.51±0.03, respectively). During rehydration following both fast and slow dehydration, p5cr was downregulated relative to controls (Fig. 4). A simplified summary of expression changes during dehydration stress is presented in Fig. 5.

Multivariate stastics

To summarize expression patterns across all genes, we conducted PCA separately on each dataset (i.e. temperature and dehydration) to visualize which treatment groups are similar and distinct from one another. In the temperature series, we see three non- overlapping groups of samples along PC1: a cluster of control samples, a cluster containing heat shock, frozen, frozen + recovery, and supercooled + recovery samples, and a cluster of supercooled samples (Fig. 6A). For the dehydration experiment, a plot of

PC2 versus PC1 shows that the control and fast dehydration + rehydration groups form an overlapping cluster of points (Fig. 6B). However, each of the remaining treatment groups forms a distinct cluster of points, with the cryoprotective dehydration group being the most distant from the control group.

196

Discussion

From a draft of the B. antarctica genome (in preparation), we isolated coding sequences for 11 genes related to energy mobilization and osmoprotectant synthesis.

Using qPCR, we measured their expression in response to both thermal and dehydration stress. Environmental stress led to expression changes in a number of metabolic genes, with the magnitude and direction of change dependent on the type of stress experienced.

On the whole, expression changes were well-correlated with a priori knowledge of the metabolic changes that accompany thermal and dehydration stress.

Gene expression changes in response to thermal stress

In response to both heat and cold stress, we observed upregulation of gp and pepck, suggesting that glucose synthesis is upregulated via both glycogenolysis and gluconeogenesis (Fig. 2,3). This is consistent with our previous study in which both freezing and supercooling resulted in significant glucose mobilization (Teets et al., 2011).

Interestingly, despite the upregulation of gp in both the supercooled and frozen group,

Teets et al. (2011) only observed significant glycogen depletion in frozen larvae.

However, in that study, we only measured glycogen content following recovery from cold, so it is possible that supercooled larvae convert glycogen to glucose at low temperatures and then recycle glucose to glycogen during recovery. With regards to heat shock, our results do not mesh with previous biochemical studies, as Michaud et al.

(2008) observed a significant decrease in glucose following heat shock. This discrepancy could be the result of differences in treatment conditions (Michaud et al. used 30°C for 1 h as their heat shock treatment, vs. the 25°C for 24 h used in this study), or it could be that elevated metabolism at higher temperatures outweighs glucose production, so that 197

net glucose levels decrease even while glycogenolysis and gluconeogenesis increase.

Also, it is possible that biochemical fluxes are not governed by changes at the level of transcription.

Indeed, upregulation of both glycogenolysis and gluconeogenesis appears to be a general feature of thermal stress responses. Glucose synthesis in response to low temperature has been observed in organisms ranging from plants (Sasaki et al., 1996), to insects (Overgaard et al., 2007), to frogs (Costanzo et al., 1993). At the molecular level, upregulation of pepck is a common response to environmental stress; recent studies have demonstrated strong induction of pepck in response to heat (Sorensen et al., 2005), cold

(Teets et al., 2012b), hypoxia (Liu et al., 2006), and oxidative stress (Girardot et al.,

2004). Furthermore, pepck is strongly upregulated during diapause in D. melanogaster

(Baker and Russell, 2009), Sarcophaga crassipalpis (Ragland et al., 2010), and

Rhagoletis pomonella (Ragland et al., 2011), perhaps to upregulate glucose production in advance of adverse conditions. While the importance of glycogenolysis during thermal stress is well-established (Storey and Storey, 1991), to our knowledge this is the first study demonstrating stress-inducible upregulation of gp transcripts. In wood frogs, there is seasonal accumulation of GP protein during hibernation (Kiss et al., 2011), but expression at the mRNA level has not been examined. Thus, while glycogenolysis is regulated at many levels by post-translational and substrate-dependent events (Arrese and

Soulages, 2010), it appears in B. antarctica that glycogenolysis is also under transcriptional control.

198

Somewhat surprisingly, while there was clear upregulation of glucose production during thermal stress, we observed downregulation of genes involved in polyol, trehalose and proline synthesis, despite their known function as cryoprotectants in other organisms.

Larvae of B. antarctica synthesize glycerol in response to freezing (Michaud et al.,

2008), but this was not reflected in the gene expression data. The two genes in our dataset involved in polyol synthesis, akr and g3pdh, were either unresponsive or downregulated during heat and cold exposure (Fig. 2,4). It is worth noting that Michaud et al. used -10°C for their cold treatment rather than the -5°C used in this study, and this could explain the discrepancy. Perhaps even more puzzling, two genes involved in trehalose production, tps and t6pp, were strongly down-regulated in response to thermal stress. Thus, despite trehalose being a potent cryoprotectant (Duman et al., 1991), its synthesis appears to be shut down in response to low temperature, in favor of glucose production. Finally, while proline has not been established as a cryoprotectant in B. antarctica, it is a potent cryoprotectant in drosophilid flies (Kostal et al., 2012; Kostal et al., 2011b). However, once again, p5cr, a key gene in proline synthesis, is downregulated during thermal stress in this species. Thus, despite the clear upregulation of glucose production during thermal stress, it appears that other cryoprotectant synthesis pathways are shut down, at least at the level of gene expression. One possible explanation for these observations would be that cryoprotecant levels are constitutively high in B. antarctica, as are other stress reponses such as heat shock protein expression (Rinehart et al., 2006) and antioxidant defenses (Lopez-Martinez et al., 2009). However, previous metabolite studies do not support this, as baseline levels of glycerol and sorbitol are very low in B. antarctica

199

(Elnitsky et al., 2008), while trehalose levels (Teets et al., 2012a) are on par with those observed in other larval insects (Kostal et al., 2011a; Storey et al., 1981).

One goal of this study was to determine whether any metabolic genes were differentially regulated between supercooled and frozen larvae. With supercooling points around -7°C (Lee et al., 2006), larvae remain unfrozen at -5°C, provided they avoid inoculative freezing from exogenous ice. In a previous study, we found that supercooled larvae fare much better than frozen larvae, with regards to whole animal survival, cell survival, and energy depletion (Teets et al., 2011). Thus, we hypothesized necessary adaptations to low temperature, such as cellular metabolic restructuring, may be inhibited when larvae are in the frozen state. Looking at the PCA analysis (Fig. 6A), it is evident that frozen and supercooled larvae are transcriptionally distinct. The gene most responsible for this separation is pepck, which was the only gene significantly different between supercooled and frozen larvae; expression was 30% higher in supercooled larvae

(Fig. 2). As the rate-limiting enzyme of gluconeogenesis, perhaps elevated expression of pepck in supercooled larvae gives them a “head-start” in glucose production at low temperature, thus permitting better survival down the line. However, this idea is speculative and needs to be verified with enzyme and substrate-level data.

Gene expression changes in response to dehydration stress

As was the case for thermal stress, dehydration stress resulted in a clear upregulation of gluconeogenesis via upregulation of pepck (Fig. 4,5). These results are supported by previous physiological studies of dehydration tolerance in B. antarctica, as glucose is accumulated both in response to repeated dehydration at a constant

200

temperature (Teets et al., 2012a) and in response to cryoprotective dehydration (Elnitsky et al., 2008). However, whereas thermal stress also caused upregulation of gp to support glycogenolysis, this was not always the case during dehydration. During fast dehydration, we observed a modest increase in gp levels, similar to that observed during thermal stress

(Fig. 4,5). In contrast, gp expression was down-regulated during slow dehydration and cryoprotective dehydration, suggesting larvae are not relying heavily on glycogen for glucose production. While slow dehydration does cause some glycogen depletion, the rate of breakdown is much more rapid during fast dehydration (Teets et al., 2012a). In addition to serving as a carbon pool, glycogen also binds 3-5x its weight in water (Gibbs et al., 1997), so perhaps slowing down glycogen breakdown is a means of conserving water during slow, prolonged dehydration.

In addition to their role as cryoprotectants, polyols are also important components in the response of B. antarctica to dehydration. Glycerol levels increase during both fast and slow dehydration (Benoit et al., 2007; Michaud et al., 2008), while sorbitol levels decrease (Michaud et al., 2008). In our dataset, we measured three genes involved in polyol metabolism, akr, sordh, and g3pdh. While none of these genes were responsive to fast dehydration, we observed downregulation of akr and sordh during slow dehydration

(Fig. 4,5). These changes are opposite of what we would expect based on Michaud et al.’s data; thus it appears polyol synthesis during dehydration is not controlled at the transcriptional level. However, in the cryoprotective dehydration group, we did see upregulation of polyol synthesis at the transcript level. Both akr and g3pdh are significantly upregulated, which would seemingly push sugars towards their respective sugar alcohols (Fig. 1). 201

Trehalose, the blood sugar of insects, is accumulated in a wide-range of organisms in response to dehydration (Clegg, 2001). Larvae of B. antarctica synthesize trehalose in response to fast, slow, and cryoprotective dehydration (Benoit et al., 2009;

Benoit et al., 2007; Elnitsky et al., 2008), presumably to help conserve water and protect cellular structures. These observations were well-supported in the gene expression data for the slow dehydration and cryoprotective dehydration groups; in response to both treatments, there was strong upregulation of the two genes that synthesize trehalose from glucose, tps and t6pp (Fig. 4,5). Furthermore, there was upregulation of tret1, a high- affinity trehalose transporter responsible for bringing trehalose into cells (Kikawada et al., 2007), as well as treh, an enzyme that breaks down trehalose into glucose monomers at the target tissue (Mitsumasu et al., 2010). Indeed, the expression patterns we observed for these genes were similar to those observed during dehydration in the anhydrobiotic sleeping midge, P. vanderplanki (Mitsumasu et al., 2010), and the Arctic collembolan,

Megaphorura arctica (Clark et al., 2009). Thus, it appears that upregulation of genes along the trehalose biosynthesis axis is a critical adaption for dehydration-tolerant arthropods.

While the role of proline as a cryoprotectant is well-established (Kostal et al.,

2011b), its potential role as an osmoprotectant during dehydration has not been examined extensively (Benoit, 2010). We observed upregulation of p5cr, the final enzyme required for proline biosynthesis, during both slow and cryoprotective dehydration (Fig. 4,5).

Proline plays an important role during drought stress in plants, although stress-induced accumulation is primarily regulated by expression of pyrroline-5-carboxylate synthetase, an enzyme upstream of P5CR (Verbruggen and Hermans, 2008). Nonetheless, our 202

expression data provide evidence that proline synthesis is a component of the dehydration response in B. antarctica. Additional substrate-level experiments are needed to determine whether proline is accumulated during dehydration or serves as a precursor for other metabolites required for dehydration.

Similar to our supercooled vs. frozen comparison above, a secondary objective of the dehydration experiment was to compare transcriptional signatures in response to fast, slow, and cryoprotective dehydration. While all three treatments resulted in ~35% water loss, the transcriptional responses were clearly distinct (Fig. 6B). Most notably, in several cases, expression patterns were opposite between fast and slow dehydration. For example, as discussed above, gp was upregulated during fast dehydration but downregulated during slow dehydration (Fig. 4,5). Also, whereas there was clear upregulation of both tps and t6pp during slow dehydration, tps was unaffected by fast dehydration while t6pp was downregulated. Larvae do accumulate modest amounts of trehalose during fast dehydration, but amounts are much higher during slow dehydration

(Benoit et al., 2007). While some of this could be due to time differences, our data show that trehalose synthesis is actually shut down at the gene expression level during fast dehydration. Instead, it appears that larvae are relying primarily on glucose mobilization, via both gluconeogenesis and glycogenolysis, during fast dehydration.

Comparison of thermal stress and dehydration stress

While the thermal stress and dehydration stress treatments were conducted and analyzed as separate experiments, we are able to makes some general comparisons between the two. Most notably, fast dehydration yielded expression changes that were

203

more similar to those observed during thermal stress than during other types of dehydration (Fig. 3,5). Fast dehydration and thermal stress both resulted in upregulation of glucose synthesis via gp and pepck, downregulation of trehalose synthesis, and no increase in p5cr. In both cases, larvae were rapidly transferred to stressful conditions (i.e. temperature extremes or low humidity), and it appears that the metabolic demands of acute thermal and water stress are very similar. It would be interesting to investigate whether gradual, prolonged chilling (in the absence of dehydration) leads to a shift towards trehalose and proline synthesis, as was observed during slow and cryoprotective dehydration. On the whole, it was clear that while thermal stress and dehydration stress shared some of the same molecular responses (most notably induction of gluconeogenesis via pepck), the transcriptional responses were largely different. Thus, even though adaptations for cold and dehydration are thought to have a common origin (Ring and

Danks, 1994), our data suggest notable metabolic differences between the two. However, since it is difficult to make thermal and dehydration stress bouts “equally stressful,” this idea warrants further investigation.

Conclusions

Our gene expression results show that larvae of B. antarctica undergo rapid metabolic restructuring during periods of environmental stress. Previous work has illustrated some of the metabolic consequences of changes in temperature and water balance (Baust and Edwards, 1979; Baust and Lee, 1983; Benoit et al., 2009; Benoit et al., 2007; Elnitsky et al., 2008; Michaud et al., 2008; Teets et al., 2011, 2012a), and now we provide molecular mechanisms that govern some of these changes. While gene expression data do not always reflect protein levels or protein activities (Feder and 204

Walser, 2005), for the most part observed changes in metabolic gene expression meshed with previous phenotypic studies. Also, in some cases, such as the central role of pepck gene expression in regulating gluconeogenesis (Hanson and Reshef, 1997) and the transcriptional regulation of trehalose synthesis (Mitsumasu et al., 2010), there is a clear link between mRNA levels and metabolic endpoints. In the case of B. antarctica, our results suggest that coordinated changes in metabolic gene expression are a critical survival mechanism for Antarctica’s southernmost free-living insect.

Acknowledgements

We thank the staff at Palmer Station for their excellent support during our field season. We acknowledge Justin Peyton for his assistance with the Belgica genome data and for providing the Matlab script for qPCR analysis. We also thank Dr. Tom Teets for help preparing Fig. 1 and Kevin Stevenson for technical assistance with qPCR standard curves.

Grants

This work was supported in part by NSF OPP-ANT-0837613 and ANT-0837559.

References

Arrese, E.L., Soulages, J.L., 2010. Insect fat body: Energy, metabolism, and regulation. Annual Review of Entomology, 55, 207-225.

205

Baker, D.A., Russell, S., 2009. Gene expression during Drosophila melanogaster egg development before and after reproductive diapause. BMC Genomics, 10, 242.

Baust, J.G., Edwards, J.S., 1979. Mechanisms of freezing tolerance in an antarctic midge, Belgica antarctica. Physiological Entomology, 4, 1-5.

Baust, J.G., Lee, R.E., 1981. Environmental "homeothermy" in an Antarctic insect. Antarctic Journal of the United States, 15, 170-172.

Baust, J.G., Lee, R.E., 1983. Population differences in antifreeze cryoprotectant accumulation patterns in an Antarctic insect. Oikos, 40, 120-124.

Benoit, J.B., 2010. Water Management by Dormant Insects: Comparisons Between Dehydration Resistance During Summer Aestivation and Winter Diapause, in: Navas, C.A.C.J.E. (Ed.), Aestivation: Molecular and Physiological Aspects, pp. 209-229.

Benoit, J.B., Lopez-Martinez, G., Elnitsky, M.A., Lee, R.E., Denlinger, D.L., 2009. Dehydration-induced cross tolerance of Belgica antarctica larvae to cold and heat is facilitated by trehalose accumulation. Comparative Biochemistry and Physiology a- Molecular & Integrative Physiology, 152, 518-523.

Chen, Q.F., Haddad, G.G., 2004. Role of trehalose phosphate synthase and trehalose during hypoxia: from flies to mammals. Journal of Experimental Biology, 207, 3125- 3129.

Clegg, J.S., 2001. Cryptobiosis - a peculiar state of biological organization. Comparative Biochemistry and Physiology B-Biochemistry & Molecular Biology, 128, 613-624.

Costanzo, J.P., Lee, R.E., Lortz, P.H., 1993. Glucose-concentration regulates freeze tolerance in the wood frog, Rana sylvatica. Journal of Experimental Biology, 181, 245- 255.

206

Duman, J.G., Wu, D.W., Xu, L., Tursman, D., Olsen, T.M., 1991. Adaptations of insects to subzero temperatures. Quarterly Review of Biology, 66, 387-410.

Elbein, A.D., Pan, Y.T., Pastuszak, I., Carroll, D., 2003. New insights on trehalose: a multifunctional molecule. Glycobiology, 13, 17R-27R.

Feder, M.E., Walser, J.C., 2005. The biological limitations of transcriptomics in elucidating stress and stress responses. Journal of Evolutionary Biology, 18, 901-910.

Gibbs, A.G., Chippindale, A.K., Rose, M.R., 1997. Physiological mechanisms of evolved desiccation resistance in Drosophila melanogaster. Journal of Experimental Biology, 200, 1821-1832.

Girardot, F., Monnier, V., Tricoire, H., 2004. Genome wide analysis of common and specific stress responses in adult Drosophila melanogaster. BMC Genomics, 5, 16.

Hanson, R.W., Reshef, L., 1997. Regulation of phosphoenolpyruvate carboxykinase (GTP) gene. Annual Review of Biochemistry, 66, 581-611.

Joanisse, D.R., Storey, K.B., 1994. Enzyme-activity profiles in an overwintering population of freeze-tolerant larvae of the gall fly, Eurosta solidaginis. Journal of Comparative Physiology. B, Biochemical, Systemic, and Environmental Physiology, 164, 247-255.

Kennedy, A.D., 1993. Water as a limiting factor in the Antarctic terrestrial environment - A biogeographical synthesis. Arctic and Alpine Research, 25, 308-315.

Kikawada, T., Saito, A., Kanamori, Y., Nakahara, Y., Iwata, K.I., Tanaka, D., Watanabe, M., Okuda, T., 2007. Trehalose transporter 1, a facilitated and high-capacity trehalose transporter, allows exogenous trehalose uptake into cells. Proceedings of the National Academy of Sciences of the United States of America, 104, 11585-11590.

Kiss, A.J., Muir, T.J., Lee, R.E., Costanzo, J.P., 2011. Seasonal variation in the hepatoproteome of the dehydration- and freeze-tolerant wood frog, Rana sylvatica. International Journal of Molecular Sciences, 12, 8406-8414.

207

Kostal, V., Simek, P., Zahradnickova, H., Cimlova, J., Stetina, T., 2012. Conversion of the chill susceptible fruit fly larva (Drosophila melanogaster) to a freeze tolerant organism. Proceedings of the National Academy of Sciences of the United States of America, 109, 3270-3274.

Larionov, A., Krause, A., Miller, W., 2005. A standard curve based method for relative real time PCR data processing. Bmc Bioinformatics, 6, 62.

Lee, R.E., 2011. A primer on insect cold tolerance, in: Denlinger, D.L., Richard E. Lee, J.r. (Eds.), Low Temperature Biology of Insects. Cambridge University Press, Cambridge, UK, pp. 3-34.

Liu, G.W., Roy, J., Johnson, E.A., 2006. Identification and function of hypoxia-response genes in Drosophila melanogaster. Physiological Genomics, 25, 134-141.

208

Misener, S.R., Chen, C.P., Walker, V.K., 2001. Cold tolerance and proline metabolic gene expression in Drosophila melanogaster. Journal of Insect Physiology, 47, 393-400.

Ring, R.A., Danks, H.V., 1994. Desiccation and cryoprotection - Overlapping adaptations. Cryo-Letters, 15, 181-190.

Sasaki, H., Ichimura, K., Oda, M., 1996. Changes in sugar content during cold acclimation and deacclimation of cabbage seedlings. Annals of Botany, 78, 365-369.

Storey, K.B., 1997. Organic solutes in freezing tolerance. Comparative Biochemistry and Physiology. Part A, Physiology, 117, 319-326.

209

Storey, K.B., Storey, J.M., 1991. Biochemistry of cryoprotectants, in: Lee, R.E., Denlinger, D.L. (Eds.), Insects at Low Temperature. Champman and Hall, New York, pp. 64-93.

Sugg, P., Edwards, J.S., Baust, J., 1983. Phenology and life-history of Belgica antarctica, an Antarctic midge (Diptera, Chironomidae). Ecological Entomology, 8, 105-113.

Teets, N.M., Peyton, J.T., Ragland, G.J., Colinet, H., Renault, D., Hahn, D.A., Denlinger, D.L., 2012b. Uncovering molecular mechanisms of cold tolerance in a temperate flesh fly using a combined transcriptomic and metabolomic approach. Physiological Genomics, 44, 764-777.

Verbruggen, N., Hermans, C., 2008. Proline accumulation in plants: a review. Amino Acids, 35, 753-759.

Winston, P.W., Bates, D.S., 1960. Saturated salt solutions for the control of humidity in biological research. Ecology, 41, 232-237.

Yi, S.-X., Benoit, J.B., Elnitsky, M.A., Kaufmann, N., Brodsky, J.L., Zeidel, M.L., Denlinger, D.L., Lee, R.E., 2011. Function and immuno-localization of aquaporins in the Antarctic midge Belgica antarctica. Journal of Insect Physiology, 57, 1096-1105.

210

Tables

Table 6.1. Bioinformatics analysis of the 11 metabolic genes profiled in this study. Coding sequences isolated from a B. antarctica draft genome were blasted against the RefSeq protein database using the blastx search algorithm. The species with closest homology, based on the expected value (E-val), is included in the table, along with the % positive for that species. The % positive is the percentage of BLAST-aligned amino acids that are either identical or have similar chemical properties. The final column includes the predicted Pfam domain(s) based on a search of the predicted B. antarctica protein sequence against the Pfam database. The Pfam database ID is included in the table; for a full description see http://pfam.sanger.ac.uk/.

Gene GenBank Predicted Function Species with closest E-value % Pfam Accession # homology positive ID(s) gp JX462658 glycogen phosphorylase Anopheles gambiae 0 91 PF00343

phosphoenolpyruvate pepck JX462659 Aedes aegypti 0 85 PF00821 carboxykinase g6pase JX462660 glucose-6-phosphatase Aedes aegypti 1.00E-45 59 PF01569 akr JX462661 aldohyde/ketone reductase Anopheles gambiae 1.00E-129 76 PF00248

PF08240, sordh JX462662 sorbitol dehydrogenase Drosophila yakuba 0 86 PF00107

glycerol-3-phosphate PF01210, g3pdh JX462663 Aedes aegypti 0 93 dehydrogenase PF07479

PF00982, tps JX462664 trehalose-6-phosphate synthase Anopheles gambiae 0 87 PF02358

trehalose-6-phosphate t6pp JX462665 Culex quinquefasciatus 9.00E-93 72 PF02358 phosphatase treh JX462666 trehalase Aedes aegypti 0 76 PF01204 tret1 JX462667 trehalose transporter Aedes aegypti 0 92 PF00083 p5cr JX462668 pyrroline-5-carboxylate reductase Aedes aegypti 2.00E-113 85 PF03807

211

Table 6.2. Primers used for qPCR gene expression assays. The R2 and efficiency values were calculated from an 8-point, 10-fold dilution series of purified PCR product.

Gene R2 Efficiency Primers Tm (°C) Product Size (%) F: 5'-TGGATCGTAACTTGGCCGAGAACA-3' 60.2 gp 0.999 88.1 125 R: 5'-AACGATATCGGCGAGTGATGCAGA-3' 60.2 F: 5'-AAATGCCTGCACTCAGTTGGAACC-3' 60.2 pepck 0.998 90.6 105 R: 5'-GCTCAGTGCTGGTTTGTGCAAGAT-3' 60.3 F: 5'-AGTGCAGCTGACTGAGAAGTCGAA-3' 60.1 g6pase 0.999 85.1 129 R: 5'-TTGAAAGCCAGTTGAACAGACGCC-3' 60.1 F: 5'-GCCAACAACATTCTGATCACCGCA-3' 60.3 akr 0.999 88.2 120 R: 5'-AACGATGACCGAGTTCTCCAGCAA-3' 60.4 F: 5'-TATCGTCGCGAAGCTCGGAAAGAA-3' 60.2 sordh 0.998 94.7 147 R: 5'-GGTGTCGCACAGAAAGCCATTTCA-3' 60.2 F: 5'-ATACTTGCCCGGACACAAATTGCC-3' 60.2 g3pdh 0.999 91.1 122 R: 5'-CGCCCAAGCCTTTGATGAACTGAT-3' 60.0 F: 5'-GACTTTGCCGCTGGAAACCAAGAA-3' 60.2 tps 0.999 86.3 148 R: 5'-CAAACCGTGATTGCCGGCATAAGT-3' 60.3 F: 5'-TCGCACAACTTTGGCGAAGAATGG-3' 60.3 t6pp 0.998 94.5 149 R: 5'-GTTGCTTTGGTCGGCAGATTTGGA-3' 60.3 F: 5'-GTTGCAATCAGGCGAACAATGGGA-3' 60.3 treh 0.992 89.5 118 R: 5'-CCATTCTTGTGCAACAGCCTTCGT-3' 60.2 F: 5'-TGCTGATCCCTGAAACACCGAGAT-3' 60.2 tret1 0.997 88 151 R: 5'-GCATGTCGTTCAGCTTCGCAATGA-3' 60.3 F: 5'-ACAAATGATTGCGAGTGCCCATCC-3' 60.2 p5cr 0.999 91.7 142 R: 5'-AACGACGTTTGGCTTCACACACAC-3' 60.3 F: 5'-TGGCAGTTCGACCAGCATTCAAAC-3' 60.2 rp49 0.996 90.6 142 R: 5'-AAGCGACGTCTGACTCTGTTGTCA-3' 60.1 F: 5'-ACATCCACAAGCGTAAGGCTGAGA-3' 60.3 rpl19 0.999 87 128 R: 5'-TTCTTGTTTCTTGGTGGCGATGCG-3' 60.1

212

Figures

Figure 6.1. Pathway diagram illustrating the biochemical reactions catalyzed by the genes examined in this study. For clarity, only the names of enzymes for which we measured expression are included in the diagram. Consecutive arrows indicate intermediate reactions that are not depicted in the diagram. Akr, aldo-keto reductase; G3pdh, glycerol-3-phosphate dehydrogenase; G6pase, glucose-6-phosphatase; GP, glycogen phosphorylase; P5cr, pyrroline-5-carboxylate reductase; Pepck, phosphoenolpyruvate carboxykinase; Sordh, sorbitol dehydrogenase, T66p, trehalose-6- phosphate phosphatase; Tps, trehalose-6-phosphate synthase; Treh, trehalase; Tret1, trehalose transporter 1.

213

Figure 6.2. Gene expression changes during thermal stress. Bars represent mean±SE fold change of each group relative to control. Different letters represent significant differences in mRNA abundance for a particular gene (ANOVA, pooled-T, FDR<0.05). HS, heat shock; SC, supercooled; SC+R, supercooled with recovery; F, frozen; F+R, frozen with recovery. See Materials and Methods for full description of treatment conditions.

214

Figure 6.3. Summary of gene expression patterns during thermal stress. Each square represents the relative fold change of a particular transcript relative to control. Expression changes that are not significant (N.S.) are denoted with gray coloration, while the color scale for significant fold changes is included in the figure. HS, heat shock; SC, supercooled; SC+R, supercooled with recovery; F, frozen; F+R, frozen with recovery. See Materials and Methods for full description of treatment conditions.

215

Figure 6.4. Gene expression changes during dehydration stress. Bars represent mean±SE fold change of each group relative to control. Different letters represent significant differences in mRNA abundance for a particular gene (ANOVA, pooled-T, FDR<0.05). FD, fast dehydration; FD+R, fast dehydration + rehydration; SD, slow dehydration; SD+R, slow dehydration + rehydration; CD, cryoprotective dehydration. See Materials and Methods for full description of treatment conditions.

216

Figure 6.5. Summary of gene expression patterns during dehydration stress. Each square represents the relative fold change of a particular transcript relative to control. Expression changes that are not significant (N.S.) are denoted with gray coloration, while the color scale for significant fold changes is included in the figure. FD, fast dehydration; FD+R, fast dehydration + rehydration; SD, slow dehydration; SD+R, slow dehydration + rehydration; CD, cryoprotective dehydration. See Materials and Methods for full description of treatment conditions.

217

Figure 6.6. Principal components analysis of thermal stress (A) and dehydration stress (B) gene expression datasets. The principal components were calculated on 2-ΔCt values for each transcript in each sample (see Methods for description of 2-ΔCt calculations). Each individual sample is represented on the graph, with samples from the same treatment group having the same symbol. F, frozen; F+R, frozen with recovery; HS, heat shock; SC, supercooled; SC+R, supercooled with recovery; CD, cryoprotective dehydration; FD, fast dehydration; FD+R, fast dehydration + rehydration; SD, slow dehydration; SD+R, slow dehydration + rehydration. See Materials and Methods for full description of treatment conditions.

218

Chapter 7: Gene Expression Changes Governing Extreme Dehydration Tolerance in an Antarctic Insect

Abstract

Among terrestrial organisms, arthropods are especially susceptible to dehydration, given their small body size and high surface area to volume ratio. This challenge is particularly acute for polar arthropods, that face near-constant desiccating conditions, as water is frozen and thus unavailable for much of the year. The molecular mechanisms that govern extreme dehydration tolerance in insects remain largely undefined. In this study, we used RNA-sequencing to quantify transcriptional mechanisms of extreme dehydration tolerance in the Antarctic midge, Belgica antarctica, the world’s southernmost insect and only insect endemic to Antarctica. Larvae of B. antarctica are remarkably tolerant of dehydration, surviving losses up to 70% of their body water. Gene expression changes in response to dehydration indicated upregulation of cellular recycling pathways including the ubiquitin-mediated proteasome and autophagy, with concurrent downregulation of genes involved in general metabolism and ATP production.

Metabolomics results revealed shifts in metabolite pools that correlated closely with changes in gene expression, indicating that coordinated changes in gene expression and metabolism are a critical component of the dehydration response. Finally, using comparative genomics, we compared our gene expression results with a transcriptomic dataset for the Arctic collembolan, Megaphorura arctica. Although B. antarctica and M.

219

arctica are adapted to similar environments, our analysis indicated very little overlap in expression profiles between these two arthropods. While several orthologous genes showed similar expression patterns, transcriptional changes were largely species-specific, indicating these polar arthropods have developed distinct transcriptional mechanisms to cope with similar desiccating conditions.

Introduction

For organisms living in arid environments, mechanisms to maintain water balance and cope with dehydration stress are an essential physiological adaptation. Insects, in particular, are at high risk of dehydration due to their small body size and consequent high surface area to volume ratio (Gibbs et al., 1997). Physiological mechanisms for maintaining water balance in insects include adaptations to reduce cuticular water permeability (Gibbs, 1998) and mechanisms to reduce respiratory water loss (Chown,

2002). When water balance cannot be maintained, insects evoke a suite of molecular mechanisms to cope with cellular osmotic stress. For example, during periods of dehydration, heat shock proteins are upregulated to minimize protein damage (Benoit et al., 2010), while aquaporins mediate water movement between cellular compartments

(Liu et al., 2011). However, we have a limited knowledge of the large scale molecular changes prompted by water loss.

Among terrestrial biomes, polar environments are particularly challenging from a water balance perspective, as water is frozen and therefore unavailable for much of the year (Kennedy, 1993). Polar arthropods are typically extremely tolerant of desiccation,

220

with some being able to survive near-anhydrobiotic conditions (Worland et al., 1998).

One such dehydration-tolerant polar arthropod is the Antarctic midge, Belgica antarctica,

Antarctica’s only endemic insect and the southernmost free-living insect. Larvae of B. antarctica are one of the most dehydration-tolerant insects known, surviving a 70% loss of water under ecologically relevant conditions (Hayward et al., 2007). In this species, the ability to tolerate dehydration is an important adaptation for successful overwintering.

The loss of water enhances acute freezing tolerance (Hayward et al., 2007). In addition, overwintering midge larvae are capable of undergoing another distinct form of dehydration, known as cryoprotective dehydration (Elnitsky et al., 2008). During cryoprotective dehydration, a gradual decrease in temperature in the presence of environmental ice creates a vapor pressure gradient that draws water out of the body, thereby depressing the body fluid melting point and allowing larvae to remain unfrozen at subzero temperatures (Holmstrup et al., 2002).

In this study, we used next-generation RNA-sequencing to quantify genome-wide mRNA changes in response to both dehydration at a constant temperature and cryoprotective dehydration. While our recent work on B. antarctica has revealed several key molecular mechanisms of dehydration tolerance, including expression of heat shock proteins (Lopez-Martinez et al., 2009), aquaporins (Goto et al., 2011; Yi et al., 2011), and metabolic genes (Teets et al., 2012a), we lack a comprehensive understanding of the genes and pathways involved in extreme dehydration tolerance. To date, only three studies have examined large-scale transcriptional changes in response to dehydration in insects, all of which were conducted on tropical species. Cornette et al. (2010) identified genes associated with anhydrobiosis in the African sleeping midge, Polypedilum 221

vanderplanki using a semi-quantitative EST approach, while Wang et al. (2010) and

Matzkin et al. (2009) used microarrays to examine genome-wide transcriptional changes following dehydration in Anopheles gambiae and Drosophila mojavensis, respectively. In addition to the insect studies, transcriptional responses to desiccation have been reported for an Arctic arthropod closely related to insects, the springtail (Collembola)

Megaphorura arctica (Clark et al., 2009), as well as a widely-distributed collembolan,

Folsomia candida (Timmermans et al., 2009). Here, in response to dehydration, we report upregulation of recycling pathways such as the proteasome and autophagy with a concurrent shutdown of central metabolism. Complementary metabolomics experiments supported a number of our transcriptome observations, indicating a strong correlation between gene expression and metabolic end products during dehydration. Using comparative genomics, we also compared the molecular response to dehydration in the

Antarctic species B. antarctica with that of the Arctic arthropod M. arctica (Clark et al.,

2009).

Materials and Methods

Animals

Larvae of B. antarctica were collected on offshore islands near Palmer Station

(64°46’S, 64°04’W) in January, 2010, and shipped to Ohio State University. Prior to an experiment, fourth instar larvae were handpicked from substrate in ice water and left at

4°C overnight on moist filter paper to standardize body water content.

222

Experimental Conditions

For these experiments, larvae were exposed to the following conditions: control

(“C”, held at 100% RH at4°C), desiccation (“D”, exposed to 93% RH for 5 d at 4°C), and cryoprotective dehydration (“CD”, temperature gradually lowered from -0.6°C to -3°C over 5 d in the presence of environmental ice, then held at -3°C for 10 d). During cryoprotective dehydration, larvae lose water through the cuticle to the surrounding ice and remain unfrozen by decreasing the hemolymph melting point to match the temperature of the surrounding ice (Elnitsky et al., 2008). Both the desiccation and cryoprotective dehydration treatments resulted in approximately 40% water loss, with survival near 100%. Immediately after treatment, larvae were frozen at -70°C, where they were held until RNA and metabolite extractions. Each treatment consisted of 3 biological replicates, with each replicate containing 20 larvae.

RNA extraction and library preparation

Total RNA was extracted from larvae using Trizol reagent (Life Technologies,

Grand Island, NY) according to the manufacturer’s protocol. RNA quantity and purity was assessed on a NanoDrop 2000 spectrophotometer (Thermo Fisher Scientific,

Waltham, MA), and integrity was measured on an Agilent Bioanalyzer 2100 (Agilent

Technologies, Santa Clara, CA). To generate RNA-seq libraries, we used the Illumina

TruSeq RNA Sample Preparation kit (Illumina, San Diego, CA) according to the manufacturer’s protocol. In short, mRNA was purified from 2 µg total RNA from each sample, fragmented, and converted to double-stranded cDNA. Sequencing barcodes were ligated to the cDNA fragments, and the resulting fragments were amplified using PCR.

223

Libraries were validated on an Agilent Bioanalyzer 2100 to ensure the libraries had the expected fragment size of ~300 bp.

Sequencing

Libraries were quantified using qPCR and sequenced at the Ohio Agricultural

Research and Development Center Molecular and Cellular Imaging Center. Sequencing libraries were multiplexed into groups of three (so that each multiplexed library contained one library from each of the three treatment groups) and sequenced on an Illumina

Genome Analyzer II. For each sample, we obtained between 1.2 and 11.7 million 76-bp reads (Table 1).

Mapping and counting reads

Reads were mapped to B. antarctica genomic contigs (in preparation) using

Bowtie and TopHat (Trapnell et al., 2009), a short read aligner that is capable or predicting exon-exon splice junctions. After mapping, alignment files were processed using SAMtools (Li et al., 2009), and counts were generated with HTSeq, a Python package for high-throughput sequencing analysis. Using HTSeq, we counted the total number of sequencing reads that aligned to each putative gene model in the draft B. antarctica genome. Our draft genome contains ~13,500 gene models that were derived from a combination of RNA-seq reads, BLAST hits, and ab initio gene prediction software using MAKER (Cantarel et al., 2008). Of these, ~11,500 had enough reads align to them to allow estimation of differential gene expression. A relatively high percentage

(>76% for all samples) of reads aligned to gene models, suggesting a good representation of the transcriptome. Using blastx (E-value cutoff of 1E-4), we compared our gene

224

models with annotated protein sequences from Aedes aegypti and Drosophila melanogaster to determine putative functions, and GO terms were assigned to each gene model using Blast2GO (Conesa et al., 2005). Full details about genome assembly and annotation will be provided in our genome report, which is in preparation.

RNA-seq data analysis

To determine which genes were DE, we used the R package DESeq (Anders and

Huber, 2010). In short, DESeq normalizes counts so that library size is equivalent for each sample, estimates a variance function, and tests for expression difference between two treatment conditions using a negative binomial distribution. We ran DESeq for each pairwise comparison of treatments (i.e. C v. D, C v. CD, and D v. CD). For clarity, throughout this manuscript, a fold change >1 for comparison “X v. Y” indicates higher expression in group Y relative to X, while fold change <1 indicates lower expression in group Y relative to X. For hierarchical clustering of the phenotypic classes, we obtained variance stabilized data from DESeq, calculated a matrix of distances, and used the R package hclust for clustering.

After identifying DE genes, enriched GO terms were determined using the R package GOseq (Young et al., 2010), which accounts for transcript length-bias associated with RNA-seq data. We separately tested for enriched GO terms in genes that were up- and downregulated, to identify which categories of genes were induced and which were repressed by a particular treatment. After enrichment testing, p-values were corrected using the Benjamini and Hochberg method (Benjamini and Hochberg, 1995) to control the false discovery rate. We restricted the output to GO terms with ontology “Biological

225

Process” to limit redundancy. Additionally, we tested for enriched KEGG pathways

(Kanehisa and Goto, 2000) using the R package GSA (Efron and Tibshirani, 2007).

Unlike traditional overrepresentation analysis, GSA uses the actual expression values for each gene in determining its enrichment score. For GSA, we mapped our gene models to the A. aegypti proteome and tested the entire set of A. aegypti KEGG pathways for enrichment. The input for GSA was a matrix of normalized counts for the B. antarctica gene models that had a significant (E-value<1E-4) BLAST hit against A. aegypti. In cases where two or more gene models mapped to the same A. aegypti protein, only the best

BLAST match was retained.

Comparative genomics of dehydration response

Using microarrays, Clark et al. (2009) identified ESTs responsive to desiccation and cryoproective dehydration in the arctic collembolan, M. arctica. We restricted our comparison to the two treatments in (Clark et al., 2009) that were analgous to our desiccation and cryoprotective dehydration treatments, the treatments named”0.9 salt” and “-2°C,” respectively. For simplicity, these treatments will also be referred to as desiccation and cryoprotective dehydration.

Putative orthologs between B. antarctica and M. arctica were determined by conducting reciprocal blast (algorithm tblastx) of our gene models against the M. arctica

ESTs found on the microarray. The M. arctica microarray data was obtained from

ArrayExpress (accession # E-MEXP-2105) and analyzed using the R package limma according to the parameters outlined in (Clark et al., 2009). Finally, using the R package

VennDiagram, we calculated the degree of overlap between orthologous up- and

226

downregulated genes among the four species/treatment combinations. Additionally, to determine the overall similarity in gene expression between groups, we conducted hierarchical clustering on the samples, restricting the analysis to orthologous transcripts.

Hierarchical clustering was conducted on the log fold change values for each transcript from each individual sample using JMP 9 (SAS Institute, Cary, NC). For our dataset, we calculated the log fold change of each transcript relative to the mean expression value of the control group. For the M. arctica data, log fold changes for each EST were obtained from the limma pipeline following between array normalization.

qPCR validation

To validate results from the RNA-seq analysis, we conducted qPCR on a subset of

13 genes. We selected genes from several functional categories of interest (i.e. heat shock proteins, detoxification enzymes, regulators of cell death, and structural components of the cuticle and cystoskeleton), including a mix of genes that were up- and downregulated by our treatments. Primers were designed using IDT’s primer design software

(www.idtdna.com) with the following parameters: length of 24 nt, melting temperature of

60°C, and product size of 100-180 bp. Primers were tested using conventional PCR and gel electrophoresis for a product of the correct size, and standard curves were conducted on a 10-fold dilution series of PCR products. The primer sequences and standard curves are presented in Table 2.

cDNA for qPCR was generated from aliquots of the same RNA samples used for

RNA-seq, thus allowing a direct correlation between RNA-seq and qPCR results. Total

RNA was further purified using the Ambion RiboPure kit (Life Technologies, Grand

227

Island, NY), and cDNA was generated with the Invitrogen SuperScript VILO cDNA

Synthesis Kit. The resulting cDNA samples were diluted 10X prior to analysis and stored at -20°C. Each qPCR reaction consisted of 2 µl cDNA, 2 µl of each primer at 250 nM concentration, 4 µl water, and 10 µl 2X iQ SYBR Green Supermix (Bio-Rad, Hercules,

CA). Reactions were carried out on a Bio-Rad iCycler iQ Real-Time PCR Detection

System, with the following temperature protocol: 94°C for 3 min, followed by 40 cycles of 94°C for 10 s, 58°C for 30 s, and 72°C for 30 s. After each run, a melt-curve was generated to verify that only one product was present in the reaction. Baseline correction, amplitude normalization, and threshold cycle (Ct) calculations were conducted according to Larionov et al. (2005) with a custom MatLab script. Relative gene expression was calculated using the 2-ΔCt method, with rpl19 serving as the reference gene. To convert to fold change, the mean 2-ΔCt value for each treatment group was divided by the mean value for the control.

Metabolomics

Since a large number of metabolic genes were differentially regulated in our treatments, we also conducted a metabolomics analysis of the same treatment conditions.

Groups of 15 larvae were homogenized in 600 µl of 2:1 methanol-chloroform, 400 µl water was added for phase separation, and 180 µl of the upper aqueous phase was vacuum dried. The extract was resuspended in 30 µl of 20 mg mL-1 methoxyaminehydrochloride in pyridine and heated for 60 min at 40°C while shaking.

Subsequently, 30 µl of MSTFA was added, and the sample was heated for an additional

60 min at 40°C. All derivatization steps were conducted with a CTC CombiPal

228

autosampler (GERSTEL GmbH & Co.KG, Mülheim an der Ruhr, Germany) to ensure uniformity of samples.

After derivatization, samples were run on a Trace GC Ultra chromatograph coupled to a Trace DSQII quadrupole mass spectrometer (Thermo Fischer Scientific Inc,

Waltham, MA, USA). Oven conditions were as follows: 70 to 170 °C at 5 °C.min-1, from

170 to 280 °C at 7 °C.min-1, from 280 to 320 °C at 15 °C.min-1, and then the oven remained 4 min at 320 °C. Spectra were screened for 60 pure reference compounds in a custom database, and quantification was accomplished by comparing samples to a 10 point standard curve of pure analyte. Data were analyzed by conducting an ANOVA followed by a pooled-t test for each compound in JMP 9. P-values were corrected using the Benjamini-Hochberg method (Benjamini and Hochberg, 1995).

Data deposition

Raw sequencing reads are available in the NCBI Short Read Archive, accession number SRA058518, while the genomic contigs are available under NCBI BioProject

PRJNA172148. Accession numbers for predicted transcripts in this study are deposited in the NCBI Transcriptome Shotgun Assembly database, accession number

GAAK01000000.

Results and Discussion

The Antarctic midge, B. antarctica, is one of the most dehydration-tolerant insects that has been characterized. In this study, we used RNA-seq to measure gene expression

229

levels in response to the following treatments that hereafter we refer to as control, desiccation, and cryoprotective dehydration:

- Control – Held at 4°C and 100% RH; fully-hydrated.

- Desiccation – Desiccation at a constant temperature of 4°C and 93% RH for 5 d,

resulting in ~40% water loss.

- Cryoprotective dehydration – Gradually chilled over 5 d from -0.6 to -3°C at

vapor pressure equilibrium with surrounding ice, then held at -3°C for 10 d

(Elnitsky et al., 2008). Also yielded ~40% water loss.

Both dehydration treatments resulted in substantial changes in gene expression. Of the ~11,500 gene models that had enough reads to support estimation of differential expression, 3,275 and 2,365 were differentially expressed during desiccation and cryoprotective dehydration, respectively (Fig. 1A). Hierarchical clustering analysis indicated that the desiccation and cryoprotective dehydration treatments yielded distinct transcriptional signatures (Fig. 1B). However, a majority of the differentially expressed genes were shared between the two treatments (Fig. 1C), and downstream analyses revealed that many enriched pathways were identical. Thus, for clarity, we will primarily discuss the results of the desiccation treatment, while specific results from the cryoprotective dehydration treatment can be found in Tables 3 and 4. Additionally, a direct comparison of the desiccation and cryoprotective dehydration treatments, highlighting the expression differences between these two conditions, is provided in

Table 5. However, it is worth mentioning that time differences between the two dehydration treatments (5 d for desiccation, 15 d for cryoprotective dehydration) may

230

also contribute to differences between these treatments. To validate our expression results, we used qPCR to measure expression of 13 genes in the same RNA samples used for RNA-seq. Overall, there was excellent agreement between the RNA-seq results and qPCR results (Fig. 2).

Functional categories of differentially expressed genes

To place these large-scale changes in gene expression into a meaningful context, we identified enriched functional categories using GO enrichment analysis (Table 6) and enriched KEGG pathways using gene set analysis (GSA; Table 7). To distinguish between functional categories of genes that are turned on and off in response to desiccation, we separated the GO enrichment analysis into lists of up- and downregulated genes.

Functional categories upregulated during desiccation

In response to desiccation, we observed enrichment of several functional terms, notably terms related to stress response, ubiquitin-dependent proteasome, actin organization, and signal transduction, specifically several GTPase enzymes involved in membrane trafficking (Table 6A). The GO term “response to heat” was enriched in the upregulated genes, and this category primarily encompasses the heat shock proteins

(hsps), cellular chaperones that repair misfolded proteins in response to various environmental stressors (Feder and Hofmann, 1999), including heat, cold (Teets et al.,

2012b), oxidative damage (Girardot et al., 2004), and dehydration (Benoit et al., 2010;

Lopez-Martinez et al., 2009). Our group has demonstrated the importance of hsps in B. antarctica stress tolerance (Lopez-Martinez et al., 2009; Rinehart et al., 2006), but

231

previous studies were limited to a few hsp genes obtained by targeted approaches. Here, we report upregulation of numerous putative hsps, including members of the small heat shock protein (3 members), hsp40 (2 members), hsp70 (8 members), and hsp90 (1 member) families. We also observed ~1.8-fold upregulation of hsf, the transcription factor that regulates hsp expression (Morimoto, 1998). In addition to chaperone activity, hsps target damaged proteins to the proteasome to prevent accumulation of dysfunctional proteins and to recycle peptides and amino acids (Goldberg, 2003). Indeed, we detected enrichment of GO terms related to ubiquitin-dependent proteolysis (Table 6A) in the desiccation-upregulated genes. Our results indicate coordinated upregulation of hsps and proteasomal genes, which cooperatively function to repair and degrade damaged proteins during dehydration.

In our GSA, we observed positive enrichment of the KEGG pathway “Regulation of autophagy” during desiccation (Table 7). Autophagy is a catabolic process in which parts of the cytoplasm and organelles are sequestered into vesicles and digested in lysosomes (Maiuri et al., 2007), thereby conserving cellular macromolecules and energy during periods of stress and nutrient deprivation. While autophagy can be an alternative means of programmed cell death, during times of stress, autophagy can reduce the amount of cell death by recycling cellular components and inhibiting apoptotic cell death

(Maiuri et al., 2007). We hypothesize that during dehydration, the level of autophagy increases, which conserves energy and promotes survival during prolonged periods of cellular stress.

232

We identified 92 homologs of genes with known function in autophagy and programmed cell death that were differentially expressed during desiccation and/or cryoprotective dehydration. Several lines of evidence support the hypothesis that dehydration promotes autophagy while concurrently inhibiting apoptosis (Fig. 3A). This evidence includes: 1) 11-fold upregulation of sestrin during desiccation. Sestrins are highly conserved genes that have an antioxidant function and promote longevity by inhibiting apoptosis and increasing autophagy via inhibition of TOR signaling (Lee et al.,

2010). 2) Significant upregulation of six authophagy-related signaling genes (atg1, atg6, atg8, atg9, atg13 and atg18) that carry out the essential cellular functions of auophagy

(He and Klionsky, 2009). 3) Upregulation of four transcription factors, eip74EF, eip75EF, cabut and maf-S, that are positive regulators of autophagy in D. melanogaster

(Gorski et al., 2003). 4) 3-fold upregulation of thread, a potent inhibitor of apoptotic cell death that prevents activity of pro-apoptotic caspases (Lisi et al., 2000). 5) Upregulation of proteasomal genes (see above), suggesting cross-talk and cooperation between these distinct cellular recycling pathways (Korolchuk et al., 2010). We suspect that the autophagy pathway serves an important protective function by limiting cell death and turnover of macromolecules during dehydration, especially during the long Antarctic winter.

Functional categories downregulated during dehydration

Upregulation of cellular recycling pathways, such as ubiquitin-mediated proteasome and autophagy, likely serves to conserve energy during prolonged dehydration. Consistent with this idea, we observed downregulation of genes related to general metabolism and ATP production (Table 6B, Fig. 3B). Larvae of B. antarctica 233

significantly depress oxygen consumption rates in response to dehydration (Benoit et al.,

2007). This is a common adaptation in dehydration-tolerant insects, presumably to minimize respiratory water loss and to minimize the loss of water bound to glycogen and other carbohydrates (Marron et al., 2003). This dehydration-mediated metabolic shutdown is strongly supported by gene expression data, as nearly 25% of all metabolic genes in our dataset were downregulated in response to desiccation (Table 6B). We noted a general shutdown of carbohydrate catabolism and ATP generation; nearly every gene involved in glycolysis, TCA cycle, and ATP synthesis is downregulated (Fig. 3B).

Furthermore, among our downregulated genes, we observed enrichment of genes related to protein, lipid, and chitin metabolism, as well as energetically expensive processes such as membrane transport, including proton, cation, carbohydrate, and amino acid transport.

A decrease in metabolic activity was further supported by our GSA results; nearly every negatively-enriched KEGG pathway (i.e. pathways in which genes tended to be downregulated) was related to metabolism, including several pathways related to carbohydrate and amino acid metabolism (Table 7). Thus, taken together, both GO enrichment analysis and GSA analysis of KEGG pathways revealed a coordinated shutdown of metabolic activity at the transcript level. We hypothesize that these mechanisms may be particularly important for overwintering larvae, contributing to energy conservation during the long Antarctic winter.

Dehydration-induced changes in the metabolome

To determine whether the above changes in metabolic gene expression correlated with changes in metabolic endpoints, we conducted a follow-up metabolomics experiment with the same treatment conditions. Using targeted GC-MS metabolomics, 234

we measured levels of 36 compounds in response to desiccation and cryoprotective dehydration. As with gene expression, desiccation and cryoprotective dehydration had a major impact on the metabolome, as the concentrations of 32 out of the 36 compounds significantly changed in at least one treatment (Fig. 4). While the metabolic changes induced by desiccation and cryoprotective dehydration were largely similar, our treatment groups were distinct from one another, as determined by hierarchical clustering

(Fig. 5).

We observed several distinct metabolic responses to desiccation, and these were generally supported by gene expression data. We noted the following: 1) Decreased levels of the glycolytic intermediates glucose-6-phosphate and fructose-6-phosphate, which reflected downregulation of glycolysis genes (Fig. 3B). Hexokinase and glucose-6- phosphate isomerase, the enzymes that synthesize glucose-6-phosphate and fructose-6- phosphate, were both significantly downregulated (>1.5-fold). Additionally, we observed decreased levels of lactate, the endpoint of anaerobic respiration through glycolysis. 2)

Accumulation of citrate, which is evidence of decreased flux through the TCA cycle, was supported by downregulation of a number of TCA cycle genes (Fig. 3B). An alternative explanation for accumulation of citrate would be increased oxidation of fatty acids, but this hypothesis is not supported by the gene expression data, as a majority of fatty acid metabolism genes were downregulated (Table 6,7). 3) Increase in proline levels from 7.8 to 21.1 nmol mg-1 dry mass in response to desiccation, which was supported by 1.5-fold upregulation of pyrroline-5-carboxylate reductase, the terminal enzyme of proline synthesis. Additionally, we observed 1.3-fold upregulation of glutamate synthase and concurrent accumulation of glutamate, a precursor of proline, from 12.6 to 29.9 nmol mg- 235

1 dry mass (Fig. 4). While proline is a potent cryoprotectant in insects (Kostal et al., 2011) and confers desiccation tolerance in plants (Verbruggen and Hermans, 2008), this is the first time proline has been linked to dehydration in insects. 4) Accumulation of several osmoprotective polyols, of which the quantities of sorbitol (increase from 0.5 to 4.3 nmol mg-1 dry mass) and mannitol (increase from 5.0 to 155.1 nmol mg-1 dry mass) exhibited the most dramatic changes. Additionally, fructose, a precursor for both mannitol and sorbitol, increased from 1.3 to 33.4 nmol mg-1 dry mass. While the genes involved in mannitol and sorbitol synthesis are poorly defined in insects, we did observe 4.6-fold upregulation of phosphoenolpyruvate carboxykinase, the rate limiting step of gluconeogenesis (Hanson and Reshef, 1997). Upregulation of this gene leads to increased glucose production via gluconeogenesis, with glucose serving as a central precursor for the synthesis of most sugar alcohols. Interestingly, we did not observe accumulation of glucose during dehydration (Fig. 4), suggesting glucose is being shunted to other pathways as soon as it is produced. On the whole, there was good agreement between gene expression and metabolomics data. However, some metabolite changes could not be correlated with changes at the transcript level, suggesting post-transcriptional levels of control. Also, in some instances changes in gene expression may alter rates of metabolic flux, which are not captured in these types of metabolomics analyses.

Comparative genomics of molecular response to dehydration

The transcriptomic response to dehydration has been studied in three other insects, the African sleeping midge P. vanderplanki (Cornette et al., 2010), the mosquito

A. gambiae (Wang et al., 2010), and the cactophilic fruit fly, D. mojavensis (Matzkin and

Markow, 2009), as well as two closely-related arthropods, the Arctic collembolan M. 236

arctica (Clark et al., 2009) and the collembolan F. candida (Timmermans et al., 2009), thus facilitating cross-species comparisons of dehydration-induced gene expression. We observed several general similarities between our dataset and the transcriptome of P. vanderplanki, which inhabits temporary pools in tropical Africa. Like B. antarctica, dehydration in P. vanderplanki induced expression of a number of heat shock proteins, including multiple members of the hsp70 family. Additionally, dehydration in P. vanderplanki causes upregulation of genes involved in cell death signaling and ubiquitin- mediated proteasome, patterns that are also quite prevalent in our dataset. However, one conspicuous difference between our dataset and that of P. vanderplanki is the absence of late embryogenesis active (LEA) proteins in the B. antarctica genome, despite B. antarctica and P. vanderplanki being in the same family, Chironomidae. LEA proteins are dehydration-associated proteins found in organisms ranging from bacteria to animals

(Hand et al., 2011), but P. vanderplanki is the only true insect in which LEA genes have been identified.

Like B. antarctica, D. mojavensis is adapted to desiccating environments, albeit warm, desert habitats. As in our dataset, severe dehydration in D. mojavensis elicited significant modulation of numerous metabolic pathways, including downregulation of genes regulating flux through glycolysis and the TCA cycle (Matzkin and Markow,

2009). Thus, it appears downregulation of metabolism may be a general feature of xeric- adapted insects. In contrast, comparing our expression data with A. gambiae revealed little overlap between our dataset and the mosquito response to desiccation. Nonetheless, similar to our results, Wang et al. (2010) observed downregulation of several metabolic genes, particularly genes related to chitin metabolism. 237

The transcriptomic study of dehydration in M. arctica (Clark et al., 2009) included two treatments very similar to our desiccation and cryoprotective dehydration treatments, allowing a formal comparison of the two datasets. M. arctica (formerly

Onychiurus arcticus) is found on numerous islands in the northern Palearctic (Hodkinson et al., 1994), and like B. antarctica is extremely dehydration-tolerant and capable of using cryoprotective dehydration as an overwintering strategy (Worland et al., 1998).

Thus, we investigated whether B. antarctica and M. arctica share common transcriptional responses to desiccation and cryoprotective dehydration, despite their geographic and phylogenetic separation.

Using reciprocal blast, we identified 1,280 putative one-to-one orthologs between the B. antarctica gene models and the M. arctica EST library. Of these, we found 12 genes that were upregulated in response to both desiccation and cryoprotective dehydration in both species, and 7 that were downregulated. Of note, common upregulated genes included an hsp40 gene, two genes involved in the ubiquitin-mediated proteasome, and a GTPase involved in membrane trafficking, thus supporting the central roles of these processes during dehydration. Among the 7 downregulated genes in common were 4 genes involved in carbohydrate hydrolysis and a single peptidase, indicating that downregulation of metabolic genes may be a common attribute of dehydration. Additionally, there were 37 genes that were either up- or downregulated in response to desiccation only, and two genes upregulated only during cryoprotective dehydration. Genes specific to cryoprotective dehydration were a gene involved in unfolded protein binding and an acid-amino acid ligase.

238

Despite the above similarities in dehydration-induced gene expression, the expression profiles of B. antarctica and M. arctica during dehydration were largely different. The Venn diagrams in Fig. 6A,B indicate that more differentially expressed genes are specific to a particular species than are shared between the two species. Also, hierarchical clustering indicates a high degree of separation in the transcript signatures of

B. antarctica and M. arctica (Fig. 6C). Thus, the transcript signature for a particular group is more dependent on the species than the dehydration treatment it experienced.

This suggests that despite being adapted to similar habitats, B. antarctica and M. arctica have evolved distinct molecular responses to dehydration. General comparisons with a second collembolan transcriptomic dataset, that of F. candida (Timmermans et al., 2009), also revealed very little similarity to B. antarctica. In F. candida, desiccation at a constant temperature likewise results in downregulation of lipid and chitin metabolism genes, but aside from these examples very few genes showed similar expression patterns.

These differences in expression patterns may reflect different strategies for combating dehydration; whereas B. antarctica shuts down metabolic activity and waits for favorable conditions to return, F. candida relies on active water vapor absorption to restore water balance during prolonged periods of desiccation. However, since B. antarctica and collembolans are so phylogenetically distant, similar comparisons with closely related chironomids are needed to better understand the evolutionary physiology of dehydration- tolerance in this taxonomic family that is so well known for its extreme tolerance of multiple environmental stresses.

239

Acknowledgements

We thank the staff of Palmer Station for their support during our field season. We also acknowledge Asela Wijeratne and members of the Ohio Agricultural Research and

Development Center Molecular and Cellular Imaging Center for running the sequencing reactions. We appreciate input from Xiaodong Bai during the initial planning phase of this study. Finally, we acknowledge Vanessa Larvor for technical assistance in the GC-

MS experiments.

Grants

This work was supported by NSF OPP-ANT-0837613 and ANT-0837559. Funding for the metabolomics experiments was provided by the French Polar Institute (IPEV 136) and is linked with the SCAR Evolution and Biodiversity in the Antarctic research program.

References

Anders, S., Huber, W., 2010. Differential expression analysis for sequence count data. Genome Biology, 11, R106.

Benoit, J.B., Lopez-Martinez, G., Phillips, Z.P., Patrick, K.R., Denlinger, D.L., 2010. Heat shock proteins contribute to mosquito dehydration tolerance. Journal of Insect Physiology, 56, 151-156.

240

Cantarel, B.L., Korf, I., Robb, S.M.C., Parra, G., Ross, E., Moore, B., Holt, C., Alvarado, A.S., Yandell, M., 2008. MAKER: An easy-to-use annotation pipeline designed for emerging model organism genomes. Genome Research, 18, 188-196.

Chown, S.L., 2002. Respiratory water loss in insects. Comparative Biochemistry and Physiology a-Molecular & Integrative Physiology, 133, 791-804.

Conesa, A., Gotz, S., Garcia-Gomez, J.M., Terol, J., Talon, M., Robles, M., 2005. Blast2GO: a universal tool for annotation, visualization and analysis in functional genomics research. Bioinformatics, 21, 3674-3676.

Cornette, R., Kanamori, Y., Watanabe, M., Nakahara, Y., Gusev, O., Mitsumasu, K., Kadono-Okuda, K., Shimomura, M., Mita, K., Kikawada, T., Okuda, T., 2010. Identification of anhydrobiosis-related genes from an expressed sequence tag database in the cryptobiotic midge Polypedilum vanderplanki (Diptera; Chironomidae). Journal of Biological Chemistry, 285, 35889-35899.

Efron, B., Tibshirani, R., 2007. On testing the significance of sets of genes. Annals of Applied Statistics, 1, 107-129.

Feder, M.E., Hofmann, G.E., 1999. Heat-shock proteins, molecular chaperones, and the stress response: Evolutionary and ecological physiology. Annual Review of Physiology, 61, 243-282.

Gibbs, A.G., 1998. Water-proofing properties of cuticular lipids. American Zoologist, 38, 471-482.

Gibbs, A.G., Chippindale, A.K., Rose, M.R., 1997. Physiological mechanisms of evolved desiccation resistance in Drosophila melanogaster. Journal of Experimental Biology, 200, 1821-1832.

Girardot, F., Monnier, V., Tricoire, H., 2004. Genome wide analysis of common and specific stress responses in adult Drosophila melanogaster. BMC Genomics, 5, 16.

Goldberg, A.L., 2003. Protein degradation and protection against misfolded or damaged proteins. Nature, 426, 895-899.

241

Gorski, S.M., Chittaranjan, S., Pleasance, E.D., Freeman, J.D., Anderson, C.L., Varhol, R.J., Coughlin, S.M., Zuyderduyn, S.D., Jones, S.J.M., Marra, M.A., 2003. A SAGE approach to discovery of genes involved in autophagic cell death. Current Biology, 13, 358-363.

Hand, S.C., Menze, M.A., Toner, M., Boswell, L., Moore, D., 2011. LEA proteins during water stress: Not just for plants anymore. Annual Review of Physiology, 73, 115-134.

Hanson, R.W., Reshef, L., 1997. Regulation of phosphoenolpyruvate carboxykinase (GTP) gene. Annual Review of Biochemistry, 66, 581-611.

He, C.C., Klionsky, D.J., 2009. Regulation mechanisms and signaling pathways of autophagy. Annual Review of Genetics, 43, 67-93.

Hodkinson, I.D., Coulson, S., Webb, N.R., Block, W., Strathdee, A.T., Bale, J.S., 1994. Feeding studies on Onychiurus arcticus (Tullberg) (Collembola, Onychiuridae) on West Spitsbergen. Polar Biology, 14, 17-19.

Kanehisa, M., Goto, S., 2000. KEGG: Kyoto Encyclopedia of Genes and Genomes. Nucleic Acids Research, 28, 27-30.

Kennedy, A.D., 1993. Water as a limiting factor in the Antarctic terrestrial environment - A biogeographical synthesis. Arctic and Alpine Research, 25, 308-315.

Korolchuk, V.I., Menzies, F.M., Rubinsztein, D.C., 2010. Mechanisms of cross-talk between the ubiquitin-proteasome and autophagy-lysosome systems. FEBS Letters, 584, 1393-1398.

242

Larionov, A., Krause, A., Miller, W., 2005. A standard curve based method for relative real time PCR data processing. Bmc Bioinformatics, 6, 62.

Lee, J.H., Budanov, A.V., Park, E.J., Birse, R., Kim, T.E., Perkins, G.A., Ocorr, K., Ellisman, M.H., Bodmer, R., Bier, E., Karin, M., 2010. Sestrin as a feedback inhibitor of TOR that prevents age-related pathologies. Science, 327, 1223-1228.

Li, H., Handsaker, B., Wysoker, A., Fennell, T., Ruan, J., Homer, N., Marth, G., Abecasis, G., Durbin, R., Genome Project Data, P., 2009. The Sequence Alignment/Map format and SAMtools. Bioinformatics, 25, 2078-2079.

Lisi, S., Mazzon, I., White, K., 2000. Diverse domains of THREAD/DIAP1 are required to inhibit apoptosis induced by REAPER and HID in Drosophila. Genetics, 154, 669- 678.

Maiuri, M.C., Zalckvar, E., Kimchi, A., Kroemer, G., 2007. Self-eating and self-killing: crosstalk between autophagy and apoptosis. Nature Reviews Molecular Cell Biology, 8, 741-752.

Marron, M.T., Markow, T.A., Kain, K.J., Gibbs, A.G., 2003. Effects of starvation and desiccation on energy metabolism in desert and mesic Drosophila. Journal of Insect Physiology, 49, 261-270.

Morimoto, R.I., 1998. Regulation of the heat shock transcriptional response: cross talk between a family of heat shock factors, molecular chaperones, and negative regulators. Genes & Development, 12, 3788-3796.

Teets, N.M., Kawarasaki, Y., Lee, R.E., Denlinger, D.L., 2012a. Expression of genes involved in energy mobilization and osmoprotectant synthesis during thermal and

243

desiccation stress in the Antarctic midge, Belgica antarctica. Journal of Comparative Physiology. B, Biochemical, Systemic, and Environmental Physiology, In Press.

Timmermans, M., Roelofs, D., Nota, B., Ylstra, B., Holmstrup, M., 2009. Sugar sweet springtails: on the transcriptional response of Folsomia candida (Collembola) to desiccation stress. Insect Molecular Biology, 18, 737-746.

Trapnell, C., Pachter, L., Salzberg, S.L., 2009. TopHat: discovering splice junctions with RNA-Seq. Bioinformatics, 25, 1105-1111.

Verbruggen, N., Hermans, C., 2008. Proline accumulation in plants: a review. Amino Acids, 35, 753-759.

Wang, M.H., Marinotti, O., Vardo-Zalik, A., Boparai, R., Yan, G.Y., 2010. Genome- wide transcriptional analysis of genes associated with acute desiccation stress in Anopheles gambiae. Plos One, 6, 8.

Worland, M.R., Grubor-Lajsic, G., Montiel, P.O., 1998. Partial desiccation induced by sub-zero temperatures as a component of the survival strategy of the Arctic collembolan Onychiurus arcticus (Tullberg). Journal of Insect Physiology, 44, 211-219.

Young, M.D., Wakefield, M.J., Smyth, G.K., Oshlack, A., 2010. Gene ontology analysis for RNA-seq: accounting for selection bias. Genome Biology, 11, R14.

244

Tables

Table 7.1. Summary of read statistics from Illumina sequencing.Total reads includes the raw number of unprocessed reads obtained from Illumina sequencing, while # of high quality reads refers to the reads that remained after read trimming and filtering during the mapping step. The last row shows the percentage of high quality reads that unambiguously mapped to a B. antarctica gene model.

C1 C2 C3 D1 D2 D3 CD1 CD2 CD3

Total # of 1,423,663 11,739,615 1,180,431 2,836,265 5,211,200 2,746,396 2,620,152 7,924,380 1,861,076 reads

# high quality 1,210,411 10,335,845 1,027,833 2,385,458 4,553,910 2,349,825 2,223,350 6,996,403 1,608,715 reads

% of high quality 77.93 79.26 78.06 77.44 78.89 76.81 77.98 79.39 77.89 mapping to a gene model

245

Table 7.2. Primers used for qPCR validation. The R2 and efficiency were determined by conducting an 8-point standard curve with purified PCR product as template. l(2)efl = lethal-2 essential for life; hsp40 = 40 kD heat shock protein; hsp70 = 70 kD heat shock protein; UDP-GlycTrans = UDP-glycosyltransferase; cyp450a and cyp450b = two different cytochrome P450 genes; tep3 = thiolester containing protein III; spermidine syn. = spermidine synthase; mlck = myosin light chain kinase; rpl19 = ribosomal protein L19.

Gene Accession # R2 E (%) Primers Tm (°C) l(2)efl GAAK01009816 0.9999 97 F: 5'-ATGGTGCGGTCCTTAACCTTGACT 60.4 R: 5'-AAATTGCGCAGCGACACCCTTATC 60.3 hsp40 GAAK01004380 0.9991 77.3 F: 5'-TCGCAATCATTCAACGTTCACGGC 60.4 R: 5'-TGTTGATGTCTTCCAGGCTGACCA-3' 60.4 hsp70 GAAK01011953 0.9956 94.9 F: 5'-CTGCTTTGGCTTACGGTTTGGACA-3' 60 R: 5'-AGATCCCTCGTCGATGGTCAAGAT-3' 59.3

UDP-GlycTrans GAAK01002922 0.9996 95.4 F: 5'-CGAACTGCTGCATTCCAAGCAAGA-3' 60.2 R: 5'-GCAACCAACGGAACGTTGAACTGA-3' 60.2 cyp450a GAAK01011671 0.9996 97.3 F: 5'-TTCGTACTGGAAGAAACTCGGCGT-3' 60.2 R: 5'-ACGGTGTGCCAAACGACTTCAATG-3' 60.3 cyp450b GAAK01006077 0.9998 96.7 F: 5'-TCATGGAGCGCGTCGTTAAAGAGA-3' 60.3 R: 5'-CGGTGCAGCGCGTATATGTTCAAA-3' 60.1 sestrin GAAK01000559 0.9999 96.7 F: 5'-GCTTGTTGCTATCCTGACCGCATT-3' 59.9 R: 5'-TGGCCTCCAGAATCATCAGGTTCA-3' 60.1

Relish GAAK01006924 0.9991 100.2 F: 5'-TCTTGCGAACTCCGCCTTACAGAA-3' 60.3 R: 5'-ACTTGTACCGAAACTCGATGGGCT-3' 60.3 tep3 GAAK01010272 0.9991 100.6 F: 5'-TGACGTCAAAGACGAGGGAAACCA-3' 60.3 R: 5'-TGAACGGGCGGATCATGAACGATA-3' 60.3 thread GAAK01000576 0.9999 94.9 F: 5'-TCGGTTCCTCGTTCTTCGTTTCCA-3' 60.2 R: 5'-ACGACAACCCTTGGGTAGAACACA-3' 60.2 spermidine syn. GAAK01013086 0.99999 94.1 F: 5'-GCCGTTTATGGCTTGTGGGTTTGA-3' 60.3 R: 5'-ACTGCTGGGCCAATAGGATCACTT-3' 60.4 cuticular protein GAAK01011152 0.9997 97.8 F: 5'-TTAACGCCCGCTTGTGATATGTGC-3' 60 R: 5'-AAGAAAGCGGCATCGTAATGCGTG-3' 60.3 mlck GAAK01011539 0.9999 96.4 F: 5'-CCGGTGATTACAAATGCATCGCCA-3' 60.1 R: 5'-ACTCAAGTGTGGTCGTTCGGTTCT-3' 60.3 rpl19 GAAK01002260 0.999 92.7 F: 5'-ACATCCACAAGCGTAAGGCTGAGA-3' 60.3 R: 5'-TTCTTGTTTCTTGGTGGCGATGCG-3' 60.1

246

Table 7.3. GO enrichment analysis of (A) upregulated and (B) downregulated genes upregulated in response to cryoprotective dehydration. GO term Description FDR # up- or Total in down category

(A) Up GO:0007264 small GTPase mediated signal transduction 7.83E-05 29 98 GO:0009408 response to heat 4.18E-03 19 50 GO:0032312 regulation of ARF GTPase activity 4.25E-03 8 13 GO:0016310 phosphorylation 1.81E-02 82 466 GO:0006950 response to stress 1.81E-02 17 59 GO:0007015 actin filament organization 2.32E-02 25 80 GO:0006468 protein phosphorylation 3.18E-02 71 363 GO:0070936 protein K48-linked ubiquitination 4.44E-02 4 6 positive regulation of proteasomal ubiquitin-dependent protein GO:0032436 4.50E-02 4 5 catabolic process GO:0007298 border follicle cell migration 5.20E-02 23 79 GO:0043405 regulation of MAP kinase activity 5.20E-02 3 4 GO:0045081 negative regulation of interleukin-10 biosynthetic process 5.20E-02 3 3 GO:0045599 negative regulation of fat cell differentiation 5.20E-02 3 3 (B) Down GO:0006508 proteolysis 4.93E-18 115 595 GO:0008152 metabolic process 3.58E-13 133 827 GO:0006030 chitin metabolic process 1.89E-08 32 104 GO:0006629 lipid metabolic process 1.95E-08 38 159 GO:0055114 oxidation-reduction process 2.23E-06 101 732 GO:0005975 carbohydrate metabolic process 1.37E-05 37 177 GO:0055085 transmembrane transport 3.85E-05 64 408 GO:0006810 transport 6.19E-05 96 756 GO:0015986 ATP synthesis coupled proton transport 4.85E-04 10 19 GO:0006096 glycolysis 4.85E-04 12 30 GO:0008643 carbohydrate transport 8.43E-04 17 61 GO:0015992 proton transport 9.27E-04 12 30 GO:0006754 ATP biosynthetic process 2.03E-03 11 32 GO:0009253 peptidoglycan catabolic process 4.58E-03 7 13 GO:0015672 monovalent inorganic cation transport 8.97E-03 4 4 GO:0050830 defense response to Gram-positive bacterium 1.30E-02 8 21 GO:0030239 myofibril assembly 4.94E-02 6 15 GO:0060361 flight 5.86E-02 5 10 GO:0001894 tissue homeostasis 6.42E-02 3 3 GO:0044262 cellular carbohydrate metabolic process 6.59E-02 4 6 GO:0031032 actomyosin structure organization 7.34E-02 3 3 GO:0045087 innate immune response 7.34E-02 16 91 GO:0016045 detection of bacterium 7.48E-02 4 6 GO:0015991 ATP hydrolysis coupled proton transport 7.48E-02 8 28

247

Table 7.4. Gene set analysis (GSA) revealing enriched KEGG pathways during cryoprotective dehydration. Gene set name Score p-value

Positive Gene Sets Jak/STAT signaling pathway 1.22 <2E-4 Ubiquitin mediated proteolysis 0.95 <2E-4 Natural killer cell mediated cytotoxicity 0.73 <2E-4 mTOR signaling pathway 0.71 <2E-4 Wnt signaling pathway 0.54 <2E-4 Purine metabolism 0.37 <2E-4 Negative Gene Sets Glycolysis/Gluconeogenesis -1.38 <2E-4 Starch and sucrose metabolism -1.16 <2E-4 Propanoate metabolism -1.15 <2E-4 Galactose metabolism -1.13 <2E-4 Pyruvate metabolism -0.92 <2E-4 Ether lipid metabolism -0.86 <2E-4 Drug metabolism - cytochrome P450 -0.83 <2E-4 Retinol metabolism -0.81 <2E-4 Valine, leucine, and isoleucine degradation -0.79 <2E-4 Metabolism of xenobiotics -0.75 <2E-4 Glutathione metabolism -0.50 <2E-4 Amino sugar and nucleotide sugar metabolism -0.42 <2E-4 *Positive Gene Sets are enriched gene sets in which genes tend to be upregulated, while Negative Gene Sets are enriched gene sets in which genes tend to be downregulated.

248

Table 7.5. GO enrichment analysis of genes more highly expressed in the cryoprotective dehydration group relative to the desiccation group. GO term Definition FDR # upregulated total in category

GO:0006950 response to stress 5.18E-04 11 59 GO:0045214 sarcomere organization 9.78E-04 11 41 GO:0030239 myofibril assembly 3.15E-02 6 15 GO:0009408 response to heat 3.99E-02 9 50 GO:0006508 proteolysis 9.64E-02 31 595

249

Table 7.6. GO enrichment analysis of genes (A) upregulated or (B) downregulated in response to desiccation. GO term Definition FDR # up or Total in down category (A) Up

GO:0006511 ubiquitin-dependent protein catabolic process 7.35E-03 29 82 GO:0007465 R7 cell fate commitment 1.20E-02 10 14 GO:0009408 response to heat 1.20E-02 21 50 GO:0007015 actin filament organization 1.96E-02 28 80 GO:0006468 protein phosphorylation 1.96E-02 86 363 GO:0007264 small GTPase mediated signal transduction 2.75E-02 26 98 GO:0042176 regulation of protein catabolic process 8.51E-02 6 8

(B) Down

GO:0006508 proteolysis 9.33E-20 152 595 GO:0008152 metabolic process 1.52E-17 193 827 GO:0055114 oxidation-reduction process 1.62E-12 160 732 GO:0006030 chitin metabolic process 4.22E-12 44 104 GO:0015992 proton transport 2.55E-08 19 30 GO:0015986 ATP synthesis coupled proton transport 8.39E-07 14 19 GO:0005975 carbohydrate metabolic process 1.48E-05 48 177 GO:0006754 ATP biosynthetic process 3.68E-05 17 32 GO:0006629 lipid metabolic process 8.70E-05 41 159 GO:0006810 transport 1.36E-04 133 756 GO:0006096 glycolysis 2.38E-04 15 30 GO:0055085 transmembrane transport 1.22E-03 82 408 GO:0006099 tricarboxylic acid cycle 2.16E-03 15 36 mitochondrial electron transport, cytochrome c to GO:0006123 3.08E-03 6 7 O2 GO:0003333 amino acid transmembrane transport 1.12E-02 11 28 GO:0006812 cation transport 1.72E-02 15 40 GO:0044262 cellular carbohydrate metabolic process 3.11E-02 5 6 GO:0008643 carbohydrate transport 3.34E-02 18 61 GO:0015672 monovalent inorganic cation transport 3.34E-02 4 4 GO:0015991 ATP hydrolysis coupled proton transport 3.34E-02 11 28 GO:0006032 chitin catabolic process 5.91E-02 9 22 GO:0019083 viral transcription 8.20E-02 6 11 GO:0006865 amino acid transport 9.50E-02 7 15 GO:0009253 peptidoglycan catabolic process 9.98E-02 6 13

250

Table 7.7. Gene set analysis (GSA) revealing enriched KEGG pathways during desiccation.

Gene Set Name Score Adj. p-value Positive Gene Sets* Regulation of Autophagy 1.27 <2E-4 TGF-beta Signaling Pathway 1.11 <2E-4 mTOR Signaling Pathway 0.82 <2E-4 Endocytosis 0.68 <2E-4 Ether lipid metabolism 0.41 <2E-4 Negative Gene Sets* Glyoxylate and dicarboxylate acid metabolism -2.07 <2E-4 Glycolysis/Gluconeogenesis -1.32 <2E-4 Starch and sucrose metabolism -1.19 <2E-4 Galactose Metabolism -1.05 <2E-4 Nicotinate and nicotinamide metabolism -1.03 <2E-4 Propanoate metabolism -1.01 <2E-4 Pyruvate metabolism -0.85 <2E-4 Tryptophan metabolism -0.78 <2E-4 beta-Alanine metabolism -0.71 <2E-4 Valine, leucine, and isoleucine degradation -0.60 <2E-4 Arginine and proline metabolism -0.58 <2E-4 Metabolism of xenobiotics -0.58 <2E-4 Glutathione metabolism -0.54 <2E-4 Fatty acid metabolism -0.54 <2E-4 Folate biosynthesis -0.45 <2E-4 Phagosome -0.35 <2E-4 *Positive Gene Sets are enriched gene sets in which genes tend to be upregulated, while Negative Gene Sets are enriched gene sets in which genes tend to be downregulated.

251

Figures

Figure 7.1. Expression summary (A), dendrogram (B) and Venn diagram (C) showing degree of similarity between the desiccation (D) and cryoprotective dehydration (CD) groups.In (A) and (B), the criteria for differentially expressed genes was FDR<0.05. In (C) the length of each branch indicates the relative distance between two nodes. C = control, D = desiccation, CD = cryoprotective dehydration.

252

Figure 7.2. Results of qPCR validation experiment.In (A) and (B), the fold changes obtained by both RNA-seq and qPCR are graphed together for the C v. D (A) and C v. CD (B) comparisons. In (C), individual log fold changes obtained by RNA-seq and qPCR for each gene in each sample are plotted with the best fit regression line. Log fold changes for each sample were determined relative to the mean of the control group and were normalized to a reference gene, rpl19. C = control, D = desiccation, and CD = cryoprotective dehydration. l(2)efl = lethal-2 essential for life; hsp40 = 40 kD heat shock protein; hsp70 = 70 kD heat shock protein; cyp450a and cyp450b = two different cytochrome P450 genes; tep3 = thiolester containing protein III; mlck = myosin light chain kinase

253

Figure 7.3. Pathway diagrams illustrating upregulation of autophagy-related genes (A) and downregulation of carbohydrate metabolism and ATP synthesis (B).Green boxes indicate significant upregulation, red boxes indicate significant downregulation, and grey boxes indicate no significant change in expression. Gene abbreviations are provided in the Supporting Information. Only the results for the closest homolog to each D. melanogaster gene (determined by BLAST) are included. Consecutive arrows indicate steps where intermediate reactions are not pictured or the intermediate reactions are unknown.

254

Figure 7.4. Changes in metabolite content in response to desiccation and cryoprotective dehydration.Bars represent mean ± SE of the fold change of each metabolite relative to control. Different letters represent significant differences between groups (ANOVA, pooled-t, FDR<0.05).

255

Figure 7.5. Hierarchical clustering of the metabolomics dataset.Hierarchical clustering was conducted on the log metabolite concentrations for each compound in each sample using the Ward method. C = control, D = desiccation, CD = cryoprotective dehydration.

256

Figure 7.6. Venn diagrams (A,B) and dendrogram (C) showing degree of similarity between the gene expression profiles of the Antarctic midge B. antarctica (Ba) and the Arctic springtail M. arctica (Ma) in response to desiccation (D) and cryoproective dehydration (CD).The numbers of shared and unique upregulated genes are depicted in (A), while the numbers of shared and unique downregulated genes are depicted in (B). In (C), hierarchical clustering was conducted on the log fold change values for each orthologous gene in each sample.

257

Chapter 8: Major Findings and Conclusions

This dissertation explored the molecular mechanisms of environmental stress tolerance in both temperate and polar insects. Specifically, Chapters 2 and 3 addressed cellular and molecular mechanisms of rapid cold-hardening (RCH) in temperate flies, while Chapters 4-7 covered the energetic and mechanisms of stress tolerance in the

Antarctic midge, B. antarctica. The major findings in this dissertation include: 1) clarifying the role of transcription during rapid cold-hardening and recovery from cold shock; 2) identifying calcium signaling as an important regulator of cold-sensing and rapid cold-hardening; 3) quantifying the energetic consequences of ecologically relevant repeated stress exposures in B. antarctica; 4) identifying the role of metabolic restructuring at the transcription level during thermal and dehydration stress in B. antarctica; and 5) providing the first genome-wide transcriptional profile of extreme stress tolerance in an Antarctic insect. Conclusions from each of the sections, as well as specific results supporting them, are summarized below.

Part 1: Cellular and Molecular Mechanisms of Rapid Cold-Hardening

In Chapters 2 and 3, I addressed the cellular and molecular mechanisms of RCH in temperate dipterans. In Chapter 2, I used transcriptomics and metabolomics to survey gene expression and biochemical changes accompanying RCH, while in Chapter 3 I used

258

targeted cell biology experiments to test the role of calcium signaling during RCH. Taken together, the results indicate that cold-sensing and RCH are second messenger-driven processes that are not dependent on the synthesis of new gene products. Specific summary points and conclusions are listed below.

1. Transcriptional regulation is not a major driver of RCH. Using a cDNA

microarray, the expression of ~15,000 ESTs was measured during rapid cold-

hardening (2 h at 0°C) in the flesh fly, Sarcophaga bullata. Despite the dramatic

increase in cold tolerance afforded by RCH (RCH increased survival at -10°C

from 0 to ~50%), we failed to detect differential expression of a single transcript

during RCH. Given the speed (minutes to hours) and low temperature (0°C) at

which RCH occurs in S. bullata, it appears there is no opportunity for

transcription of new gene products. Thus, even though rapid transcriptional

changes during cold-acclimation have been documented in plants (Thomashow,

1999), our results indicate this is not the case in insects.

2. RCH does not alter the transcriptional signature during recovery from cold

shock. Our microarray results clearly indicated that transcriptional regulation is

not a major mechanism of RCH during the hardening period. We also tested the

hypothesis that RCH alters the transcriptional signature during recovery from

subsequent cold-shock. RCH activates cell-signaling pathways such as MAP

kinase (Fujiwara and Denlinger, 2007) and apoptosis (Yi and Lee, 2011)

pathways, and we expected this to be reflected in the transcriptional signature

during recovery from a subsequent cold shock. However, once again, this was not

the case. Comparing flies that were given 2 h of RCH prior to cold shock with 259

flies that were directly exposed to cold shock, only 5 transcripts differed in

expression when measured after 2 h of recovery, and none of these differed by

more than 33%.

3. An abundance of transcripts are differentially regulated during recovery from

cold shock. While RCH failed to elicit major changes in gene expression,

recovery from cold shock elicited significant changes in gene expression, as

roughly 10% of the transcriptome was differentially expressed within 2 h of

recovery. Enrichment analyses identified numerous over-represented functional

categories, including an abundance of heat shock protein chaperones and genes

involved in actin cytoskeleton organization. In addition, several enriched

pathways, such as insulin signaling and JAK/Stat signaling, have not been

previously implicated during cold stress. However, additional targeted

experiments are needed to verify the roles of these pathways during recovery

from cold shock. Finally, there was considerable overlap between our dataset and

previous transcriptional studies of stress tolerance in Drosophila melanogaster.

From this we conclude there are common axes involved in diverse forms of

environmental stress, including cold, heat, and oxidative stress.

4. RCH and cold shock have significant effects on the metabolome. While RCH

failed to significantly alter the transcriptome, several metabolic pathways were

directly impacted by RCH. RCH stimulated glycolysis and gluconeogenesis,

presumably to provide the necessary energy for hardening as well as provide

glucose intermediates for downstream cryoprotectant synthesis. In addition, RCH

altered the metabolic signature during recovery from cold-shock; flies exposed to

260

RCH generally had a lower degree of metabolic perturbation compared to those

directly exposed to cold shock. In fact, cold shock caused major disruptions in

metabolism; the level of nearly every metabolite measured was significantly

altered during recovery from cold shock.

5. Recovery from cold shock involves coordinated changes in gene expression and

metabolism. In several instances, changes in metabolic gene expression during

recovery from cold shock were closely correlated with changes in metabolic end

points. For example, 4-fold upregulation of phosphoenolpyruvate carboxykinase

(PEPCK), the rate limiting enzyme of gluconeogenesis, correlated with a 5-fold

increase in glucose levels. Also, four biochemical pathways were enriched at both

the gene and metabolite level, indicating a high level of agreement between the

two datasets.

6. Calcium signaling mediates cellular cold-sensing and RCH in insects. Since

RCH does not require transcription of new gene products, it is likely mediated by

post-transcriptional, second messenger pathways. In Chapter 3, we explored the

role of calcium-signaling in mediating cold-sensing and RCH the freeze-tolerant

gall fly, Eurosta solidaginis, and the freeze-intolerant S. bullata. In E.

solidaginis, chilling to 0°C caused a 40% increase in intracellular calcium and

activated calcium/calmodulin-dependent protein kinase II. Blocking various

components of calcium signaling pathways with pharmacological agents

prevented cellular RCH, indicating calcium is required for this process to occur.

Similar results were obtained with S. bullata, suggesting calcium signaling plays

a general role in cold-sensing for both freeze-tolerant and freeze-avoiding

261

species. These results provide the first evidence of the signaling pathway that

senses low temperature and triggers downstream hardening events.

Part 2: Mechanisms of Stress Tolerance in B. antarctica

In chapters 4-7, I addressed various aspects of environmental stress tolerance in the Antarctic midge, B. antarctica. Chapters 4-5 covered the survival and energetic costs of repeated environmental stress exposure in B. antarctica, while chapters 6-7 examined transcriptional mechanisms of stress tolerance using both targeted and non- targeted approaches. Overall, results from Chapters 4-5 indicate that repeated stress bouts exacts a significant energetic cost in B. antarctica. In Chapter 6, I found significant restructuring of metabolism at the transcript level in response to environmental stress, with the magnitude and direction of change dependent on the nature of the stress. Finally, in Chapter 7, I used RNA-seq to identify molecular mechanisms of extreme dehydration tolerance, and found a number of pathways (e.g. autophagy) that have not been implicated in insect stress responses. Specific summary points and conclusions are listed below.

1. Repeated cold exposure in the frozen state has significant survival and

energetic costs. While previous studies in our lab have characterized the limits of

stress tolerance in B. antarctica, these studies have only examined exposure to a

single bout of stress. In nature, larvae are exposed to multiple bouts of stress, and

in many instances repeated stress exposures produce novel physiological

responses that cannot be predicted on the basis of a single exposure (Marshall and

262

Sinclair, 2010). Furthermore, climate change models predict an increased

frequency of extreme events and freeze-thaw cycles (Easterling et al., 2000),

which could limit the survival and range of certain insect populations. In B.

antarctica, while all larvae survived a single 12-h bout of freezing at -5°C, 5

cycles of freeze-thaw resulted in significant mortality and energetic costs.

Compared to control larvae, larvae that experience 5 freeze-thaw events had 15%

less total energy content and 30% less carbohydrate content. Thus, there are

cumulative costs to repeated freeze-thaw cycles that result in significant mortality,

tissue damage, and energetic deficits.

2. Freezing, but not chilling per se, is deleterious to larvae of B. antarctica. The

microhabitat temperature for B. antarctica rarely dips below -5°C, while the

supercooling point is around -7°C (Lee et al., 2006). Thus, when environmental

moisture conditions are low, larvae have the capability to avoid freezing and

remain supercooled at subzero temperatures. However, when conditions are

moist, larvae are susceptible to inoculative freezing from environmental ice

crystals (Elnitsky et al., 2008). Thus, we conducted experiments to determine

which strategy, freezing or supercooling, is preferable during both repeated and

prolonged cold exposure. Our results indicated that is preferable, both from a

survival and energetic perspective, to remain supercooled during acute cold

exposure. While frozen larvae experience significant lethal and sublethal

consequences during repeated cold exposure (see above), supercooled larvae

experienced little mortality and no measurable energetic consequences following

repeated cold exposure. Similar results were obtained for larvae exposed to -5°C

263

for 60 h continuously. Thus, we conclude that it is preferable for larvae to seek

dry microhabitats and avoid freezing during acute bouts of low temperature.

3. Repeated dehydration and rehydration is energetically costly for larvae of B.

antarctica. In Chapter 5, we similarly quantified the energetic costs of repeated

dehydration exposure. Compared to larvae exposed to a single

dehydration/rehydration cycle (75% RH for 24 h, 100% RH for 24 h), larvae

exposed to multiple cycles lost significantly less water and had higher water

contents following rehydration, suggesting some degree of adaptation to repeated

bouts of dehydration. However, as with freezing, there were significant energetic

consequences of repeated dehydration exposure. Larvae exposed to 5 dehydration

bouts had 89% less glycogen and 48% less trehalose than control counterparts,

indicating a dramatic depletion of carbohydration energy stores to cope with

dehydration stress. Thus, we conclude that while larvae of B. antarctica are one of

the most dehydration-tolerant insects known, there are severe costs to tolerating

dehydration, which likely have fitness consequences and limit the potential range

of B. antarctica.

4. Transcriptional regulation of metabolism is an important component of stress

tolerance in B. antarctica. Previous work has demonstrated that larvae of B.

antarctica undergo significant biochemical changes in response to environmental

stress, including mobilization of energy reserves (see above) and synthesis of

protective osmolytes (e.g. Elnitsky et al., 2008; Michaud et al., 2008). However,

the molecular mechanisms of these biochemical changes had not been addressed.

We measured the expression of 11 different metabolic transcripts, and found that

264

gene expression patterns generally reflected expected biochemical changes. For

example, both thermal stress (i.e. heat and cold) and dehydration stress caused

significant upregulation of PEPCK, reflecting the need to synthesize glucose to

support the energetic demands of stress tolerance and provide substrates for the

synthesis of osmoprotectants. Also, genes involved in trehalose synthesis were

strongly upregulated during prolonged dehydration, supporting the essential role

of trehalose as an osmoprotectant during dehydration stress. However, in some

cases, changes in gene expression were counter to expectations, indicating post-

transcriptional regulation of biochemical flux in certain pathways.

5. Expression changes of metabolic genes depend on the nature and duration of

stress. In some instances, the magnitude and direction of transcriptional regulation

depended on the type and severity of the stress experienced. For example, while

PEPCK was upregulated in all stress treatments, the magnitude of upregulation

was positively correlated with the length of stress exposure. Also, larvae

supercooled at -5°C had significantly higher levels of PEPCK than those frozen at

-5°C; perhaps this in part explains the improved survival and energetic status of

supercooled larvae (see above). Finally, in some cases expression patterns of

larvae exposed to rapid, severe dehydration were opposite those of larvae exposed

to gradual, prolonged dehydration. For example, while trehalose and proline

synthesis genes were downregulated following rapid dehydration, they were

upregulated following gradual, prolonged dehydration. The transcript signature

following fast dehydration was similar to that following acute cold stress,

265

suggesting sudden exposure to both extreme cold and water stress produce similar

metabolic gene expression patterns.

6. Extreme dehydration causes widespread changes in gene expression in B.

antarctica. Using RNA-seq, we profiled gene expression patterns in response to

desiccation at a constant temperature and cryoprotective dehydration. Both

treatments caused large-scale changes in gene expression, as >20% of 13,500

genes were differentially expressed. Expression patterns following both types of

dehydration were similar, suggesting water loss is the driving factor for the

observed transcriptional changes. In addition to upregulation of well-known stress

genes such as heat shock proteins and PEPCK, we observed upregulation of cell-

recycling pathways such as the ubiquitin-proteasome and autophagy. Numerous

genes involved in regulating and executing autophagy were upregulated, and we

hypothesize that prolonged stress induces an “autophagy switch” that recycles

macromolecules and prolongs cell survival during times of stress. Furthermore,

we observed downregulation of many metabolic pathways, including glycolysis,

TCA cycle, and lipid metabolism, indicating a general shutdown of metabolism

during extreme dehydration. This metabolic shutdown was reflected in the

metabolome, as the levels of key glycolytic and TCA cycle intermediates

supported reduced activity of these pathways.

7. Molecular responses to extreme dehydration are largely species specific. The

dehydration-induced transcriptome of several other invertebrates, including a

mosquito, a tropical midge, and an arctic collembolan, have also been measured.

General comparisons between these datasets and ours revealed some

266

commonalities, such as upregulation of heat shock proteins, but on the whole

expression changes were species-specific. A microarray study of desiccation and

cryoprotective dehydration in an Arctic collembolan, Megaphorura arctica,

included treatments very similar to ours, allowing a formal comparison of

expression data. However, very few orthologous genes showed similar expression

profiles, and multivariate clustering indicated a high degree of separation in the

transcriptomes of B. antarctica and M. arctica. Thus, despite being similarly

adapted to cold, dry, polar environments, the molecular mechanisms of extreme

dehydration tolerance are largely different.

References

Easterling, D.R., Meehl, G.A., Parmesan, C., Changnon, S.A., Karl, T.R., Mearns, L.O., 2000. Climate extremes: Observations, modeling, and impacts. Science, 289, 2068-2074.

Michaud, M.R., Benoit, J.B., Lopez-Martinez, G., Elnitsky, M.A., Lee, R.E., Denlinger, D.L., 2008. Metabolomics reveals unique and shared metabolic changes in response to

267

heat shock, freezing and desiccation in the Antarctic midge, Belgica antarctica. Journal of Insect Physiology, 54, 645-655.

Thomashow, M.F., 1999. Plant cold acclimation: Freezing tolerance genes and regulatory mechanisms. Annual Review of Plant Physiology and Plant Molecular Biology, 50, 571- 599.

Yi, S.-X., Lee, R.E., 2011. Rapid cold-hardening blocks cold-induced apoptosis by inhibiting the activation of pro-caspases in the flesh fly Sarcophaga crassipalpis. Apoptosis, 16, 249-255.

268

Bibliography