Investigating the Interplay of Physiological and Molecular Mechanisms Underpinning Programmable Aspects of Heat Stress in Pigs Jacob Seibert Iowa State University

Iowa State University Capstones, Theses and Graduate Theses and Dissertations Dissertations

2018 Investigating the interplay of physiological and molecular mechanisms underpinning programmable aspects of heat stress in pigs Jacob Seibert Iowa State University

Follow this and additional works at: https://lib.dr.iastate.edu/etd Part of the Agriculture Commons, Animal Sciences Commons, Molecular Biology Commons, and the Physiology Commons

Recommended Citation Seibert, Jacob, "Investigating the interplay of physiological and molecular mechanisms underpinning programmable aspects of heat stress in pigs" (2018). Graduate Theses and Dissertations. 16878. https://lib.dr.iastate.edu/etd/16878

This Dissertation is brought to you for free and open access by the Iowa State University Capstones, Theses and Dissertations at Iowa State University Digital Repository. It has been accepted for inclusion in Graduate Theses and Dissertations by an authorized administrator of Iowa State University Digital Repository. For more information, please contact [email protected]. Investigating the interplay of physiological and molecular mechanisms underpinning programmable aspects of heat stress in pigs

Jacob Todd Seibert

A dissertation submitted to the graduate faculty

in partial fulfillment of the requirements for the degree of

DOCTOR OF PHILOSOPHY

Major: Molecular, Cellular, and Developmental Biology

Program of Study Committee: Lance Baumgard, Co-major Professor Jason Ross, Co-major Professor Steven Lonergan James Reecy Christopher Tuggle

The student author, whose presentation of the scholarship herein was approved by the program of study committee, is solely responsible for the content of this dissertation. The Graduate College will ensure this dissertation is globally accessible and will not permit alterations after a degree is conferred.

Iowa State University

Ames, Iowa

2018

DEDICATION

I dedicate this dissertation to every person that has pushed me and encouraged me to keep going and to do my best with anything that I set my mind to, family, friends, coaches, mentors, and teachers.

"The price of success is hard work, dedication to the job at hand, and the determination

that whether we win or lose, we have applied the best of ourselves to the task at hand."

- Vince Lombardi

iii

TABLE OF CONTENTS

Page

LIST OF FIGURES ...... vi

LIST OF TABLES ...... ix

ACKNOWLEDGMENTS ...... xi

ABSTRACT ...... xiv

CHAPTER 1. GENERAL INTRODUCTION ...... 1 The Impact of Heat Stress on Livestock Agriculture and Food Security ...... 1 Physiological Responses to HS ...... 2 Thermotolerance ...... 6 Genetics of Thermotolerance ...... 6 Acquired Thermotolerance ...... 8 Heat Shock Proteins and Cellular Thermotolerance ...... 9 Developmental Programming and Epigenetics ...... 11 Developmental Programming and Effects in Postnatal Life ...... 11 Heat Stress Effects on the Developing Offspring ...... 14 Types of Epigenetic Marks and Effects on Transcriptional Regulation ...... 15 Epigenetic Events in Developmental Programming ...... 21 Heat Stress Influence on Epigenetic Marks ...... 23 Summary ...... 27 References ...... 28

CHAPTER 2. CHARACTERIZING THE ACUTE HEAT STRESS RESPONSE IN GILTS: I. THERMOREGULATORY AND PRODUCTION VARIABLES ...... 47 Abstract ...... 47 Introduction ...... 49 Materials and Methods ...... 50 Animals and Experimental Design ...... 50 Production and Thermoregulatory Measurements ...... 51 Determination of HS Tolerance and Susceptibility Based on the Thermoregulatory Response ...... 52 Statistical Analysis ...... 52 Results ...... 53 Thermoregulatory and Productivity Variables have Marginal Relationships during TN Conditions ...... 53 Minor Relationships between Thermoregulation and Production Parameters are Observed during HS Conditions ...... 53 Relationship between Thermoregulatory Indices and Production Characteristics during TN and HS Conditions ...... 54 iv

Pigs Retrospectively Classified as SUS had Improved Production Characteristics during TN Conditions yet Impaired Thermoregulatory Ability during HS Compared to TOL Pigs ...... 54 Discussion ...... 55 Conclusion ...... 59 Acknowledgements ...... 60 Declaration of Interest ...... 60 Literature Cited ...... 60 Tables and Figures ...... 64

CHAPTER 3. EFFECTS OF HEAT STRESS AND INSULIN SENSITIZERS ON PIG ADIPOSE TISSUE ...... 73 Abstract ...... 74 Introduction ...... 75 Materials and Methods ...... 76 Animals and Experimental Design ...... 76 Production and Thermoregulatory Measurements ...... 78 Blood Sampling and Analysis ...... 79 Tissue Collection and Fatty Acid Composition Analysis ...... 79 Adipose Moisture Content and Morphology Analysis ...... 80 Statistical Analysis ...... 81 Results ...... 82 Thermoregulatory, Production, and Blood Indices ...... 82 Fatty Acid Composition ...... 83 Back Fat Thickness, Adipocyte Area, and Adipose Moisture Content ...... 83 Discussion ...... 84 Conclusion ...... 87 Acknowledgements ...... 88 Literature Cited ...... 88 Tables and Figures ...... 93

CHAPTER 4. DIFFERENTIATING BETWEEN THE EFFECTS OF HEAT STRESS AND LIPOPOLYSACCHARIDE ON PORCINE OVARIAN HEAT SHOCK PROTEIN EXPRESSION ...... 106 Abstract ...... 107 Introduction ...... 109 Materials and Methods ...... 111 Animals and Experimental Design ...... 111 Quantitative One-step RT-PCR ...... 112 Western Blot ...... 113 Statistical Analysis ...... 115 Results ...... 115 HS or LPS have Marginal Effects on Ovarian HSP Transcript Abundance ...... 115 The Protein Abundance of Ovarian HSPs are Selectively Influenced by HS and LPS ...... 115 Whole-ovarian Hypoxia Markers are not Influenced by HS or LPS during the Follicular Phase, but are Reduced in the CL during HS in the Luteal Phase ...... 117 v

Discussion ...... 118 Conclusion ...... 122 Acknowledgements ...... 123 Declaration of Interest ...... 123 Literature Cited ...... 123 Tables and Figures ...... 132

CHAPTER 5. IMPACT OF PRENATAL AND POSTNATAL HEAT STRESS ON THE ADIPOSE, LIVER, AND MUSCLE-SPECIFIC TRANSCRIPTOME AND TARGETED DNA METHYLATION RESPONSES OF MUSCLE IN PIGS ...... 139 Abstract ...... 140 Introduction ...... 141 Materials and Methods ...... 143 Animals and Experimental Design ...... 143 RNA Extraction, RNA-Seq Library Preparation, Sequencing, and Alignment ... 144 Bioinformatic Analysis ...... 145 Quantitative One-step RT-PCR ...... 145 Bisulfite Sequencing of Targeted Genomic Regions ...... 146 Western Blot ...... 147 Statistical Analysis ...... 148 Results ...... 150 Gene Expression ...... 150 RNA Sequencing of Adipose, Liver, and Muscle ...... 150 Gene Ontology and Bioinformatics ...... 151 Quantitative PCR Analysis and Comparison to RNA-Seq ...... 152 Transcriptional Control ...... 153 Protein Expression ...... 155 Discussion ...... 157 Conclusions ...... 163 Literature Cited ...... 163 Tables and Figures ...... 173

CHAPTER 6. GENERAL DISCUSSION ...... 179 Possible Contributions of Rectal Temperature Variation during Acute HS ...... 180 Does HS Result in Faster Corpus Luteum Regression? ...... 183 HIF1A Regulation and Potential Role during HS ...... 184 Other Thoughts Regarding HS during Pre- or Postnatal Development ...... 185 Literature Cited ...... 187

APPENDIX A. SUPPLEMENTARY TABLES AND FIGURES FOR CHAPTER 5 .. 191

APPENDIX B. DAILY 12 HOUR RECTAL TEMPERATURE AND WEEKLY AVERAGE DAILY GAIN MEASUREMENTS DURING HEAT STRESS FOR 21 DAYS ...... 232

APPENDIX C. HOURLY RECTAL TEMPERATURE RESPONSE DURING ACUTE HEAT STRESS ...... 233 vi

LIST OF FIGURES

Page

Figure 1.1. Timeline and effects of the Dutch famine birth cohort comprised of 2414 people ...... 13

Figure 1.2. A summary of DNA methylation dynamics during development ...... 18

Figure 1.3. DNA methylation and histone modification during development ...... 21

Figure 2.2. Relationships between rectal temperature (TR) and the change (Δ) in feed intake (FI) of pigs during period (P) 2 (heat stress [HS] conditions; A) and of HS TR with the ΔBW (B) ...... 68

Figure 2.5. Correlations between the percent change (Δ) in feed intake (FI) and the percent ΔBW ...... 71

Figure 3.3. Effect of thermoneutral (TN) ad libitum (TNAL), TN pair-fed (TNPF), and all heat stress (HS) treatments combined on adipose tissue moisture content during period 2 ...... 105

Figure 4.1. Experimental scheme for the study ...... 133

Figure 4.2. Effects of heat stress (HS) or lipopolysaccharide (LPS) on ovarian or CL transcript abundance of heat shock protein (HSPs) or hypoxia- responsive genes ...... 134

Figure 4.3. Heat stress (HS) during the follicular phase alters heat shock protein (HSP) abundance in the ovary ...... 135

Figure 4.4. Lipopolysaccharide (LPS) during the follicular phase alters heat shock protein (HSP) abundance in the ovary ...... 136

Figure 4.5. Heat stress (HS) during the luteal phase alters heat shock protein (HSP) abundance in the corpus luteum (CL) ...... 137

Figure 4.6. Effects of heat stress (HS) or lipopolysaccharide (LPS) on ovarian VEGFA protein abundance ...... 138

Figure 5.1. Gestational thermal treatment alters transcriptomic profile in adipose, liver, and longissimus dorsi muscle ...... 176

Figure 5.2. Effects of pre- and postnatal thermal treatment and sex on cumulative methylation levels of CpGs overlapping PYGM, HSF1, and HIF1A genomic loci in longissimus dorsi muscle ...... 177

Figure 6.1. Relationship between rectal temperature (TR) and feed intake during period 1 (thermoneutral [TN] conditions) from the study in chapter 2 ...... 182

Figure A.2. Effects of pre- and postnatal thermal treatment and sex on protein abundance of HSPA1A in longissimus dorsi muscle ...... 231 viii

Figure B.1. Effect of thermoneutral ad libitum (TNAL), thermoneutral pair-fed (TNPF), heat stress ad libitum (HSAL), heat stress sterculic oil (HSSO), and heat stress chromium (HSCr) on 12 h rectal temperature (TR) area under the curve (AUC) by day (top) and average daily gain by week (bottom) from study 2 (chapter 3) ...... 232

Figure C.1. Rectal temperature (TR) response during 24 h of heat stress (HS) from the first replication of study 1 (chapter 2) ...... 233 ix

LIST OF TABLES

Page

Table 1.1. The effects of heat stress on epigenetic modifications in various species ...... 25

Table 2.1. Ingredients and chemical composition of diet for growing pigs (as-fed basis) ...... 64

Table 2.2. Effects of environment on thermoregulatory indices and production variables in growing pigs ...... 65

Table 2.3. Effects of retrospective classification and environment on thermoregulatory indices and production variables for growing pigs during the study ...... 66

Table 3.1. Ingredients and chemical composition of diet for growing pigs (as-fed basis) ...... 93

Table 3.2 Effects of period 2 treatment on body temperature indices ...... 94

Table 3.3. Effects of period 2 treatment on production parameters ...... 95

Table 3.4. Effects of period 2 treatment on fatty acid (FA) profile of abdominal adipose tissue ...... 96

Table 3.5 Effect of period 2 treatment on fatty acid (FA) profile of inner s.c. adipose tissue ...... 98

Table 3.6. Effects of period 2 treatment on fatty acid (FA) profile of outer s.c. adipose tissue ...... 100

Table 3.7. Effects of period 2 treatment on backfat thickness and fatty acid (FA) profile, adipocyte area, and moisture content of abdominal, inner, and outer s.c. adipose tissue depots during the study ...... 102

Table 4.1. Primer information for quantitative amplification of each target gene ...... 132

Table 5.1. Primer information for quantitative amplification of each target gene ...... 173

Table 5.2. Comparison of qRT-PCR and RNA-Seq results for selected genes ...... 174

Table A.1. Genes selected based on the RNA-Seq analysis for bisulfite sequencing of targeted genomic regions ...... 191

Table A.2. Functional annotation clusters of biological terms representing processes affected from gestational heat stress (HS) in adipose tissue ...... 192

Table A.3. Functional annotation clusters of biological terms representing processes affected from gestational heat stress (HS) in liver ...... 193

Table A.4. Functional annotation clusters of biological terms representing processes affected from gestational heat stress (HS) in longissimus dorsi muscle .... 195

Table A.5. Functional annotation clusters from Ingenuity Pathway Analysis of biological terms representing processes affected from gestational heat stress (HS) in longissimus dorsi muscle ...... 196

Table A.6. Genomic loci of amplicons from bisulfite sequencing and number of CpGs assessed/amplicon ...... 197

Table A.7. Quality of the targeted bisulfite sequencing analysis ...... 199

Table A.9. Effects of gestational treatment, sex, and their interactions on protein abundance in muscle from the western blot analysis ...... 227

Table A.10. Effects of thermal treatment during the first half of gestation, sex, and their interactions on protein abundance in muscle from the western blot analysis ...... 228

Table A.11. Effects of postnatal treatment, sex, and their interactions on protein abundance in muscle from the western blot analysis ...... 229

ACKNOWLEDGMENTS

Drs. Ross and Baumgard, I cannot thank you both enough for allowing me to pursue my doctorate degree under your guidance. I am grateful for everything you both have done for me during my time here. I appreciate your mentorship, encouragement to pursue new ideas, and also both of you challenging me to think critically. It has been challenging, especially with my “cell-centric” thinking early on, as you both instilled the concept of “going from the bench to the beast”. You both held the bar high and I truly thank you for all the training and treating me like a colleague from day one.

I want to thank the members of my PhD committee for your guidance and tutelage. Dr. Tuggle, Dr. Reecy, and Dr. Lonergan, thank you for taking the time to discuss scientific concepts and questions along the way, troubleshooting, providing your expert advice, and for the hallway small talk along the way (and sometimes providing edits on a scientific abstract on short notice!). I also want to thank you Dr. Lonergan for encouraging me to go to graduate school (and also to do it at ISU) back in 2011. It meant a lot to me at the time, especially being an undergraduate that did not think he was good enough. Again, thank you all for your time and input on my graduate education.

I cannot express enough gratitude for my past and current labmates from both the

Baumgard and Ross groups: Dr. Sara Kvidera, Dr. Mo Abuajamieh, Dr. Maria Victoria

Sanz Fernandez, Dr. Jay Johnson, Erin Horst, Johana Mayorga, Mohmmad Al-Qaisi, Sam

Lei, Dr. Malavika Adur, Dr. Yunsheng Li, Dr. Benjamin Hale, Dr. Elizabeth Hines, Matt

Romoser, Blythe Shultz, Zoe Kiefer, Theresa Johnson, Candi Hager, and Kody Graves.

Thank you guys for helping me conduct the live phase experiments, talking science, troubleshooting lab assays, and just being great people to talk to. Thank you also for the xii scientific and nonscientific discussions from Dr. Josh Selsby, Dr. Aileen Keating, and past and present members from the Selsby and Keating groups: Kendra Clark, Hannah

Spaulding, Dr. Shanthi Ganesan, Dr. Jill Madden, Dr. Katrin Hollinger, Mackenzie

Dickson, Katie Bidne, and Porsha Thomas. Dr. James Koltes, thank you for your statistical expertise and helping me during critical times during my PhD. I also want to thank Mike Kaiser and Dr. Kyle Grubbs for helping me troubleshoot assays and for the nonscientific discussions, the latter being needed in graduate school at times.

Thank you to the past and present ISU research farm staff that have helped address issues when conducting live phase experiments: Trey Faaborg, Jacob Schulte, and the late Jacob Myers. All others in the Animal Science department that have helped me along the way, thank you.

I also want to thank my past teachers of science. Denise Philipp, thank you for getting me initially interested in biology during a time in my life when I did not really care about science. I have been told that everybody has one influential person that inspires a turning point in oneself, and you did that for me. Thank you also to Drs. Jodi

Enos-Berlage, Kevin Kraus, and Marian Kaehler for your high standards, teaching, one- on-one mentoring, and just being someone to talk to during my time at Luther. You inspired me and also encouraged me to pursue a life in science and to this day, and after, I truly cherish it.

Mom and Dad, thank you for all of your love and support. I could not have gotten this far without the work ethic you taught me and the “do your best” mentality you instilled in me long ago. Thank you for always believing in me, encouraging me to pursue what I am passionate about, and reminding me that God does not give us more xiii than we can handle. Mike, Anna, and Emily, thank you guys for being great siblings, your love and support, and being people that I can rely on. To all my other family and friends, thanks for listening and the support.

Last but not least, I’d like to thank my family. To my wife Paige, you have been here with me every step of the way, during the ups and downs of this journey they call graduate school. Thank you for your support, patience, listening, and being a great mother. I am glad God brought you into my life. I love you and could not have done this without you. Landrie and Kandrik, your Dad loves you and you both don’t even realize how you have helped me along the way. Looking forward to the next step for us!

xiv

ABSTRACT

Heat stress (HS) undermines production efficiency in all animal agriculture, culminating in major economic losses annually. The effects of HS in pigs is realized through depressed growth, altered body composition, and impaired reproductive performance. Pigs, unlike other mammals, do not possess functional sweat glands, making them rely on other mechanisms to thermoregulate during a heat challenge.

Continual investigation of hyperthermia in pigs is crucial for developing new strategies and/or technologies that mitigate the effects of HS imposed on the pork industry. The studies conducted in this dissertation investigate the complex interplay of whole-animal physiology as well as cellular and molecular mechanisms within specific tissues.

Identifying pigs demonstrating HS tolerance without compromising productivity would be valuable. Thus, study 1 (chapter 2) was carried out to evaluate the degree to which thermoregulatory responses are associated with production phenotypes during acute HS.

Hyperthermic pigs accumulate more adipose tissue than their feed intake predicts, which may be mediated by HS-induced hyperinsulinemia. However, less is known how HS influences fatty acid composition and adipocyte size, both of which are regulated by insulin and can influence carcass fat quality. To better understand the underlying physiology of how HS influences body composition, study 2 (chapter 3) investigated insulin’s potential role in affecting adipose characteristics during HS. A cardinal physiological response to HS is compromised intestinal integrity and concomitant increased gut-derived lipopolysaccharide (LPS) into systemic circulation. While HS causes female infertility, many of the reproductive consequences imposed by the abiotic factor are also recapitulated by direct exposures to LPS independently. Additionally, heat xv shock proteins are known as molecular stress responders, but their roles during in vivo periods of stress in the ovary are ill-defined. Thus, study 3 (chapter 4) was conducted to evaluate the direct and indirect (i.e., LPS) effects of HS on ovarian heat shock protein machinery. Recent studies also demonstrate that body composition and metabolic responses to HS can also be programmed via gestational HS, directly impacting the progeny’s postnatal development. Developmental programming is mediated through specific epigenetic mechanisms and study 4 (chapter 5) investigated alterations to DNA methylation and protein responses in muscle as a result of intrauterine HS. In conclusion, the cumulative discoveries within this dissertation expands our knowledge regarding the negative consequences of thermal biology in pigs, a pivotal step for mitigating the deleterious effects of environmental HS on animal health and performance.

CHAPTER 1. GENERAL INTRODUCTION

The Impact of Heat Stress on Livestock Agriculture and Food Security

Now more than ever, efficient and sustainable agriculture is crucial for ensuring food security for a rapidly growing human population. Despondently, there is an estimated 815 million people (equating to one in nine people) that are undernourished according to a United Nations Food and Agriculture Organization 2016 evaluation (FAO et al., 2017). Moreover, global agriculture will need to produce at least 70% more food to meet the needs of a predicted 9 billion people by 2050 (World Bank, 2007; United

Nations, 2017), which includes the livestock sector doubling meat output (Alexandratos and Bruinsma, 2012).

Heat stress (HS) is one factor that limits production potential and efficiency in all animal agriculture, resulting in major economic losses annually (St-Pierre et al., 2003).

The effects of HS on the swine industry alone translates to approximately $900 million every year (Pollmann, 2010), encompassing not just increased mortalities, but also negatively affecting every level of pork production, including compromised feed efficiency and growth, altered body compositions, and reduced fecundity (Bank, 2007;

Baumgard and Rhoads, 2013; Ross et al., 2017). Notably, the economic effect of HS is likely underestimated as the approximation does not include the effects of intrauterine programming due to gestational HS (Johnson et al., 2013; Boddicker et al., 2014;

Johnson et al., 2015a; Johnson et al., 2015b). Swine in particular have a limited capacity to dissipate heat effectively as they possess nonfunctional sweat glands under normal physiological conditions, which impedes evaporative heat loss (Ingram, 1967; D'Allaire et al., 1996). Moreover, genetic selection emphasizing rapid lean tissue accretion is 2 accompanied by increased endogenous heat production; a thermal scenario that exacerbates their innate inability to tolerate heat (Bianca, 1976; Renaudeau et al., 2011).

Even with modern advancements in nutritional strategies and management practices, HS remains a major seasonal challenge for the livestock sector as a whole. Additionally, the effects of HS may become increasingly severe if climate patterns continue to change as predicted (Hoffmann, 2010).

There is a need to better understand the biological responses underlying production losses in response to HS. Predicted increases in swine populations, at least in some parts of the world (Seo and McCarl, 2011), would ostensibly compensate for production losses under hotter conditions while also meeting the demand for a larger human population. However, at some point, increased livestock numbers may prove to be unsustainable as they would compete for more resources, such as land, water, grain usage, etc. (Godfray et al., 2010). Therefore, in order to minimize production losses due to the abiotic stress and strive for improved production efficiency, it is critical to expand our understanding of the effects of HS on whole animal physiology as well as the underlying molecular mechanisms.

Physiological Responses to HS

Heat stress occurs when the animal’s body temperature increases (through both intrinsic and extrinsic sources of thermal energy) and body heat dissipated to the environment is less than required to maintain euthermia (Bianca, 1968, 1976). From an agricultural perspective, heat-stressed livestock ultimately produce less and are thus obviously less efficient and profitable (Baumgard and Rhoads, 2013). 3

Reduced feed intake is a cardinal feature of HS, which ostensibly occurs to limit metabolic heat production (Bianca, 1976; Blaxter, 1989; Collin et al., 2001; Baumgard and Rhoads, 2013). Considering the consequential decrease in body weight, heat-induced hypophagia has traditionally been attributed as the major factor underlying impaired productivity in pigs (Collin et al., 2001; West, 2003; DeShazer et al., 2009). However, body weight is not as markedly reduced in heat-stressed pigs as compared with pair-fed thermoneutral controls, despite both being on the same plane of nutrition (Pearce et al.,

2013a). Dairy cows and chickens, on the other hand, have more pronounced decreases in milk production and body weights, respectively, compared with pair-fed animals (Geraert et al., 1996; Rhoads et al., 2009; Wheelock et al., 2010). Regardless of these differences,

HS has significant physiological ramifications independent of feed intake.

Interestingly, despite HS-induced hypophagia, heat-stressed pigs and dairy cows actually have reduced circulating nonesterified fatty acids (NEFA) and undergo increased adipose tissue accretion compared with thermoneutral pair-fed counterparts (Rhoads et al., 2009; Wheelock et al., 2010; Pearce et al., 2013a; Sanz Fernandez et al., 2015b).

Simultaneously, there is an increased circulating markers of muscle catabolism (i.e., creatine and Nτ-methylhistidine), corresponding with an overall shift in body composition

(i.e., less lean and more adipose tissue) that is well-documented in pigs, rodents, and chickens (Close et al., 1971; Katsumata et al., 1990; Ain Baziz et al., 1996; Geraert et al.,

1996; Collin et al., 2001; Pearce et al., 2013a). This phenomenon paradoxically contradicts what is energetically predicted as pigs on a restricted diet (i.e., decreased energy concentration) typically have a decreased carcass lipid-to-protein ratio (Oresanya et al., 2008). Moreover, heat-stressed cattle and pigs have increased concentrations of 4 lipolytic hormones (i.e., epinephrine and glucagon) in circulation (Alvarez and Johnson,

1973; de Braganca and Prunier, 1999). There is also an unexplainable increase in circulating insulin in heat-stressed pigs, which is a response conserved in mice, sheep, cows, and snakes (Wheelock et al., 2010; Morera et al., 2012; Pearce et al., 2013a;

Mahjoubi et al., 2014; Gangloff et al., 2016). Even with exogenous epinephrine, heat- stressed pigs and cows have blunted NEFA mobilization, which may be attributed to the antilipolytic and lipogenic nature of insulin (Baumgard et al., 2011; Sanz Fernandez et al., 2015a). Additionally, whereas increases in circulating glucose and insulin in heat- stressed pigs may be suggestive of insulin resistance, HS actually improves whole-body insulin sensitivity as determined by the hyperinsulinemic-euglycemic clamp in growing pigs (Sanz Fernandez et al., 2015b), which may be critical for reasons described below.

Heat stress also results in marked changes in glucose metabolism. In the liver, there is an overall increase in gluconeogenesis in rats (Collins et al., 1980), which is in agreement with an upregulation of pyruvate carboxylase (first step of gluconeogenesis) transcript abundance in bovine (Rhoads et al., 2011) and overall increased hepatic glucose output in humans during HS (Rowell et al., 1968; Angus et al., 2001).

Intriguingly, dietary supplementation of carbohydrates is unable to blunt the heat-induced increase in hepatic glucose production (Angus et al., 2001). Intracellular metabolism in muscle is altered during HS, consisting of enhanced glycolytic capacity as evidenced by increases in abundance or activity of glycolytic (phosphofructokinase, phosphoglycerate kinase, pyruvate kinase) and glycogenolytic enzymes (glycogen phosphorylase) in pigs

(Cruzen et al., 2015) and chickens (Zhang et al., 2012), which is consistent with overall increased glycogen usage (Fink et al., 1975; Febbraio et al., 1994). Pyruvate 5 dehydrogenase kinase 4 (negatively regulates pyruvate dehydrogenase) and lactate dehydrogenase A (converts pyruvate to lactate) transcript abundances are both increased in muscle during HS in rats (Sanders, 2010), which may suggest a reduced emphasis on oxidative phosphorylation via reduced entry of pyruvate into the TCA cycle. However, these molecular mechanisms may be dependent on muscle fiber type as the aforementioned markers are only upregulated in glycolytic muscle. In all, enhanced glycolysis, despite ostensible maintenance of blood flow to the tissue (i.e., aerobic glycolysis) (Nielsen et al., 1990), may contribute to hyperlactatemia during HS (Febbraio et al., 1994; Zhang et al., 2012).

These metabolic changes may occur to support an activated immune system during HS (Baumgard and Rhoads, 2013). This is critical as another physiological aspect of HS is impaired intestinal integrity due to the redistribution of blood supply to the periphery to maximize heat dissipation, which corresponds with limited perfusion to the splanchic bed (Hall et al., 1999). Consequently, the HS-induced hypoxic environment in the gut leads to epithelial damage and altered tight junction complexes (Dokladny et al.,

2006; Pearce et al., 2013b) and is accompanied by increased passage of gut pathogens and affiliated endotoxins into circulation (Hall et al., 2001; Pearce et al., 2013c; Pearce et al., 2014). Activated immune cells undergo a metabolic shift from oxidative phosphorylation to aerobic glycolysis and utilize copious amounts of glucose (Palsson-

McDermott and O'Neill, 2013). This metabolic shift increases the generation and export of lactate (consistent with HS-induced hyperlactatemia), which can in turn be used as a gluconeogenic precursor to help fuel the immune response (Maciver et al., 2008;

Baumgard and Rhoads, 2013; O'Neill et al., 2016). 6

Collectively, hyperthermia involves systemic physiological and metabolic changes that have direct and indirect consequences on the animal. This whole-animal response to HS is a critical concept as an organ may not only be responding to a thermal stimulus, but also receiving endocrine, metabolite, immunogenic, and other signals, resulting in altered function and activity. These alterations throughout culminate in changes to nutrient partitioning and overall body compositions in the pig.

Thermotolerance

Genetics of Thermotolerance

Livestock production efficiency has improved (markedly in recent years) due to multiple factors related to genetics, nutrition, and management (Thornton, 2010).

However, higher producing pigs (i.e., improved efficiency in growth and lean tissue accretion) have an increased basal heat level, which corresponds with an increased sensitivity to HS (Brown-Brandl et al., 2004; Renaudeau et al., 2011). Breeds with better reproductive capability also demonstrate HS sensitivity as Dutch Yorkshire sows have better reproductive performance (i.e., total born or farrowing rate after first insemination) compared to Large White sows during thermoneutral conditions, but Large White sows outperform Dutch Yorkshires when temperatures exceed 25°C (Bloemhof et al., 2008).

Daughters of specific sires also show similar patterns with respect to total born or farrowing rate when considering increasing temperature on the day of insemination or 21 days before the first insemination, respectively. One daughter group outperforms the other for both variables, depending on the temperature conditions (Bloemhof et al.,

2013). High-producing dairy cows and sheep have similar effects as improved milk 7

production traits correspond with increased sensitivity to HS (Ravagnolo and Misztal,

2000; Bohmanova et al., 2005; Finocchiaro et al., 2005).

In some livestock species, a single gene can improve thermoregulation, as is the case with the “naked-neck” and “slick” genes in chickens and cattle, respectively

(Cahaner et al., 2008; Dikmen et al., 2014). While a key gene conferring enhanced thermoregulation has not been discovered in pigs, certain breeds appear to be more phenotypically tolerant to HS. For example, the Creole breed possesses several skin histological differences compared to the more conventional Large White breed, which may contribute to improved heat dissipation and reduced rectal temperature and respiration rate during HS (Renaudeau, 2005; Renaudeau et al., 2006b). However, increased thermoregulatory capacity during HS could also be partly due to the Creole pig being less productive (i.e., slower growth rate, less lean and more adipose tissue) than the

Large White (Renaudeau, 2005; Renaudeau et al., 2006a; Renaudeau and Mourot, 2007).

Regardless, genetic selection for some valued traits (e.g., lean tissue accretion rates) may be possible under HS conditions due to increased heritability compared to thermoneutral conditions in pigs (Zumbach et al., 2008). Interestingly, single nucleotide polymorphisms (SNPs) have been identified that account for variation in thermoregulatory responses during HS (Kim et al., 2018). Although SNPs at a single gene have been shown to improve thermoregulation in dairy cows, it may not necessarily confer improved production per se as reduced rectal temperature during HS dairy cows is not accompanied by changes in milk yield (Xiong et al., 2013). However, genome-wide association studies have identified distinct SNPs associated with thermoregulatory or production variables separately in dairy cows (Dikmen et al., 2013). Furthermore, dairy 8 cows predicted to be tolerant or susceptible to HS using genomic data have corresponding differences in rectal temperature and milk yield (Garner et al., 2016).

Thus, it may be that a “tolerant” phenotype consists in both thermoregulation and production-related variables. Specific SNPs located at heat shock protein loci may support this notion (described below).

Importantly, identifying genetic loci that are associated not only with production- related variables, but also with thermoregulation may be critical for enhancing HS tolerance. While previous studies have investigated tolerance with respect to minimizing the effects of HS on a single production characteristic (Bloemhof et al., 2008; Zumbach et al., 2008), it may be critical to avoid selecting a single trait during HS to avoid inadvertent effects in the future. Taking into account multiple components of the animal

(perhaps unrelated to thermal stimuli) may be key to generating more robustness in the future that can respond to and maintain productivity when exposed to a variety of conditions (Friggens et al., 2017).

Acquired Thermotolerance

Acquired thermotolerance (a.k.a. programmed thermotolerance) has been demonstrated through thermal acclimation or thermal conditioning, which differ in the duration of thermal exposure. Thermal acclimation involves long-term periods

(Horowitz, 2014) while thermal conditioning consists of relatively brief stints of heat

(Yahav and Hurwitz, 1996; Yahav and McMurtry, 2001; Kisliouk et al., 2009). Although few studies have investigated the biological responses of each in agricultural species, they provide compelling evidence for their potential impact on subsequent heat challenges. 9

In chickens, thermal conditioning on day 5 post hatching results in reduced circulating T3 hormone (major regulator of metabolic rate), cloacal temperature, and mortality compared to thermally naïve controls at 42 d of age (Yahav and Hurwitz, 1996;

Yahav and McMurtry, 2001), which is partially attributed to epigenetic mechanisms in the preoptic area of the hypothalamus (Kisliouk et al., 2009). Physically conditioned sheep via treadmill training improves blood flow to the ileum, which also corresponds with reduced rectal temperatures (Sakurada and Hales, 1998). Thus, maintaining gastrointestinal blood flow may be critical for the onset of acquired thermotolerance via limiting the aforementioned HS-induced leaky gut. Additionally, heat-acclimated animals have improved splanchnic blood flow without restricting vasodilatory mechanisms to the periphery, which may allow for more effective heat dissipation (Horowitz, 2014) and perhaps also limit the heat-induced hypoxic effects on the gut (Pearce et al., 2013c).

Heat Shock Proteins and Cellular Thermotolerance

In general, heat shock proteins (HSPs) are stress responsive proteins functioning to maintain cellular homeostasis. Discovery of HSPs was initially associated with their involvement in the heat shock response (HSR). Early work on the HSR consisted of cytological evidence of heat-induced “puffs” at certain loci in the salivary gland polytene chromosomes of Drosophila (Ritossa, 1962). Eventually, using radiolabeled amino acids, researchers were able to observe the synthesis of key proteins during the HSR and the naming of these proteins corresponded to their molecular weights (kDa) (Tissieres et al.,

1974; McKenzie et al., 1975; Lindquist, 1980). Over time, much like multiple aliases for other genes, updated nomenclature has been developed for a more precise system to distinguish between HSPs (Kampinga et al., 2009). For example, it is now known that 10 there is a family of HSP70 proteins consisting of multiple unique genes and isoforms

(Radons, 2016) and that they respond to multiples stressors (i.e. not just heat; (Rafiee et al., 2006; El Golli-Bennour and Bacha, 2011).

Heat shock factor 1 (HSF1) is a major transcription factor involved in the cellular

HSR. During the stress response, HSF1 self-oligomerizes and then translocates to the nucleus, where it binds to DNA at heat shock elements within the regulatory regions of the HSP genes, ultimately driving their transcription (Sarge et al., 1993; Westwood and

Wu, 1993; Fulda et al., 2010; Tetievsky and Horowitz, 2010). Heat shock protein 90

(HSP90) can interact with HSF1 and also prevent its oligomerization and DNA-binding ability, implicating HSP90 as a negative regulator of the transcription factor (Ali et al.,

1998; Zou et al., 1998). More recent studies have shown EP300, an acetyltransferase, as another regulator of HSF1 as it prevents degradation and promotes chromatin interactions via acetylation of certain residues in the transcription factor (Raychaudhuri et al., 2014).

The involvement of HSPs during cellular stress responses is well-known. For example, heat shock 70 kDa protein 1A (HSPA1A) and HSP beta-1 (HSPB1) are two stress-inducible proteins that aid in protein refolding, the latter specifically involved in cytoskeleton maintenance (Bakthisaran et al., 2015; Seo et al., 2016). Interestingly, in addition to its regulating HSP transcription, HSF1 has been directly implicated in the maintenance of the cytoskeleton as well, corresponding with thermotolerance in nematodes, characterize by decreased mortality and increased motility (Baird et al.,

2014). There are also expanding roles of HSPs outside of stress response pathways. In the addition to refolding proteins in the mitochondria, HSP family D (HSP60) member 1

(HSPD1) is involved in protein import for the organelle, a process dependent on HSP10 11 and ATP availability (Martin et al., 1992; Levy-Rimler et al., 2001). Heat shock protein

90 is known to interact with multiple proteins and is involved in many functions, including reproductive and immune-related pathways (Dhamad et al., 2016; Schopf et al.,

2017). Immune-related functions have also been linked to members of the HSP70 family

(Dokladny et al., 2010; Afrazi et al., 2012).

Many studies have determined that SNPs within many HSP genes corresponds with improved thermotolerance (i.e., improved thermoregulatory and/or production variables) in many livestock species (Marcos-Carcavilla et al., 2010; Basirico et al., 2011;

Li et al., 2011a; Li et al., 2011b; Charoensook et al., 2012; Chen et al., 2013; Deb et al.,

2013; Xiong et al., 2013; Kumar et al., 2015; Sajjanar et al., 2015; Kumar et al., 2016;

Verma et al., 2016; Badri et al., 2018). Interestingly, rather than direct impacts on the amino acid sequence of the HSPs, some of the SNPs confer increased transcript abundance for the particular gene, which can be the result of altered transcriptional regulation through an miRNA-mediated mechanism (Li et al., 2011b; Badri et al., 2018).

Collectively, while important for maintaining homeostasis at the cellular level, HSPs may be integral in influencing the overall sensitivity to HS in the whole animal.

Developmental Programming and Epigenetics

Developmental Programming and Effects in Postnatal Life

Developmental programming refers to the critical interactions occurring between the embryo or the fetus and the maternal environment, which has a direct impact on the progeny’s postnatal life. This concept is commonly referred to as the Barker hypothesis and, as reviewed by Barker (2007), initially emerged from epidemiological research relating high rates of infant mortality and adult deaths caused by coronary heart disease in 12

England and Wales (Barker and Osmond, 1986; Barker et al., 1989). In general, those initial publications helped generate the “developmental origins of health and disease” hypothesis and has spurred a wave of research connecting environmental influences in utero to postnatal diseases. More specifically, chronic exposure to environmental factors, such as nutritional status or temperature, during intrauterine development can culminate into altered future phenotypes (i.e., physiological, metabolic, or body composition; Chen et al., 2010; Boddicker et al., 2014; Johnson et al., 2015a).

Many studies have investigated environmental influence on intrauterine development and subsequent postnatal effects. Perhaps one of the most notable examples of maternal environment influence on development was demonstrated with nutritional effects on the offspring born during the Dutch Famine. The Nazi war strategy that banned food transportation resulted in different physiological and metabolic phenotypes, such as glucose intolerance and/or obesity, depending on the timing of nutrient deprivation during gestation (Roseboom et al., 2006); those exposed to the nutritional challenge during the first trimester exhibited more adverse consequences (Figure 1.1). Elevated stress hormones accompany malnutrition during maternal starvation (Rooij, 2011) and it has been shown that hyperglucocorticoidemia during intrauterine development can result in lower birth weights, increased risk for obesity later in life (Seckl, 2004; Entringer et al., 2012) and impaired immune function (Cavigelli and Chaudhry, 2012).

Figure 1.1. Timeline and effects of the Dutch famine birth cohort comprised of 2414 people. Various health effects of those exposed to the famine depending on timing of exposure to undernourishment during gestation. Figure adapted from Roseboom et al., 2006.

Studies modeling intrauterine growth restriction (IUGR) show similar obese and glucose intolerance phenotypes as well as hyperinsulinemia (Simmons et al., 2001).

Interestingly, ovine research has demonstrated obesity programming can occur prior to gestational development as nutritional deprivation during the periconceptional period results in an increased fat-to-lean body mass phenotype in males postnatally (Begum et al., 2013). Other research suggests that IUGR can lead to decreased skeletal muscle mass in postnatal life as well, leading to serious farm-animal production consequences

(Foxcroft et al., 2006, 2009; Tse et al., 2008). An overnourished state during gestation also has postnatal effects on the offspring. Progeny of obese pregnant rats exhibit 14 increased insulin and leptin levels as well as increased susceptibility of obesity later in life (Shankar et al., 2008). Early developmental programming in this manner may even have potential effects during female gametogenesis as morphological and metabolic traits have been altered in oocytes derived from overweight females (Leary et al., 2014).

Heat Stress Effects on the Developing Offspring

Many studies have detailed the teratogenic effects of HS exposure during gestation in many species, including humans and pigs (Graham and Edwards, 1998).

Aside from the neuronal-related teratogenic defects (Shiota and Kayamura, 1989;

Graham and Edwards, 1998), it has also been shown that hyperthermia during pregnancy in ewes results in reduced blood flow to the placenta by as much as 30% (Dreiling et al.,

1991). This heat induced “placental insufficiency” corresponds with a decrease in oxygenation and nutrition to the developing offspring, resulting in lower birth weights

(Galan et al., 1999). In this same model, adrenergic receptor desensitization in utero results in a blunted non-esterified fatty acid (NEFA) response to epinephrine challenges and future obesity (Chen et al., 2010). Other effects, such as altered behavior, learning, and postnatal growth have been demonstrated in mice exposed to HS during days 12-15 of gestation (Shiota and Kayamura, 1989). Similarly, gestational thermal insults in pigs alters body composition by increasing adipose deposition and decreasing muscle tissue accretion (Boddicker et al., 2014; Johnson et al., 2015a). These observations, coupled with increased average core body temperatures from in utero thermal exposure (Johnson 15 et al., 2013) indicate altered and inefficient use of dietary energy with respect to production characteristics.

Types of Epigenetic Marks and Effects on Transcriptional Regulation

Aside from immune cells, it is remarkable that all cell types comprising an adult mammalian organism are derived from a single zygotic cell and, therefore, contain the same genetic information. The molecular mechanisms acting upon chromatin are associated with its dynamic packaging processes and allow for critical events during development, such as cellular differentiation, maintenance of function, organogenesis, and ultimately the generation of a viable organism (Reik, 2007). Controlling the condensation or accessibility of chromatin at certain regions allows the cell to regulate gene expression at the transcriptional level, eventually resulting in cell-specific transcriptional profiles that confer morphology, function, and, ultimately, cell type (Reik,

2007; Smith and Meissner, 2013). Euchromatic or heterochromatic genomic loci are transcriptionally active or inactive regions, respectively, and are determined by the types of chemical modifications or epigenetic marks that occur to the chromatin structure

(Cosgrove and Wolberger, 2005). Two major types of epigenetic marks affecting chromatin accessibility are DNA methylation and posttranslational modifications that occur to the histone N terminal tails.

DNA methylation mainly occurs through the transfer of a methyl group to the 5th carbon position of a cytosine (commonly referred to as 5-methylcytosine or 5mC) in a cytosine-guanine dinucleotide context. DNA methyltrasnferases (DNMTs) carry out these methylation events and certain enzymes belonging to this group have been associated with specific roles, either in de novo (Dnmt3a and Dnmt3b) or maintenance 16

(Dnmt1) of methylation marks (Perera and Herbstman, 2011; Hackett and Surani, 2013;

Smith and Meissner, 2013). Although DNA methylation is traditionally associated with transcriptional downregulation of the respective gene (Weaver et al., 2004; Dolinoy et al.,

2006; Coolen et al., 2010; Mamrut et al., 2013), the genomic context actually determines the effect of the epigenetic mark on transcriptional activity (Gilsbach et al., 2014;

Wagner et al., 2014; Yang et al., 2014). Imprinted genes, for example, are expressed in a parent-of-origin manner, meaning that a particular gene may be expressed paternally, but not maternally, or vice versa (Peters, 2014). At the IGF2-H19 locus, an imprinting control region (ICR) containing multiple CpGs regulates the transcriptional expression of the IGF2 and H19 genes. On the maternal chromosome, the ICR is methylated and results in transcriptional silencing and activation of IGF2 and H19, respectively. The same ICR is unmethylated paternally, which results in the opposite transcriptional effects for the two imprinted genes (Engel et al., 2004).

Further exemplifying the complexity on transcriptional regulation, genomic studies have revealed the methylation status of CpGs within a promoter element (Weaver et al., 2004; Dolinoy et al., 2006; Mamrut et al., 2013) or gene body (Wagner et al., 2014;

Yang et al., 2014) can result in transcriptional downregulation and upregulation, respectively. Within the promoter specifically, the actual number of CpGs or density within the regulatory element may also play a role as to whether DNA methylation status of a promoter element confers transcriptional regulation (Vinson and Chatterjee, 2012).

There are cases when DNA methylation occurs at silencer elements, preventing repressive proteins from binding and ultimately resulting in transcriptional upregulation

(Eden et al., 2001). Interestingly, DNA methylation regulatory ability may even have 17 tissue-specific effects on transcription for the same gene as shown in mice, wherein the methylation status of CpGs in the oxytocin receptor gene promoter is associated with positive or negative transcript abundance of oxytocin receptor in the mammary glands or uterus, respectively (Mamrut et al., 2013).

In a developmental context, DNA methylation signatures are anything but static and also involve expression and localization differences of DNA methyltransferases in the gamete and the subsequent embryo (Rivera and Ross, 2013). Interestingly, there are two important erasure events (Perera and Herbstman, 2011), occurring at primordial germ cell (PGC) development and postfertilization (Figure 1.2). During PGC migration in male and female mice, there are reductions in the degree of DNA methylation at imprinted and non-imprinted genes, but not at transposable elements, such as Line1; the retention of methylation at the latter regions may serve to prevent the expression and activity of transposable elements during a stage where global DNA methylation is decreased (Hajkova et al., 2002). The second erasure occurs following fertilization and then remethylation events occur from the blastocyst stage when the inner cell mass is present and onward (Smith et al., 2012). Interestingly, some studies have indicated that specific genomic regions may undergo differential methylation between maternal and paternal contributions to the embryo and, thus undergo different DNA methylation dynamics postfertilization (Tomizawa et al., 2011).

Figure 1.2. A summary of DNA methylation dynamics during development. Figure adapted from Perera and Herbstman, 2011.

Nucleosomes are comprised of a core histone octamer, in which four different histone proteins, H2A, H2B, H3, and H4 are represented twice in the DNA-protein complex (Luger et al., 1997). Typically referred to as “beads-on-a-string”, 146 base pairs of the DNA “string” actually wrap around the protein complex, representing first-order packaging of the DNA molecule (Luger et al., 1997). Other histone variants called linker histones, such as H1, promote further packaging of DNA via internucleosomal linkages

(Hayakawa et al., 2012). Posttranslational modifications (PTMs) can be catalyzed to the functional groups of the amino acid residues at the N terminal regions of the histone tails

(Turner, 2002; Cosgrove and Wolberger, 2005; Tan et al., 2011). These chemical alterations influence the binding of chromatin associating factors and the interactions made between the DNA molecule and the histone complex, which can directly affect the accessibility of the chromatin (Cosgrove and Wolberger, 2005). Although acetylation and methylation represent classic covalent additions to histone residues, numerous types of 19 chemical modifications have been reported (Turner, 2002; Margueron et al., 2005; Tan et al., 2011) and new chemical moieties, such as crotonylation, are still being discovered

(Tan et al., 2011). These histone PTMs have been implicated in regulation of gene expression at the transcriptional level and some suggest that combinations of certain chemical and positional covalent modifications constitute a “histone code” for regulation

(Turner, 2002). During early mammalian development, histone modifiers, such as those belonging to the repressive polycomb group of proteins, are needed to establish and maintain repression of certain genomic loci for successful differentiation and development (Reik, 2007; Schuettengruber et al., 2007). Linker H1 variants have also been implicated in controlling the cellular differentiation process during early stages of embryonic development (Hayakawa et al., 2012).

In many cases, these two types of epigenetic marks work together to impact the accessibility (the degree of chromatin packaging) at many genomic loci. Studies have demonstrated that DNA methylation can mediate the activity of histone modifiers. For instance, the targeting ability of histone deacetylase 1 and 2 is impaired in DNMT1 knock out human cancer cells, resulting in globally increased acetylation and decreased methylation modifications to H3K9, and the wildtype phenotype can be rescued upon transfection of murine DNMT1 (Espada et al., 2004). DNMT1 can even make direct interactions with histone modifiers, such as G9a (an enzyme that catalyzes mono- and dimethylation at H3K9), and RNAi mediated knockdown of DNMT results in reduction of H3K9me2 levels (Esteve et al., 2006). Conversely, histone modifiers can direct the activity of DNMTs. For example, RNAi mediated knockdown of EZH2, a polycomb group protein that catalyzes H3K27 methylation (also makes direct binding ability with 20 all DNMTs 1, 3A, and 3B), inhibits CpG methylation at the Myt1 promoter, resulting in transcriptional upregulation of the gene (Vire et al., 2006). In addition, double knockouts for Suv39h1 and -2, which are both histone methyltransferases involved in H3K9 trimethylation, results in altered methylated CpG patterns at major satellite repeats in pericentric regions of mouse embryonic stem cells (Lehnertz et al., 2003).

Importantly, the coordination of both types of epigenetic marks to promote the correct expression of pluripotency-associated genes and more differentiated-related genes, in a spatial and temporal manner, is key to allow for successful development

(Figure 1.3; Reik, 2007). Environmental factors, like the ones already discussed, can have a major impact on the reestablishment of epigenetic patterns, which can result in the programming of phenotypic limitations (e.g., reduced lean tissue accretion) later in postnatal development. This would especially be critical to more pluripotent cells during earlier stages of development, as a single epigenetic change during early embryonic development may be maintained and amplified through proliferation from that original cell. All of the cells derived from that single cell may be significantly altered and could lead to ineffective organ function or poorly producing livestock.

Figure 1.3. DNA methylation and histone modification during development

. Gestation marks a sensitive time for a newly developing organism as it undergoes dynamic gene expression patterns that ultimately result in cellular differentiation and tissue formation. This requires sophisticated spatial and temporal orchestration of epigenetic marks, DNA methylation and histone modifications (histone 3 [H3] lysine residue 27 [K27] and H3 lysine residue 4 [K4]) to regulate genes associated with a pluripotent or differentiated state, depending on the developmental stage. In addition to the immediate embryo, primordial germ cells (PGCs) development also consists of changes to the epigenetic landscape in order to result in functional gametes for future reproduction. Alterations to epigenetic signatures from environmental factors can be programmed in the developing offspring and PGCs, resulting in altered trait phenotypes with the immediate offspring and, potentially, subsequent generations. Figure adapted from Reik, 2007.

Epigenetic Events in Developmental Programming

As mentioned, the maternal environment can have a major influence on the epigenome during in utero development of the offspring, leading to altered phenotypes in postnatal life (Reik, 2007; Perera and Herbstman, 2011). One prime example 22 demonstrating extrinsic factor influence on epigenetic establishment has been conducted in mouse models (Dolinoy et al., 2006). The agouti gene is a signal that causes a shift in black pigment production to yellow pigment production in mice. In normal mice, there is brief expression of the wildtype agouti allele (A), which results in a brown coat pigment phenotype. Mice that contain an insertion from the retrotransposon Intracisternal Particle

A (IPA) within the agouti locus (Avy) exhibit a yellow coat phenotype because agouti is overexpressed due to a “cryptic promoter” within the IPA insertion. There is also an allele that codes for only black pigment production and is referred to the nonagouti allele

(a). A normal cross between homozygotic a/a females with heterozygotic Avy/a males results in a variety of different coat phenotypes (including yellow, a mix between yellow and brown referred to as mottled, and a “pseudoagouti” or brown appearance) for the heterozygous offspring. However, it was demonstrated that dietary supplementation of genistein, an estrogen agonist, during gestation resulted in increased methylation status of the IPA insertion sequence. This epigenetic alteration caused a shift toward more pseudoagouti coat phenotypes, which was also accompanied by a reduction in the frequency of the obesity phenotype, typically associated with the heterozygotes (Dolinoy et al., 2006).

Epigenetic changes maintained in postnatal life have also been investigated in maternal malnourishment and IUGR models. There is a sustained upregulation of hypothalamic glucocorticoid receptor in 5-year-old male ewes subjected to periconceptional nutrient deprivation, which is linked to their obese phenotype as well.

This upregulation was accompanied by promoter hypomethylation, an increase in H3K9 acetylation, and a decrease in H3K27 trimethylation (the latter two being markers for 23 transcriptional upregulation and downregulation, respectively) (Begum et al., 2013).

Similar modifications have been implicated in the reduced expression of pancreatic and duodenal homeobox 1 (Pdx1) and glucose transporter type 4 (GLUT4) in b cells and muscle, respectively, contributing to symptoms and onset of type II diabetes in IUGR rodents (Pinney and Simmons, 2010). The promoters of each gene should contain markers for transcriptional upregulation, such as histone 3 acetylation, but undergo extensive epigenetic restructuring and exhibit markers for a more condensed chromatin state, such as H3K9 dimethylation at both promoters. Directly knocking out JHDM2a, an

H3K9 demethylase, has even been demonstrated to result in metabolic instability (i.e. elevated levels of insulin, triglycerides, leptin) and obese phenotypes due to the dysregulation of genes involved in metabolism and energy balance (Inagaki et al., 2009).

Clearly, the importance of correct epigenetic establishment and maintenance is crucial in order to prevent altered trait phenotypes, including metabolism, later in postnatal life.

Heat Stress Influence on Epigenetic Marks

Both DNA methylation and histone modifications are influenced by thermal stimuli as exemplified by organisms belonging to the Animalia (Desrosiers and Tanguay,

1988; Thomson et al., 2004; Zhu et al., 2008; Kisliouk et al., 2009; Tetievsky and

Horowitz, 2010; Seong et al., 2011; Norouzitallab et al., 2014; Hao et al., 2016), Fungi

(Shivaswamy and Iyer, 2008), and Plantae (Pecinka et al., 2010) kingdoms, underscoring the importance of conserved epigenetic machinery enabling environmental responses

(Table 1.1). Interestingly, epigenetic mechanisms have specifically been implicated in both thermal conditioning and acclimation. Short-term neonatal thermal conditioning in chickens results in global alterations to H3K9 acetylation and H3K9 dimethylation in the 24 preoptic area of the hypothalamus, which partly functions as the thermoregulation center of the brain. Importantly, these histone alterations also corresponded with an increased mRNA abundance of a translational initiation factor (Eif2B5), which has been previously implicated in the thermally conditioned phenotype (Kisliouk et al., 2009). Thermal acclimation confers the maintenance of acetylation modifications at H4 even after a 30- day withdrawal period, corresponding with increased HSP gene expression and cytoprotection in cardiac muscle during subsequent periods of HS in rats (Tetievsky and

Horowitz, 2010). Additionally, in Artemia species, acquired thermotolerance (i.e., higher survival to lethal HS) established in the parental generation is maintained transgenerationally, which is accompanied by globally altered DNA methylation, acetylation modifications at H3 and H4, and increased HSP70 protein abundance

(Norouzitallab et al., 2014).

Heat-induced alterations to the epigenetic landscape likely explain the different phenotypes observed in pigs exposed to gestational HS (Boddicker et al., 2014; Johnson et al., 2015a). Interestingly, the DNA methylation status of multiple genomic loci corresponds with adipose and muscle composition differences for breeds, sexes, and anatomical locations (Li et al., 2012). Some of these regions could be susceptible to HS- induced epigenetic changes as specific genomic loci are differentially methylated in response to HS in porcine longissimus dorsi muscle (Hao et al., 2016). Furthermore, paternally expressed quantitative trait loci that are involved with growth and development in pigs (Thomsen et al., 2004) may be susceptible to thermal insults as DNA methylation at some paternally imprinted genes is decreased in response to HS in murine blastocysts

(Zhu et al., 2008). Collectively, the epigenome can respond to thermal stimuli and result 25 in altered trait phenotypes, likely representing an underlying mechanism by which prenatal heat exposure affects body compositions (Johnson et al., 2015a) and metabolism

(Boddicker et al., 2014) postnatally in pigs.

Table 1.1. The effects of heat stress on epigenetic modifications in various species Species Tissue Modificationa Outcomeb Locusc Reference Drosophila Cell line H2Bme ↑ - 1

H3Kme3 ↓ - H4Kme3 ↓ - Mouse Cell line H4ac ↑ Hsp70 2 S. - H4ac ↑ 248 locid 3 cerevisiae H4ac ↓ 295 locie Mouse Blastocysts DNAme ↓ Igf2r 4 DNAme ↓ H19 DNAme = Peg1 DNAme = Peg3 Chicken Hypothalamus H3K9ac ↑ GW 5 H3K9me2 ↑ GW H3K9ac ↑ EIF2B5 H3K9me2 ↓ EIF2B5 A. thaliana - DNAme = TSI 6 DNAme = GUS DNAme ↓ COPIA78 H3K9me2 ↓ TSI, COPIA78 H3K4me3 = TSI H3K4me3 ↓ COPIA78 Rat Cardiac H3K9, K18, and = HSP70 HSE 7 muscle K23 ac H4K5, K8, K12, ↑ HSP70 & - and K16 ac 90 HSE H3S10phos ↑ HSP70 & - 90 HSE Drosophila Embryos and Heterochromatin ≠ GW 8 salivary gland Chicken Liver, brain, DNAme = HSP70 9 leg muscle promoter

Table 1.1 continued Species Tissue Modificationa Outcomeb Locusc Reference A. thaliana Progeny of DNAme ≠ GW 10 heat-stress H3K9ac, ↑, ↑ HSFA2 plants H3K9me2 gene body H3K9ac, =, = SUVH2 H3K9me2 gene body H3K9ac, ↓, = SUVH2 H3K9me2 promoter H3K9ac, =, = SUVH5 H3K9me2 gene body Artemia Whole DNAme, H3ac, ≠ GW 11 juveniles H4ac A. thaliana Seedlings H3K9ac, ↑ HSP22.0 & 12 H3K4me2, -me3 HSP70 Pig Longissimus DNAme ≠ GW 13 dorsi Chicken Brain DNAme = HSP90α, - 14 (embryos) 90β, -70 ≠ (42 d HSP90α, - chicks) 90β, -70 aModification: me = methylation (followed by: 2 = dimethylation; 3 = trimethylation); ac = acetylation; phos = phosphorylation; H2B = histone 2B; H3 = histone 3; H4 = histone 4; K = lysine (followed by numerical amino acid position in the protein); S = serine bOutcome: ↑ = increased; ↓ = decreased; = = unaffected; ≠ = altered cLocus: specific genes are listed; GW = genome wide; HSE = heat shock element d248 upregulated heat shock responsive genes e295 downregulated heat shock responsive genes 1 Desrosiers and Tanguay (1988) 2 Thomson et al. (2004) 3 Shivaswamy and Iyer (2008) 4 Zhu et al. (2008) 5 Kisliouk et al., (2009) 6 Pecinka et al. (2010) 7 Tetievsky and Horowitz (2010) 8 Seong et al. (2011) 9 Gan et al. (2013) 10 Migicovsky et al. (2014) 11 Norouzitallab et al., (2014) 12 Lamke et al. (2016) 13 Hao et al. (2016) 14 Vinoth et al. (2018) 27

Summary

Heat stress negatively impacts many economically important traits and represents a major economic burden to the livestock industry. Production efficiency in pigs and other important agriculture-related species may be further compromised if climate patterns change as predicted. Thus, it is crucial to further our understanding of the biological responses to the abiotic factor. Identifying heat tolerant pigs would be valuable, thus study 1 (chapter 2) was conducted to assess the relationship between thermoregulatory and production responses to HS. Although pigs accumulate more adipose tissue during HS, the response of adipose tissue characteristics pertaining to quality (i.e., fatty acid composition and adipocyte size) are ill-defined and could be influenced by HS-induced hyperinsulinemia; study 2 (chapter 3) investigated insulin’s potential role in affecting carcass fat quality during HS. Heat stress can influence a variety of aspects relating to female fertility, but whether this is due to direct or indirect effects of HS on reproductive processes is not known. Additionally, our knowledge of how ovarian heat shock proteins respond to stressors, such as HS or LPS, is insufficient.

Study 3 (chapter 4) evaluated heat shock protein responses to both HS and LPS exposures. Many HS-induced phenotypes in pigs (e.g., altered body composition and hyperinsulinemia) can be programmed gestationally before postnatal development.

Therefore, the last objective of this dissertation was aimed at evaluating possible epigenetic mechanisms and protein responses as a result of thermal exposure during intrauterine development. 28

References

Afrazi, A., C. P. Sodhi, M. Good, H. Jia, R. Siggers, I. Yazji, C. Ma, M. D. Neal, T. Prindle, Z. S. Grant, M. F. Branca, J. Ozolek, E. B. Chang, and D. J. Hackam. 2012. Intracellular heat shock protein-70 negatively regulates TLR4 signaling in the newborn intestinal epithelium. J Immunol 188: 4543-4557.

Ain Baziz, H., P. A. Geraert, J. C. Padilha, and S. Guillaumin. 1996. Chronic heat exposure enhances fat deposition and modifies muscle and fat partition in broiler carcasses. Poultry Science 75: 505-513.

Alexandratos, N., and J. Bruinsma. 2012. World agriculture towards 2030/2050: the 2012 revision, Rome, FAO.

Ali, A., S. Bharadwaj, R. O'Carroll, and N. Ovsenek. 1998. HSP90 interacts with and regulates the activity of heat shock factor 1 in Xenopus oocytes. Molecular and Cellular Biology 18: 4949-4960.

Alvarez, M. B., and H. D. Johnson. 1973. Environmental heat exposure on cattle plasma catecholamine and glucocorticoids. J Dairy Sci 56: 189-194.

Angus, D. J., M. A. Febbraio, D. Lasini, and M. Hargreaves. 2001. Effect of carbohydrate ingestion on glucose kinetics during exercise in the heat. J Appl Physiol (1985) 90: 601-605.

Badri, T. M., K. L. Chen, M. A. Alsiddig, L. Li, Y. Cai, and G. L. Wang. 2018. Genetic polymorphism in Hsp90AA1 gene is associated with the thermotolerance in Chinese Holstein cows. Cell Stress & Chaperones 23: 639-651.

Baird, N. A., P. M. Douglas, M. S. Simic, A. R. Grant, J. J. Moresco, S. C. Wolff, J. R. Yates, 3rd, G. Manning, and A. Dillin. 2014. HSF-1-mediated cytoskeletal integrity determines thermotolerance and life span. Science 346: 360-363.

Bakthisaran, R., R. Tangirala, and M. Rao Ch. 2015. Small heat shock proteins: Role in cellular functions and pathology. Biochim Biophys Acta 1854: 291-319.

Barker, D. J. 2007. The origins of the developmental origins theory. J Intern Med 261: 412- 417. 29

Barker, D. J., C. Osmond. 1986. Infant mortality, childhood nutrition, and ischaemic heart disease in England and Wales. Lancet 1: 1077-1081.

Barker, D. J., P. D. Winter, C. Osmond, B. Margetts, S. J. Simmonds. 1989. Weight in infancy and death from ischaemic heart disease. Lancet 2: 577-580.

Basirico, L., P. Morera, V. Primi, N. Lacetera, A. Nardone, and U. Bernabucci. 2011. Cellular thermotolerance is associated with heat shock protein 70.1 genetic polymorphisms in Holstein lactating cows. Cell Stress & Chaperones 16: 441-448.

Baumgard, L. H., and R. P. Rhoads. 2013. Effects of Heat Stress on Postabsorptive Metabolism and Energetics. Annu Rev of Anim Biosci 1: 311-337.

Baumgard, L. H., J. B. Wheelock, S. R. Sanders, C. E. Moore, H. B. Green, M. R. Waldron, and R. P. Rhoads. 2011. Postabsorptive carbohydrate adaptations to heat stress and monensin supplementation in lactating Holstein cows. J Dairy Sci 94: 5620-5633.

Begum, G., A. Davies, A. Stevens, M. Oliver, A. Jaquiery, J. Challis, J. Harding, F. Bloomfield, and A. White. 2013. Maternal Undernutrition Programs Tissue-Specific Epigenetic Changes in the Glucocorticoid Receptor in Adult Offspring. Endocrinology 154: 4560-4569.

Bianca, W. 1968. Thermoregulation. In: E. S. E. Hafez (ed.) Adaptation of Domestic Animals. p 97-118. Lea and Febriger, Philadelphia.

Bianca, W. 1976. The signifiance of meterology in animal production. Int J Biometeorol 20: 139-156.

Blaxter, K. L. 1989. Energy metabolism in animals and man. Cambridge University Press, Cambridge, England.

Bloemhof, S., P. K. Mathur, E. F. Knol, and E. H. van der Waaij. 2013. Effect of daily environmental temperature on farrowing rate and total born in dam line sows. J Anim Sci 91: 2667-2679.

Bloemhof, S., E. H. van der Waaij, J. W. Merks, and E. F. Knol. 2008. Sow line differences in heat stress tolerance expressed in reproductive performance traits. J Anim Sci 86: 3330-3337. 30

Boddicker, R. L., J. T. Seibert, J. S. Johnson, S. C. Pearce, J. T. Selsby, N. K. Gabler, M. C. Lucy, T. J. Safranski, R. P. Rhoads, L. H. Baumgard, and J. W. Ross. 2014. Gestational Heat Stress Alters Postnatal Offspring Body Composition Indices and Metabolic Parameters in Pigs. PloS ONE 9: e110859.

Bohmanova, J., I. Misztal, S. Tsurauta, H. D. Norman, and T. J. Lawlor. 2005. National genetic evaluation of milk yield for heat tolerance of united state holsteins. Interbull Bulletin 33: 160-162.

Brown-Brandl, T. M., J. A. Nienaber, H. Xin, and R. S. Gates. 2004. A literature review of swine heat production. Trans. ASAE 47: 259-270.

Cahaner, A., J. A. Ajuh, M. Siegmund-Schultze, Y. Azoulay, S. Druyan, and A. V. Zarate. 2008. Effects of the genetically reduced feather coverage in naked neck and featherless broilers on their performance under hot conditions. Poult Sci 87: 2517- 2527.

Cavigelli, S. A., and H. S. Chaudhry. 2012. Social status, glucocorticoids, immune function, and health: can animal studies help us understand human socioeconomic-status- related health disparities? Hormones and Behavior 62: 295-313.

Charoensook, R., K. Gatphayak, A. R. Sharifi, C. Chaisongkram, B. Brenig, and C. Knorr. 2012. Polymorphisms in the bovine HSP90AB1 gene are associated with heat tolerance in Thai indigenous cattle. Trop Anim Health Prod 44: 921-928.

Chen, X., A. L. Fahy, A. S. Green, M. J. Anderson, R. P. Rhoads, and S. W. Limesand. 2010. 2- Adrenergic receptor desensitization in perirenal adipose tissue in fetuses and lambs w ith placentalinsufficiency-induced intrauterine growth restriction. J Physiol 588.

Chen, Z. Y., J. K. Gan, X. Xiao, L. Y. Jiang, X. Q. Zhang, and Q. B. Luo. 2013. The association of SNPs in Hsp90beta gene 5' flanking region with thermo tolerance traits and tissue mRNA expression in two chicken breeds. Mol Biol Rep 40: 5295-5306.

Close, W. H., L. E. Mount, and I. B. Start. 1971. The influence of environmental temperature and plane of nutrition on heat losses from groups of growing pigs. Anim Prod 13: 285-294.

Collin, A., J. van Milgen, S. Dubois, and J. Noblet. 2001. Effect of high temperature and feeding level on energy utilization in piglets. J Anim Sci 79: 1849-1857. 31

Collins, F. G., F. A. Mitros, and J. L. Skibba. 1980. Effect of palmitate on hepatic biosynthetic functions at hyperthermic temperatures. Metabolism 29: 524-531.

Coolen, M. W., C. Stirzaker, J. Z. Song, A. L. Statham, Z. Kassir, C. S. Moreno, A. N. Young, V. Varma, T. P. Speed, M. Cowley, P. Lacaze, W. Kaplan, M. D. Robinson, and S. J. Clark. 2010. Consolidation of the cancer genome into domains of repressive chromatin by long-range epigenetic silencing (LRES) reduces transcriptional plasticity. Nat Cell Biol 12: 235-246.

Cosgrove, M. S., and C. Wolberger. 2005. How does the histone code work? Biochemistry and Cell Biology 83: 468-476.

Cruzen, S. M., S. C. Pearce, L. H. Baumgard, N. K. Gabler, E. Huff-Lonergan, and S. M. Lonergan. 2015. Proteomic changes to the sarcoplasmic fraction of predominantly red or white muscle following acute heat stress. J Proteomics 128: 141-153.

D'Allaire, S., R. Drolet, and D. Brodeur. 1996. Sow mortality associated with high ambient temperatures. Can Vet J 37: 237-239. de Braganca, M. M., and A. Prunier. 1999. Effects of low feed intake and hot environment on plasma profiles of glucose, nonesterified fatty acids, insulin, glucagon, and IGF-I in lactating sows. Domest Anim Endocrinol 16: 89-101.

Deb, R., B. Sajjanar, U. Singh, S. Kumar, M. P. Brahmane, R. Singh, G. Sengar, and A. Sharma. 2013. Promoter variants at AP2 box region of Hsp70.1 affect thermal stress response and milk production traits in Frieswal cross bred cattle. Gene 532: 230-235.

DeShazer, J. A., G. L. Hahn, and H. Xin. 2009. Chapter 1: Basic principles of the thermal environment and livestock energetics. In: J. A. DeShazer (ed.) Livestock energetics and thermal environment management. p 181–209. Am. Soc. Agric. Biol. Eng., St. Joseph, MI.

Desrosiers, R., and R. M. Tanguay. 1988. Methylation of Drosophila histones at proline, lysine, and arginine residues during heat shock. J Biol Chem 263: 4686-4692.

Dhamad, A. E., Z. Zhou, J. Zhou, and Y. Du. 2016. Systematic Proteomic Identification of the Heat Shock Proteins (Hsp) that Interact with Estrogen Receptor Alpha (ERalpha) and Biochemical Characterization of the ERalpha-Hsp70 Interaction. PloS ONE 11: e0160312. 32

Dikmen, S., J. B. Cole, D. J. Null, and P. J. Hansen. 2013. Genome-wide association mapping for identification of quantitative trait loci for rectal temperature during heat stress in Holstein cattle. PloS ONE 8: e69202.

Dikmen, S., F. A. Khan, H. J. Huson, T. S. Sonstegard, J. I. Moss, G. E. Dahl, and P. J. Hansen. 2014. The SLICK hair locus derived from Senepol cattle confers thermotolerance to intensively managed lactating Holstein cows. J Dairy Sci 97: 5508-5520.

Dokladny, K., R. Lobb, W. Wharton, T. Y. Ma, and P. L. Moseley. 2010. LPS-induced cytokine levels are repressed by elevated expression of HSP70 in rats: possible role of NF-kappaB. Cell Stress & Chaperones 15: 153-163.

Dokladny, K., P. L. Moseley, and T. Y. Ma. 2006. Physiologically relevant increase in temperature causes an increase in intestinal epithelial tight junction permeability. Am J Physiol Gastrointest Liver Physiol 290: G204-212.

Dolinoy, D. C., J. R. Weidman, R. A. Waterland, and R. L. Jirtle. 2006. Maternal genistein alters coat color and protects Avy mouse offspring from obesity by modifying the fetal epigenome. Environ Health Perspect 114: 567-572.

Dreiling, C. E., F. S. Carman III, and D. E. Brown. 1991. Maternal Endocrine and Fetal Metabolic Responses to Heat Stress. J Dairy Sci 74: 312-327.

Eden, S., M. Constancia, T. Hashimshony, W. Dean, B. Goldstein, A. C. Johnson, I. Keshet, W. Reik, and H. Cedar. 2001. An upstream repressor element plays a role in Igf2 imprinting. The EMBO journal 20: 3518-3525.

El Golli-Bennour, E., and H. Bacha. 2011. Hsp70 expression as biomarkers of oxidative stress: mycotoxins' exploration. Toxicology 287: 1-7.

Engel, N., A. G. West, G. Felsenfeld, and M. S. Bartolomei. 2004. Antagonism between DNA hypermethylation and enhancer-blocking activity at the H19 DMD is uncovered by CpG mutations. Nature Genetics 36: 883-888.

Entringer, S., C. Buss, J. M. Swanson, D. M. Cooper, D. A. Wing, F. Waffarn, and P. D. Wadhwa. 2012. Fetal programming of body composition, obesity, and metabolic function: the role of intrauterine stress and stress biology. J Nutr Metab 2012: 632548. 33

Espada, J., E. Ballestar, M. F. Fraga, A. Villar-Garea, A. Juarranz, J. C. Stockert, K. D. Robertson, F. Fuks, and M. Esteller. 2004. Human DNA methyltransferase 1 is required for maintenance of the histone H3 modification pattern. J Biol Chem 279: 37175-37184.

Esteve, P. O., H. G. Chin, A. Smallwood, G. R. Feehery, O. Gangisetty, A. R. Karpf, M. F. Carey, and S. Pradhan. 2006. Direct interaction between DNMT1 and G9a coordinates DNA and histone methylation during replication. Genes & Development 20: 3089-3103.

FAO, IFAD, UNICEF, WFP, and WHO. 2017. The State of Food Security and Nutrition in the World 2017. Building resilience for peace and food security., Rome, FAO.

Febbraio, M. A., R. J. Snow, C. G. Stathis, M. Hargreaves, and M. F. Carey. 1994. Effect of heat stress on muscle energy metabolism during exercise. J Appl Physiol (1985) 77: 2827-2831.

Fink, W. J., D. L. Costill, and P. J. Van Handel. 1975. Leg muscle metabolism during exercise in the heat and cold. Eur J Appl Physiol Occup Physiol 34: 183-190.

Finocchiaro, R., J. B. van Kaam, B. Portolano, and I. Misztal. 2005. Effect of heat stress on production of Mediterranean dairy sheep. J Dairy Sci 88: 1855-1864.

Foxcroft, G. R., W. T. Dixon, M. K. Dyck, S. Novak, J. C. Harding, and F. C. Almeida. 2009. Prenatal programming of postnatal development in the pig. Soc Reprod Fertil Suppl 66: 213-231.

Foxcroft, G. R., W. T. Dixon, S. Novak, C. T. Putman, S. C. Town, and M. D. Vinsky. 2006. The biological basis for prenatal programming of postnatal performance in pigs. J Anim Sci 84 Suppl: E105-112.

Friggens, N. C., F. Blanc, D. P. Berry, and L. Puillet. 2017. Review: Deciphering animal robustness. A synthesis to facilitate its use in livestock breeding and management. Animal 11: 2237-2251.

Fulda, S., A. M. Gorman, O. Hori, and A. Samali. 2010. Cellular stress responses: cell survival and cell death. Int J Cell Biol 2010: 214074. 34

Galan, H. L., M. J. Hussey, A. Barbera, E. Ferrazzi, M. Chung, J. C. Hobbins, and F. C. Battaglia. 1999. Relationship of fetal growth to duration of heat stress in an ovine model of placental insufficiency. Am J Obstet Gynecol 180: 1278-1282.

Gan, J. K., D. X. Zhang, D. L. He, X. Q. Zhang, Z. Y. Chen, and Q. B. Luo. 2013. Promoter methylation negatively correlated with mRNA expression but not tissue differential expression after heat stress. Genet Mol Res 12: 809-819.

Gangloff, E. J., K. G. Holden, R. S. Telemeco, L. H. Baumgard, and A. M. Bronikowski. 2016. Hormonal and metabolic responses to upper temperature extremes in divergent life-history ecotypes of a garter snake. J Exp Biol 219: 2944-2954.

Garner, J. B., M. L. Douglas, S. R. Williams, W. J. Wales, L. C. Marett, T. T. Nguyen, C. M. Reich, and B. J. Hayes. 2016. Genomic Selection Improves Heat Tolerance in Dairy Cattle. Sci Rep 6: 34114.

Geraert, P. A., J. C. Padilha, and S. Guillaumin. 1996. Metabolic and endocrine changes induced by chronic heat exposure in broiler chickens: growth performance, body composition and energy retention. Br J Nutr 75: 195-204.

Gilsbach, R., S. Preissl, B. A. Gruning, T. Schnick, L. Burger, V. Benes, A. Wurch, U. Bonisch, S. Gunther, R. Backofen, B. K. Fleischmann, D. Schubeler, and L. Hein. 2014. Dynamic DNA methylation orchestrates cardiomyocyte development, maturation and disease. Nat Commun 5: 5288.

Godfray, H. C., J. R. Beddington, I. R. Crute, L. Haddad, D. Lawrence, J. F. Muir, J. Pretty, S. Robinson, S. M. Thomas, and C. Toulmin. 2010. Food security: the challenge of feeding 9 billion people. Science 327: 812-818.

Graham, J. M., and M. J. Edwards. 1998. Teratogen Update: Gestational Effects of Maternal Hyperthermia Due to Febrile Illnesses and Resultant Patterns of Defects in Humans. Teratology 58: 209-221.

Hackett, J. A., and M. A. Surani. 2013. DNA methylation dynamics during the mammalian life cycle. Philos Trans R Soc Lond B Biol Sci 368.

Hajkova, P., S. Erhardt, N. Lane, T. Haaf, O. El-Maarri, W. Reik, J. Walter, and M. A. Surani. 2002. Epigenetic reprogramming in mouse primordial germ cells. Mech Dev 117: 15-23. 35

Hall, D. M., K. R. Baumgardner, T. D. Oberley, and C. V. Gisolfi. 1999. Splanchnic tissues undergo hypoxic stress during whole body hyperthermia. Am J Physiol 276: G1195- 1203.

Hall, D. M., G. R. Buettner, L. W. Oberley, L. Xu, R. D. Matthes, and C. V. Gisolfi. 2001. Mechanisms of circulatory and intestinal barrier dysfunction during whole body hyperthermia. Am J Physiol Heart Circ Physiol 280: H509-521.

Hao, Y., Y. Cui, and X. Gu. 2016. Genome-wide DNA methylation profiles changes associated with constant heat stress in pigs as measured by bisulfite sequencing. Sci Rep 6: 27507.

Hayakawa, K., J. Ohgane, S. Tanaka, S. Yagi, and K. Shiota. 2012. Oocyte-specific linker histone H1foo is an epigenomic modulator that decondenses chromatin and impairs pluripotency. Epigenetics 7: 1029-1036.

Hoffmann, I. 2010. Climate change and the characterization, breeding and conservation of animal genetic resources. Anim Genet 41 Suppl 1: 32-46.

Horowitz, M. 2014. Heat acclimation, epigenetics, and cytoprotection memory. Comprehensive Physiology 4: 199-230.

Inagaki, T., M. Tachibana, K. Magoori, H. Kudo, T. Tanaka, M. Okamura, M. Naito, T. Kodama, Y. Shinkai, and J. Sakai. 2009. Obesity and metabolic syndrome in histone demethylase JHDM2a-deficient mice. Genes to cells : devoted to molecular & cellular mechanisms 14: 991-1001.

Ingram, D. L. 1967. Stimulation of cutaneous glands in the pig. J. Comp. Pathol. 77: 93-98.

Johnson, J. S., R. L. Boddicker, M. V. Sanz-Fernandez, J. W. Ross, J. T. Selsby, M. C. Lucy, T. J. Safranski, R. P. Rhoads, and L. H. Baumgard. 2013. Effects of mammalian in utero heat stress on adolescent body temperature. International Journal of Hyperthermia: the official journal of European Society for Hyperthermic Oncology, North American Hyperthermia Group 29: 696-702.

Johnson, J. S., M. V. Sanz Fernandez, J. F. Patience, J. W. Ross, N. K. Gabler, M. C. Lucy, T. J. Safranski, R. P. Rhoads, and L. H. Baumgard. 2015a. Effects of in utero heat stress on postnatal body composition in pigs: II. Finishing phase. J Anim Sci 93: 82- 92. 36

Johnson, J. S., M. V. Sanz Fernandez, J. T. Seibert, J. W. Ross, M. C. Lucy, T. J. Safranski, T. H. Elsasser, S. Kahl, R. P. Rhoads, and L. H. Baumgard. 2015b. In utero heat stress increases postnatal core body temperature in pigs. J Anim Sci 93: 4312-4322.

Kampinga, H. H., J. Hageman, M. J. Vos, H. Kubota, R. M. Tanguay, E. A. Bruford, M. E. Cheetham, B. Chen, and L. E. Hightower. 2009. Guidelines for the nomenclature of the human heat shock proteins. Cell Stress & Chaperones 14: 105-111.

Katsumata, M., H. Yano, N. Ishida, and A. Miyazaki. 1990. Influence of a high ambient temperature and administration of clenbuterol on body composition in rats. J Nutr Sci Vitaminol (Tokyo) 36: 569-578.

Kim, K. S., J. T. Seibert, Z. Edea, K. L. Graves, E. S. Kim, A. F. Keating, L. H. Baumgard, J. W. Ross, and M. F. Rothschild. 2018. Characterization of the acute heat stress response in gilts: III. Genome-wide association studies of thermotolerance traits in pigs. J Anim Sci 96: 2074-2085.

Kisliouk, T., M. Ziv, and N. Meiri. 2009. Epigenetic control of translation regulation: alterations in histone H3 lysine 9 post-translation modifications are correlated with the expression of the translation initiation factor 2B (Eif2b5) during thermal control establishment. Dev Neurobiol 70: 100-113.

Kumar, R., I. D. Gupta, A. Verma, S. V. Singh, N. Verma, M. R. Vineeth, A. Magotra, and R. Das. 2016. Novel SNP identification in exon 3 of HSP90AA1 gene and their association with heat tolerance traits in Karan Fries (Bos taurus x Bos indicus) cows under tropical climatic condition. Trop Anim Health Prod 48: 735-740.

Kumar, R., I. D. Gupta, A. Verma, N. Verma, and M. R. Vineeth. 2015. Genetic polymorphisms within exon 3 of heat shock protein 90AA1 gene and its association with heat tolerance traits in Sahiwal cows. Vet World 8: 932-936.

Lamke, J., K. Brzezinka, S. Altmann, and I. Baurle. 2016. A hit-and-run heat shock factor governs sustained histone methylation and transcriptional stress memory. The EMBO journal 35: 162-175.

Leary, C., H. J. Leese, and R. G. Sturmey. 2014. Human embryos from overweight and obese women display phenotypic and metabolic abnormalities. Hum Reprod 30: 122-132. 37

Lehnertz, B., Y. Ueda, A. A. H. A. Derijck, U. Braunschweig, L. Perez-Burgos, S. Kubicek, T. Chen, E. Li, T. Jenuwein, and A. H. F. M. Peters. 2003. Suv39h-Mediated Histone H3 Lysine 9 Methylation Directs DNA Methylation to Major Satellite Repeats at Pericentric Heterochromatin. Current Biology 13: 1192-1200.

Levy-Rimler, G., P. Viitanen, C. Weiss, R. Sharkia, A. Greenberg, A. Niv, A. Lustig, Y. Delarea, and A. Azem. 2001. The effect of nucleotides and mitochondrial chaperonin 10 on the structure and chaperone activity of mitochondrial chaperonin 60. Eur J Biochem 268: 3465-3472.

Li, M., H. Wu, Z. Luo, Y. Xia, J. Guan, T. Wang, Y. Gu, L. Chen, K. Zhang, J. Ma, Y. Liu, Z. Zhong, J. Nie, S. Zhou, Z. Mu, X. Wang, J. Qu, L. Jing, H. Wang, S. Huang, N. Yi, Z. Wang, D. Xi, J. Wang, G. Yin, L. Wang, N. Li, Z. Jiang, Q. Lang, H. Xiao, A. Jiang, L. Zhu, Y. Jiang, G. Tang, M. Mai, S. Shuai, N. Li, K. Li, J. Wang, X. Zhang, Y. Li, H. Chen, X. Gao, G. S. Plastow, S. Beck, H. Yang, J. Wang, J. Wang, X. Li, and R. Li. 2012. An atlas of DNA methylomes in porcine adipose and muscle tissues. Nat Commun 3: 850.

Li, Q., J. Han, F. Du, Z. Ju, J. Huang, J. Wang, R. Li, C. Wang, and J. Zhong. 2011a. Novel SNPs in HSP70A1A gene and the association of polymorphisms with thermo tolerance traits and tissue specific expression in Chinese Holstein cattle. Mol Biol Rep 38: 2657-2663.

Li, Q. L., Z. H. Ju, J. M. Huang, J. B. Li, R. L. Li, M. H. Hou, C. F. Wang, and J. F. Zhong. 2011b. Two novel SNPs in HSF1 gene are associated with thermal tolerance traits in Chinese Holstein cattle. DNA Cell Biol 30: 247-254.

Lindquist, S. 1980. Varying patterns of protein synthesis in Drosophila during heat shock: implications for regulation. Dev Biol 77: 463-479.

Luger, K., A. W. Mader, R. K. Richmond, D. F. Sargent, and T. J. Richmond. 1997. Crystal structure of the nucleosome core particle at 2.8 A resolution. Nature 389: 251-260.

Maciver, N. J., S. R. Jacobs, H. L. Wieman, J. A. Wofford, J. L. Coloff, and J. C. Rathmell. 2008. Glucose metabolism in lymphocytes is a regulated process with significant effects on immune cell function and survival. J Leukoc Biol 84: 949-957.

Mahjoubi, E., H. Amanlou, H. R. Mirzaei-Alamouti, N. Aghaziarati, M. H. Yazdi, G. R. Noori, K. Yuan, and L. H. Baumgard. 2014. The effect of cyclical and mild heat stress on productivity and metabolism in Afshari lambs. J Anim Sci 92: 1007-1014. 38

Mamrut, S., H. Harony, R. Sood, H. Shahar-Gold, H. Gainer, Y. J. Shi, L. Barki-Harrington, and S. Wagner. 2013. DNA methylation of specific CpG sites in the promoter region regulates the transcription of the mouse oxytocin receptor. PloS ONE 8: e56869.

Marcos-Carcavilla, A., M. Mutikainen, C. Gonzalez, J. H. Calvo, J. Kantanen, A. Sanz, N. S. Marzanov, M. D. Perez-Guzman, and M. Serrano. 2010. A SNP in the HSP90AA1 gene 5' flanking region is associated with the adaptation to differential thermal conditions in the ovine species. Cell Stress & Chaperones 15: 67-81.

Margueron, R., P. Trojer, and D. Reinberg. 2005. The key to development: interpreting the histone code? Curr Opin Genet Dev 15: 163-176.

Martin, J., A. L. Horwich, and F. U. Hartl. 1992. Prevention of protein denaturation under heat stress by the chaperonin Hsp60. Science 258: 995-998.

McKenzie, S. L., S. Henikoff, and M. Meselson. 1975. Localization of RNA from heat- induced polysomes at puff sites in Drosophila melanogaster. Proc Natl Acad Sci USA 72: 1117-1121.

Migicovsky, Z., Y. Yao, and I. Kovalchuk. 2014. Transgenerational phenotypic and epigenetic changes in response to heat stress in Arabidopsis thaliana. Plant Signal Behav 9: e27971.

Morera, P., L. Basirico, K. Hosoda, and U. Bernabucci. 2012. Chronic heat stress upregulates leptin and adiponectin secretion and expression and improves leptin, adiponectin and insulin sensitivity in mice. J Mol Endocrinol 48: 129-138.

Nielsen, B., G. Savard, E. A. Richter, M. Hargreaves, and B. Saltin. 1990. Muscle blood flow and muscle metabolism during exercise and heat stress. J Appl Physiol 69: 1040- 1046.

Norouzitallab, P., K. Baruah, M. Vandegehuchte, G. Van Stappen, F. Catania, J. Vanden Bussche, L. Vanhaecke, P. Sorgeloos, and P. Bossier. 2014. Environmental heat stress induces epigenetic transgenerational inheritance of robustness in parthenogenetic Artemia model. FASEB J 28: 3552-3563.

O'Neill, L. A., R. J. Kishton, and J. Rathmell. 2016. A guide to immunometabolism for immunologists. Nat Rev Immunol 16: 553-565. 39

Oresanya, T. F., A. D. Beaulieu, and J. F. Patience. 2008. Investigations of energy metabolism in weanling barrows: the interaction of dietary energy concentration and daily feed (energy) intake. J Anim Sci 86: 348-363.

Palsson-McDermott, E. M., and L. A. O'Neill. 2013. The Warburg effect then and now: from cancer to inflammatory diseases. Bioessays 35: 965-973.

Pearce, S. C., N. K. Gabler, J. W. Ross, J. Escobar, J. F. Patience, R. P. Rhoads, and L. H. Baumgard. 2013a. The effects of heat stress and plane of nutrition on metabolism in growing pigs. J Anim Sci 91: 2108-2118.

Pearce, S. C., V. Mani, R. L. Boddicker, J. S. Johnson, T. E. Weber, J. W. Ross, R. P. Rhoads, L. H. Baumgard, and N. K. Gabler. 2013b. Heat stress reduces intestinal barrier integrity and favors intestinal glucose transport in growing pigs. PloS ONE 8: e70215.

Pearce, S. C., V. Mani, T. E. Weber, R. P. Rhoads, J. F. Patience, L. H. Baumgard, and N. K. Gabler. 2013c. Heat stress and reduced plane of nutrition decreases intestinal integrity and function in pigs. J Anim Sci 91: 5183-5193.

Pearce, S. C., M. V. Sanz-Fernandez, J. H. Hollis, L. H. Baumgard, and N. K. Gabler. 2014. Short-term exposure to heat stress attenuates appetite and intestinal integrity in growing pigs. J Anim Sci 92: 5444-5454.

Pecinka, A., H. Q. Dinh, T. Baubec, M. Rosa, N. Lettner, and O. Mittelsten Scheid. 2010. Epigenetic regulation of repetitive elements is attenuated by prolonged heat stress in Arabidopsis. Plant Cell 22: 3118-3129.

Perera, F., and J. Herbstman. 2011. Prenatal environmental exposures, epigenetics, and disease. Reproductive Toxicol 31: 363-373.

Peters, J. 2014. The role of genomic imprinting in biology and disease: an expanding view. Nature reviews. Genetics 15: 517-530.

Pinney, S. E., and R. A. Simmons. 2010. Epigenetic mechanisms in the development of type 2 diabetes. Trends Endocrinol Metab 21: 223-229.

Pollmann, D. 2010. Seasonal effects on sow herds: industry experience and management strategies. J Anim Sci 88: 9. 40

Radons, J. 2016. The human HSP70 family of chaperones: where do we stand? Cell Stress & Chaperones 21: 379-404.

Rafiee, P., M. E. Theriot, V. M. Nelson, J. Heidemann, Y. Kanaa, S. A. Horowitz, A. Rogaczewski, C. P. Johnson, I. Ali, R. Shaker, and D. G. Binion. 2006. Human esophageal microvascular endothelial cells respond to acidic pH stress by PI3K/AKT and p38 MAPK-regulated induction of Hsp70 and Hsp27. Am J Physiol Cell Physiol 291: C931-945.

Ravagnolo, O., and I. Misztal. 2000. Genetic component of heat stress in dairy cattle, parameter estimation. J Dairy Sci 83: 2126-2130.

Raychaudhuri, S., C. Loew, R. Korner, S. Pinkert, M. Theis, M. Hayer-Hartl, F. Buchholz, and F. U. Hartl. 2014. Interplay of acetyltransferase EP300 and the proteasome system in regulating heat shock transcription factor 1. Cell 156: 975-985.

Reik, W. 2007. Stability and flexibility of epigenetic gene regulation in mammalian development. Nature 447: 425-432.

Renaudeau, D. 2005. Effects of short-term exposure to high ambient temperature and relative humidity on thermoregulatory responses of European (Large White) and Caribbean (Creole) restrictively-fed growing pigs. Anim. Res. 54: 81-93.

Renaudeau, D., M. Giorgi, F. Silou, and J. L. Weisbecker. 2006a. Effect of Breed (Lean or Fat Pigs) and Sex on Performance and Feeding Behaviour of Group Housed Growing Pigs in a Tropical Climate. Asian-Australas J Anim Sci 19: 593-600.

Renaudeau, D., J. L. Gourdine, and N. R. St-Pierre. 2011. A meta-analysis of the effects of high ambient temperature on growth performance of growing-finishing pigs. J Anim Sci 89: 2220-2230.

Renaudeau, D., M. Leclercq-Smekens, and M. Herin. 2006b. Differences in skin characteristics in European (Large White) and Caribbean (Creole) growing pigs with reference to thermoregulation. Anim. Res. 55: 209-217.

Renaudeau, D., and J. Mourot. 2007. A comparison of carcass and meat quality characteristics of Creole and Large White pigs slaughtered at 90kg BW. Meat Sci 76: 165-171. 41

Rhoads, M. L., R. P. Rhoads, M. J. VanBaale, R. J. Collier, S. R. Sanders, W. J. Weber, B. A. Crooker, and L. H. Baumgard. 2009. Effects of heat stress and plane of nutrition on lactating Holstein cows: I. Production, metabolism, and aspects of circulating somatotropin. J Dairy Sci 92: 1986-1997.

Rhoads, R. P., A. J. La Noce, J. B. Wheelock, and L. H. Baumgard. 2011. Alterations in expression of gluconeogenic genes during heat stress and exogenous bovine somatotropin administration. J Dairy Sci 94: 1917-1921.

Ritossa, F. 1962. A new puffing pattern induced by temperature shock and DNP in drosophila. Experientia 18: 571–573.

Rivera, R. M., and J. W. Ross. 2013. Epigenetics in fertilization and preimplantation embryo development. Prog Biophys Mol Biol 113: 423-432.

Rooij, S. R. d. 2011. Prenatal Diet and Stress Responsiveness. In: V. R. Preedy, R. R. Watson and C. R. Martin (eds.) Handbook of Behavior, Food and Nutrition. p 2023- 2039. Springer New York, New York, NY.

Roseboom, T., S. de Rooij, and R. Painter. 2006. The Dutch famine and its long-term consequences for adult health. Early Hum Dev 82: 485-491.

Ross, J. W., B. J. Hale, J. T. Seibert, M. R. Romoser, M. K. Adur, A. F. Keating, and L. H. Baumgard. 2017. Physiological mechanisms through which heat stress compromises reproduction in pigs. Mol Reprod Dev 84: 934-945.

Rowell, L. B., G. L. Brengelmann, J. R. Blackmon, R. D. Twiss, and F. Kusumi. 1968. Splanchnic blood flow and metabolism in heat-stressed man. J Appl Physiol 24: 475- 484.

Sajjanar, B., R. Deb, U. Singh, S. Kumar, M. Brahmane, A. Nirmale, S. K. Bal, and P. S. Minhas. 2015. Identification of SNP in HSP90AB1 and its association with the relative thermotolerance and milk production traits in Indian dairy cattle. Anim Biotechnol 26: 45-50.

Sakurada, S., and J. R. Hales. 1998. A role for gastrointestinal endotoxins in enhancement of heat tolerance by physical fitness. J Appl Physiol (1985) 84: 207-214. 42

Sanders, S. R. 2010. Effects of heat stress on energetic metabolism in rats. Dissertation, The University of Arizona.

Sanz Fernandez, M. V., J. S. Johnson, M. Abuajamieh, S. K. Stoakes, J. T. Seibert, L. Cox, S. Kahl, T. H. Elsasser, J. W. Ross, S. C. Isom, R. P. Rhoads, and L. H. Baumgard. 2015a. Effects of heat stress on carbohydrate and lipid metabolism in growing pigs. Physiol Rep 3.

Sanz Fernandez, M. V., S. K. Stoakes, M. Abuajamieh, J. T. Seibert, J. S. Johnson, E. A. Horst, R. P. Rhoads, and L. H. Baumgard. 2015b. Heat stress increases insulin sensitivity in pigs. Physiol Rep 3.

Sarge, K. D., S. P. Murphy, and R. I. Morimoto. 1993. Activation of heat shock gene transcription by heat shock factor 1 involves oligomerization, acquisition of DNA- binding activity, and nuclear localization and can occur in the absence of stress. Mol Cell Biol 13: 1392-1407.

Schopf, F. H., M. M. Biebl, and J. Buchner. 2017. The HSP90 chaperone machinery. Nat Rev Mol Cell Biol 18: 345-360.

Schuettengruber, B., D. Chourrout, M. Vervoort, B. Leblanc, and G. Cavalli. 2007. Genome regulation by polycomb and trithorax proteins. Cell 128: 735-745.

Seckl, J. R. 2004. Prenatal glucocorticoids and long-term programming. Eur J Endocrinol 151 Suppl 3: U49-62.

Seo, J. H., J. H. Park, E. J. Lee, T. T. Vo, H. Choi, J. Y. Kim, J. K. Jang, H. J. Wee, H. S. Lee, S. H. Jang, Z. Y. Park, J. Jeong, K. J. Lee, S. H. Seok, J. Y. Park, B. J. Lee, M. N. Lee, G. T. Oh, and K. W. Kim. 2016. ARD1-mediated Hsp70 acetylation balances stress-induced protein refolding and degradation. Nat Commun 7: 12882.

Seo, S. N., and B. McCarl. 2011. Managing Livestock Species under Climate Change in Australia. Animals (Basel) 1: 343-365.

Seong, K. H., D. Li, H. Shimizu, R. Nakamura, and S. Ishii. 2011. Inheritance of stress- induced, ATF- 2-dependent epigenetic change. Cell 145: 1049-1061. 43

Shankar, K., A. Harrell, X. Liu, J. M. Gilchrist, M. J. Ronis, and T. M. Badger. 2008. Maternal obesity at conception programs obesity in the offspring. Am J Physiol Regul Integr Comp Physiol 294: R528-538.

Shiota, K., and T. Kayamura. 1989. Effect of Prenatal Heat Stress on Postnatal Growth, Behavior and Learning Capacity in Mice. Neonatology 56: 6-14.

Shivaswamy, S., and V. R. Iyer. 2008. Stress-dependent dynamics of global chromatin remodeling in yeast: dual role for SWI/SNF in the heat shock stress response. Mol Cell Biol 28: 2221-2234.

Simmons, R. A., L. J. Templeton, and S. J. Gertz. 2001. Intrauterine growth retardation leads to the development of type 2 diabetes in the rat. Diabetes 50: 2279-2286.

Smith, Z. D., M. M. Chan, T. S. Mikkelsen, H. Gu, A. Gnirke, A. Regev, and A. Meissner. 2012. A unique regulatory phase of DNA methylation in the early mammalian embryo. Nature 484: 339-344.

Smith, Z. D., and A. Meissner. 2013. DNA methylation: roles in mammalian development. Nat Rev Genet 14: 204-220.

St-Pierre, N. R., B. Cobanov, and G. Schnitkey. 2003. Economic Losses from Heat Stress by US Livestock Industries. J Dairy Sci 86: E52-E77.

Tan, M., H. Luo, S. Lee, F. Jin, J. S. Yang, E. Montellier, T. Buchou, Z. Cheng, S. Rousseaux, N. Rajagopal, Z. Lu, Z. Ye, Q. Zhu, J. Wysocka, Y. Ye, S. Khochbin, B. Ren, and Y. Zhao. 2011. Identification of 67 histone marks and histone lysine crotonylation as a new type of histone modification. Cell 146: 1016-1028.

Tetievsky, A., and M. Horowitz. 2010. Posttranslational modification in histones underlie heat acclimation-mediated cytoprotective memory. J Appl Physiol 109: 1552-1561.

Thomsen, H., H. K. Lee, M. F. Rothschild, M. Malek, and J. C. Dekkers. 2004. Characterization of quantitative trait loci for growth and meat quality in a cross between commercial breeds of swine. J Anim Sci 82: 2213-2228.

Thomson, S., A. Hollis, C. A. Hazzalin, and L. C. Mahadevan. 2004. Distinct stimulus- specific histone modifications at hsp70 chromatin targeted by the transcription factor heat shock factor-1. Mol Cell 15: 585-594. 44

Thornton, P. K. 2010. Livestock production: recent trends, future prospects. Philos Trans R Soc Lond B Biol Sci 365: 2853-2867.

Tissieres, A., H. K. Mitchell, and U. M. Tracy. 1974. Protein synthesis in salivary glands of Drosophila melanogaster: relation to chromosome puffs. J Mol Biol 84: 389-398.

Tomizawa, S., H. Kobayashi, T. Watanabe, S. Andrews, K. Hata, G. Kelsey, and H. Sasaki. 2011. Dynamic stage-specific changes in imprinted differentially methylated regions during early mammalian development and prevalence of non-CpG methylation in oocytes. Development 138: 811-820.

Tse, W. Y., S. C. Town, G. K. Murdoch, S. Novak, M. K. Dyck, C. T. Putman, G. R. Foxcroft, and W. T. Dixon. 2008. Uterine crowding in the sow affects litter sex ratio, placental development and embryonic myogenin expression in early gestation. Reprod Fertil Dev 20: 497-504.

Turner, B. M. 2002. Cellular Memory and the Histone Code. Cell 111: 285-291.

United Nations. 2017. World population projected to reach 9.8 billion in 2050, and 11.2 billion in 2100. https://www.un.org/development/desa/en/news/population/world- population-prospects-2017.html.

Verma, N., I. D. Gupta, A. Verma, R. Kumar, R. Das, and M. R. Vineeth. 2016. Novel SNPs in HSPB8 gene and their association with heat tolerance traits in Sahiwal indigenous cattle. Trop Anim Health Prod 48: 175-180.

Vinoth, A., T. Thirunalasundari, M. Shanmugam, A. Uthrakumar, S. Suji, and U. Rajkumar. 2018. Evaluation of DNA methylation and mRNA expression of heat shock proteins in thermal manipulated chicken. Cell Stress & Chaperones 23: 235-252.

Vinson, C., and R. Chatterjee. 2012. CG methylation. Epigenomics 4: 655-663.

Vire, E., C. Brenner, R. Deplus, L. Blanchon, M. Fraga, C. Didelot, L. Morey, A. Van Eynde, D. Bernard, J. M. Vanderwinden, M. Bollen, M. Esteller, L. Di Croce, Y. de Launoit, and F. Fuks. 2006. The Polycomb group protein EZH2 directly controls DNA methylation. Nature 439: 871-874. 45

Wagner, J. R., S. Busche, B. Ge, T. Kwan, T. Pastinen, and M. Blanchette. 2014. The relationship between DNA methylation, genetic and expression inter-individual variation in untransformed human fibroblasts. Genome Biol 15: R37.

Weaver, I. C., N. Cervoni, F. A. Champagne, A. C. D'Alessio, S. Sharma, J. R. Seckl, S. Dymov, M. Szyf, and M. J. Meaney. 2004. Epigenetic programming by maternal behavior. Nat Neurosci 7: 847-854.

West, J. W. 2003. Effects of heat-stress on production in dairy cattle. J Dairy Sci 86: 2131- 2144.

Westwood, J. T., and C. Wu. 1993. Activation of Drosophila heat shock factor: conformational change associated with a monomer-to-trimer transition. Mol Cell Biol 13: 3481-3486.

Wheelock, J. B., R. P. Rhoads, M. J. Vanbaale, S. R. Sanders, and L. H. Baumgard. 2010. Effects of heat stress on energetic metabolism in lactating Holstein cows. J Dairy Sci 93: 644-655.

World Bank. 2007. World Development Report 2008 : Agriculture for Development., Washington, D.C.

Xiong, Q., J. Chai, H. Xiong, W. Li, T. Huang, Y. Liu, X. Suo, N. Zhang, X. Li, S. Jiang, and M. Chen. 2013. Association analysis of HSP70A1A haplotypes with heat tolerance in Chinese Holstein cattle. Cell Stress & Chaperones 18: 711-718.

Yahav, S., and S. Hurwitz. 1996. Induction of thermotolerance in male broiler chickens by temperature conditioning at an early age. Poult Sci 75: 402-406.

Yahav, S., and J. P. McMurtry. 2001. Thermotolerance acquisition in broiler chickens by temperature conditioning early in life--the effect of timing and ambient temperature. Poult Sci 80: 1662-1666.

Yang, X., H. Han, D. D. De Carvalho, F. D. Lay, P. A. Jones, and G. Liang. 2014. Gene body methylation can alter gene expression and is a therapeutic target in cancer. Cancer Cell 26: 577-590. 46

Zhang, Z. Y., G. Q. Jia, J. J. Zuo, Y. Zhang, J. Lei, L. Ren, and D. Y. Feng. 2012. Effects of constant and cyclic heat stress on muscle metabolism and meat quality of broiler breast fillet and thigh meat. Poult Sci 91: 2931-2937.

Zhu, J. Q., J. H. Liu, X. W. Liang, B. Z. Xu, Y. Hou, X. X. Zhao, and Q. Y. Sun. 2008. Heat stress causes aberrant DNA methylation of H19 and Igf-2r in mouse blastocysts. Mol Cells 25: 211-215.

Zou, J., Y. Guo, T. Guettouche, D. F. Smith, and R. Voellmy. 1998. Repression of heat shock transcription factor HSF1 activation by HSP90 (HSP90 complex) that forms a stress- sensitive complex with HSF1. Cell 94: 471-480.

Zumbach, B., I. Misztal, S. Tsuruta, J. P. Sanchez, M. Azain, W. Herring, J. Holl, T. Long, and M. Culbertson. 2008. Genetic components of heat stress in finishing pigs: Parameter estimation. J Anim Sci 86: 2076-2081.

CHAPTER 2. CHARACTERIZING THE ACUTE HEAT STRESS RESPONSE IN GILTS: I. THERMOREGULATORY AND PRODUCTION VARIABLES

Modified from a paper published in 2018 by the Journal of Animal Science 96: 941-949

J.T. Seibert*, K.L. Graves*, B.J. Hale*, A.F. Keating*, L.H. Baumgard*, and J.W. Ross*

*Department of Animal Science, Iowa State University, Ames, IA, 50011

Corresponding author: Jason W. Ross, PhD Department of Animal Science Iowa State University Ames, IA, 50011, United States Phone number: (515) 294-8647 Fax number: 515-294-4471 E-mail: [email protected]

Authors contributions:

J.T.S conducted the experiment, interpreted the data, drafted the manuscript, and prepared the figures. K.L.G. and B.J.H. helped conduct the live phase of the experiment or provided intellectual contributions. J.T.S., A.F.K., L.H.B. and J.W.R. contributed to the experimental design of the study. All authors edited the manuscript.

Results described here within were supported by the National Pork Board, the Iowa Pork

Producers Association, and Agriculture and Food Research Initiative Competitive Grant no.

2011-67003-30007 from the USDA National Institute of Food and Agriculture.

Abstract

Identifying traits associated with susceptibility or tolerance to heat stress (HS) is prerequisite for developing strategies to improve efficient pork production during the 48 summer months. Study objectives were to determine the relationship between the thermoregulatory and production responses to acute HS in pigs. Prepubertal gilts (n = 235;

77.9 ± 1.2 kg BW) were exposed to a thermoneutral (TN) period (P1, 24 h; 21.9 ± 0.5°C, 62

± 13% RH; fed ad libitum) followed immediately by a subsequent acute HS period (P2, 24 h;

29.7 ± 1.3°C, 49 ± 8% RH; fed ad libitum). Rectal temperature (TR), skin temperature (TS), and respiration rate (RR) were monitored and BW and feed intake (FI) were determined. All pigs had increased TR, TS, and RR (0.80°C, 5.65°C, and 61.2 bpm, respectively; P < 0.01) and decreased FI and BW (29% and 1.10 kg, respectively; P < 0.01) during P2 compared to

2 P1. Interestingly, body temperature indices did not explain variation in FI during P2 (R ≤

0.02). Further, the percent change in BW during P2 was only marginally explained by each

2 2 body temperature index (R ≤ 0.06) or percent change in FI (R = 0.14). During HS, TR was strongly correlated with P1 TR (r = 0.72, P < 0.01), indicating a pig’s body temperature during TN conditions predicts the severity of hyperthermia during HS. Additionally, the change in TR (ΔTR, HS TR – TN TR) was larger in pigs retrospectively classified as susceptible (SUS) as compared to tolerant (TOL) pigs (1.05 vs. 0.51°C, respectively; P <

0.01). In summary, thermoregulatory responses and production variables during acute HS are only marginally related. Further, changes in BW and FI were unexpectedly poorly correlated during acute HS (r = 0.34; P < 0.01). Collectively, suboptimal growth is largely independent on the thermoregulatory response and hypophagia during acute HS. Consequently, incorporating solely body temperature indices into a genetic index is likely insufficient for substantial progress in selecting HS tolerant pigs.

Keywords: thermoregulation, heat tolerance, gilts 49

Introduction

Heat stress (HS) annually constrains pork profitability, independent of farm size and system. Lost economic opportunity primarily results from increased mortality and morbidity, altered carcass composition, reduced fecundity, decreased and inconsistent growth, and compromised feed and facility efficiency (Baumgard and Rhoads, 2013). Porcine sweat glands are non-functional and the pig’s thermoregulatory ability is further complicated by a thick subcutaneous adipose tissue layer, which impedes radiant heat loss (D'Allaire et al.,

1996). Furthermore, genetic selection emphasizing rapid lean tissue accretion is accompanied by increased endogenous heat production; a thermal scenario that exacerbates their innate inability to tolerate heat (Bianca, 1976; Renaudeau et al., 2011). Additionally, while HS is already a significant burden to the pig industry, production efficiency may be further compromised if climate patterns change as predicted (IPCC, 2007; Hoffmann, 2010).

A conserved response to HS is reduced feed intake, ostensibly an attempt to minimize metabolic heat production (Bianca, 1976; Blaxter, 1989; Collin et al., 2001; Baumgard and

Rhoads, 2013). Consequently, reduced feed intake has traditionally been assumed to be the major reason underlying negative effects of HS on production parameters in animal agriculture (Collin et al., 2001; West 2003; DeShazer et al. 2009). However, utilizing a pair- feeding experimental design demonstrates that heat-induced hypophagia cannot fully explain the negative consequences of HS (Prunier et al., 1997; Baumgard and Rhoads, 2013; Pearce et al., 2013; Sanz Fernandez et al., 2015). Therefore, there are direct effects of HS

(independent of reduced nutrient intake) that contribute to poor summer performance

(Baumgard and Rhoads, 2013).

Incorporating a thermal sensitivity index into a genetic selection program that emphasizes HS tolerance without compromising other economically important traits would 50 be valuable. However, identifying pigs that maintain productivity during HS, or that exhibit heat tolerance, is difficult due to the biological complexity of the HS response. Previous studies have shown “heat tolerance” with respect to minimizing the HS effects on a single production characteristic (Bloemhof et al., 2008; Zumbach et al., 2008). Thus, study objectives were to examine whether or not multiple thermoregulatory responses are associated with production phenotypes during acute HS. We hypothesized that animals retrospectively classified as tolerant (TOL) and susceptible (SUS) to HS based on their thermoregulatory response to an acute heat challenge would also have industry relevant productivity differences during HS.

Materials and Methods

Animals and Experimental Design

All procedures were approved by the Iowa State University Institutional Animal Care and Use Committee. Crossbred gilts (n = 235; PIC maternal x Duroc terminal sire) from the same cohort were received on the 24th day of age and arrived immediately after weaning.

Due to logistical constraints, the experiment was conducted in 5 replications (n = 44 to

48/replicate). The initial BW from replications 1 to 5 were 59±1.0, 64±1.2, 77±1.2, 88±1.1, and 103±1.6 kg, respectively. Gilts were randomly assigned and housed in individual crates

(57 x 221 cm; 24 crates/room) at the Iowa State University Swine Nutrition Farm research facility (Ames, IA). Each crate was equipped with a stainless steel feeder and a nipple drinker. Water and feed were provided ad libitum during the entire experiment.

All pigs were fed a standard diet consisting mainly of corn and soybean meal formulated to meet or exceed nutrient requirements (NRC, 2012; Table 2.1). The study was divided into three experimental periods (P) for each replicate: P0, P1, and P2. Period 0 (72 h) 51 served as an acclimation period in which all pigs were housed individually in thermoneutral

(TN) conditions (21.9 ± 0.5°C, 62 ± 13% relative humidity [RH]). After P0, pigs remained in

TN conditions for 24 h (Period 1; P1) and then exposed to HS (29.7 ± 1.3°C, 49 ± 8% RH) conditions for 24 h (Period 2; P2). Pigs were exposed to a 12:12 h light-dark cycle during P0, but continuous light during P1 and P2 to allow for accurate data collection.

Production and Thermoregulatory Measurements

Rectal temperature (TR) was measured with a calibrated and lubricated digital thermometer (Welch Allyn SureTemp® Plus 690; accuracy: ± 0.1°C; Skaneateles Falls, NY,

USA), skin temperature (TS) was measured using a calibrated infrared thermometer (HDE

ST380A Infrared Thermometer; accuracy ± 2.0°C, HDE, Allentown, PA), and respiration rate (RR) was determined by counting flank movements for 15 s and then multiplying by 4 to obtain breaths/min (bpm). Feed intake (FI) was measured daily and thermoregulatory indices were recorded for each pig hourly during both P1 and P2. Body weights were collected at the beginning of P0 and after P1 and P2. The percent change in BW (ΔBW%) and FI (ΔFI%) were calculated by subtracting each variable’s P1 measurement from P2, dividing that value by the P1 measurement, and expressing the fraction as a percentage. Ambient temperature was controlled, but humidity was not governed and both parameters were recorded every 30 min by four data loggers (Lascar EL-USB-2-LCD, Erie, PA) in each room and later condensed into average values.

Determination of HS Tolerance and Susceptibility Based on the Thermoregulatory Response

All TR data recorded hourly during P1 were condensed into an average value to represent each individual pig’s basal core temperature (TN TR). Only TR data recorded hourly between the 4th and 12th h of HS were condensed into an average value to reflect core body temperature during P2 (HS TR). These time points were chosen to minimize variability associated with both the rate of TR increase (h 1 to 4) and the counter regulatory mechanisms

(13 to 24 h) associated with TR acclimation. The difference in core body temperature (ΔTR) was calculated by subtracting TN TR from the HS TR. The Δ TR was plotted against the HS

TR in order to determine the relationship between each pig’s average HS TR and ΔTR. Since

ΔTR was highly associated with HS TR (r = 0.83; P < 0.01; Supplementary Table 2.1), each pig’s tolerance to the heat load was classified solely on the basis of HS TR (i.e. higher and lower HS TR were considered reflective of HS susceptibility or tolerance, respectively). For each of the five replicates, the 10 most tolerant (TOL; n = 50) and susceptible (SUS; n = 50) pigs were identified (based only on HS TR) and allocated to their retrospective classifications for statistical analysis.

Statistical Analysis

All data were statistically analyzed using SAS University Edition software, version

9.4 (SAS Institute Inc., Cary, NC). Thermoregulatory indices and production data were analyzed using PROC MIXED; the model included P and room as fixed effects, replication as a random effect, and the BW collected at the beginning of P0 as a covariate. To investigate relationships between variables, PROC CORR was used to generate Pearson’s correlation coefficients. Thermoregulatory indices and production data associated with the retrospective 53

TOL and SUS treatments were analyzed using PROC MIXED; the model included classification, P, and room as fixed effects, replication as a random effect, and the BW collected at the beginning of P0 as a covariate. Data are reported as LSmeans and statistical significance (P ≤ 0.05) and tendency (0.05 < P ≤ 0.10) thresholds were utilized for interpretation of results.

Results

Thermoregulatory and Productivity Variables have Marginal Relationships during TN Conditions

During P1, TR , TS, and RR were 39.03 ± 0.03°C, 36.63 ± 0.07°C, and 39.0 ± 0.8 bpm, respectively (Supplementary Table 2.1). Period 1 FI was 2.48 ± 0.04 kg and BW was

84.4 ± 0.7 kg. A marginal association was observed for TN FI and TN TR for all pigs in the study (r = 0.03; P = 0.64; Supplementary Table 2.1). Thermoneutral FI and ADG were correlated to each other (r = 0.45; P < 0.01; Supplementary Table 2.1), but TN TR and ADG were poorly associated (R2 = 0.04).

Minor Relationships between Thermoregulation and Production Parameters are Observed during HS Conditions

As expected, TR, TS, and RR were increased (0.80°C, 5.65°C, and 61.2 bpm, respectively; P < 0.01) during P2 compared to P1 (Table 2.2). The variation in each thermoregulatory parameter also increased during HS as the SD of TR, TS, and RR was increased during P2 compared to P1 (78, 21, and 197%, respectively; Fig. 2.1). Decreased FI

(0.71 kg; P < 0.01; Table 2.2) and BW (1.09 kg; P < 0.01; Table 2.2) occurred during P2 relative to P1. During P2, ΔFI% and TR were weakly correlated (r = -0.16; P = 0.03;

Supplementary Table 2.1 and Fig. 2.2A). Additionally, the heat-induced ΔBW% was poorly associated with HS TR (r = -0.23; P < 0.01; Supplementary Table 2.1 and Fig. 2.2B). 54

Predictably, a general trend of increasing TS (r = 0.29; P < 0.01; Supplementary Table 2.1 and Fig. 2.3A) or RR (r = 0.35; P < 0.01; Supplementary Table 2.1 and Fig. 2.3B) with increasing TR was observed during P2. During P2, RR and TS were also correlated (r = 0.23;

P < 0.01; Supplementary Table 2.1).

Relationship between Thermoregulatory Indices and Production Characteristics during TN and HS Conditions

Interestingly, HS TR was strongly correlated to TN TR (r = 0.72; P < 0.01;

Supplementary Table 2.1 and Fig. 2.4A). In addition, ΔTR was correlated with TN TR (r =

0.21; P < 0.01; Supplementary Table 2.1 and Fig. 2.4B). During HS, a low proportion of variation in ΔBW% was explained by the ΔFI% (R2 = 0.14; Fig. 2.5).

Pigs Retrospectively Classified as SUS had Improved Production Characteristics during TN Conditions yet Impaired Thermoregulatory Ability during HS Compared to TOL Pigs

Based on our retrospective classification of TOL and SUS pigs, SUS pigs had increased ADG (1.13 vs. 0.89 kg/d; P = 0.01; Table 2.3) and numerically, but not significantly, increased FI (2.53 vs. 2.42 kg; P = 0.30; Table 2.3) and G:F (0.47 vs. 0.42; P =

0.21; Table 2.3) compared to TOL pigs during P0 and P1. Interestingly, TR was increased in

SUS pigs compared to TOL pigs even during P1 (39.19 vs. 38.87°C; P < 0.01; Table 2.3;

Fig. 2.6A) and also in P2 (40.24 vs. 39.38°C; P < 0.01; Table 2.3; Fig 2.6A), and was also accompanied by increased ΔBW% (-1.60 vs. -0.77%; P < 0.01; Table 2.3; Fig. 2.6B) and numerically increased ΔFI% (-30 vs. -22 %; P = 0.12; data not shown). Additionally, SUS pigs had an increased rate of change in TR compared to TOL pigs (0.12 vs. 0.07 °C/h; P < 55

0.01; Table 2.3) during P2 (1 to 14 h), indicating SUS pigs not only exhibited a higher maximum TR during HS, but achieved it at a faster rate than did the TOL pigs.

Discussion

Heat stress is an enormous economic burden to the pork industry and animal agriculture as a whole (St-Pierre et al., 2003; Pollmann 2010). Despite employing HS abatement strategies, HS continues to be a seasonal challenge to animal production efficiency and prevents livestock from expressing their full genetic potential. Notably, the negative effects of HS on animal agriculture could become increasingly severe (Hoffmann, 2010).

Thus, incorporating a HS marker(s) into the genetic selection program may facilitate improved productivity during the warm summer months.

In this study, all pigs had markedly increased body temperature indices, indicating an activated thermoregulatory response to the acute heat load. The magnitude of the thermoregulatory response to HS has potential to be predicated based on intrinsic thermoregulatory set points, as TN TR positively correlated with HS TR, a phenotype reproducible later in life (Graves et al., 2017). Surprisingly, the TR response was highly variable among pigs during HS and variation was notably larger in HS than in TN conditions with some animals demonstrating minimal to no TR increase during HS compared to TN conditions. This high degree of variation within individuals suggests the TR response is a complex phenotype regulated in part by multiple independent or interdependent factors influenced by environment.

Altering FI is a major mechanism for thermoregulation in pigs as a physiological strategy to reduce metabolic heat production (Bianca, 1976; Blaxter, 1989; Collin et al.,

2001; Baumgard and Rhoads, 2013). Feed intake and the associate thermic effect of 56 feeding’s contribution to whole body thermal energy has been demonstrated using TN pair- fed models, in which pigs pair-fed to HS counterparts have decreased TR (~0.5°C) relative to pigs fed ad libitum in the same TN conditions (Pearce et al., 2013). In this study, as expected, FI was reduced during HS, although the relationship between TR with decreased FI was also highly variable between individuals, suggesting that FI and TR (while related) are governed by partially independent mechanisms during HS. While not specifically addressed in this study, perhaps this variability could be attributed in part to mitochondrial function

(Brand, 2005) or mechanisms regulating appetite control (Wynne et al., 2005), but future studies are needed to fully elucidate their involvement during HS.

Surprisingly, we observed only a marginal association between ΔFI% and ΔBW% from P1 to P2. Some pigs even gained weight while others lost weight despite similar FI during the acute heat challenge (Fig. 2.3), again highlighting the variability in the HS response between individuals. Furthermore, animals representing the extremes with respect to ΔFI% had strikingly similar ΔBW% during P2 (Fig. 2.3). Reasons for this are unclear, but may pertain to the amount of water consumed during the acute heat challenge. However, we have previously observed similar hematocrits, a crude measure of hydration, between TN and

HS pigs (Boddicker et al., 2014; Pearce et al., 2015). Another potential explanation is variation in digesta passage rates during HS. Heat-stressed pigs are thought to have improved digestibility (Collin et al., 2001; Cosgrove et al., 2002), which is attributed to slower solids passage rate (Christopherson and Kennedy, 1983), but whether or not this contributes to the variation in production responses (i.e. ΔBW) during acute HS will require further investigation. 57

Intriguingly, we observed weak associations between TN production variables and their changes in response to acute HS. For example, TN FI and ΔFI% as well as ADG and

ΔBW% were poorly related (R2 = 0.12 and 0.03, respectively). Thus, traditionally utilized

TN production variable measurements (FI and ADG) actually appear to have low predictive value of the magnitude of ΔFI% and ΔBW% during acute HS. However, when comparing the extremes on the basis of thermoregulation, SUS pigs performed better with ADG under TN conditions, but were more sensitive to the acute HS exemplified through the increase loss of

BW as compared to TOL pigs (Table 2.3 and Fig. 2.6B).

One factor that inherently decreases pig tolerance to HS is the genetic selection emphasis that has been placed on economically important production traits (Nienaber and

Hahn, 2007). In particular, improved lean tissue accretion rates are accompanied by increased basal heat production (Brown-Brandl et al., 2004). Our data corroborate previous studies as SUS pigs (those having higher TR during HS) had better ADG and numerically increased G:F compared to TOL during TN conditions. Despite an inadvertent reduction in thermotolerance in current high producing commercial lines, it does appear that genetic control over specific phenotypes are seasonally or environmentally influenced (Zumbach et al., 2008). Furthermore, selection for improved growth has deleterious consequences on productivity in specific environmental conditions, as evidenced by a negative genetic correlation of growth or farrowing rate with heat tolerance (Zumbach et al., 2008; Bloemhof et al., 2012). This agrees with ruminant reports as dairy cows and sheep selected for improved milk production are more sensitive to increases in temperature-humidity indices

(Ravagnolo and Misztal, 2000; Bohmanova et al., 2005; Finocchiaro et al., 2005). 58

Few previous studies have intensely collected both thermoregulatory indices and production traits simultaneously; thus the relationships between the aforementioned parameters are ill-defined. This study establishes that a pig’s thermoregulatory response to

HS appears to be partially set prior to experiencing a heat load, evident by the fact that TN

TR is predictive of TR and ΔTR during an acute HS challenge. Comparing TR in pigs retrospectively classified as “tolerant” or “susceptible” based on thermoregulatory parameters under acute HS indicates the potential predictability of the HS response, in terms of thermoregulation, as ΔTR and the TR rate of change was exacerbated to a greater degree in

SUS than TOL classified pigs.

Based on the above observations, animals having a low core temperature during TN conditions could be presumed to be most tolerant to HS. While true with regard to the thermoregulatory response, from a production perspective, classification of TOL or SUS to

HS should not be determined based solely on thermoregulatory indices (e.g. TR and RR). For example, critical production efficiency parameters (FI and BW), while severely vulnerable to

HS, are not necessarily related to the thermoregulatory response during HS. In other words, thermoregulatory indices are not accurate predictors of production responses during acute

HS.

If current genetic selection methods for economically important production traits continue, sensitivity to HS will ostensibly intensify. In dairy cows, genome-wide association studies have discovered single nucleotide polymorphisms (SNP) associated with TR, but, more importantly, none of the identified SNP were linked with valued production traits

(Dikmen et al., 2013). Specific SNP in the HSP70A1A (heat shock 70 kDa protein 1A) gene have even been shown to alter thermoregulatory indices (TR and RR) without compromising 59 milk yield (Xiong et al., 2013). Furthermore, dairy cows genomically predicted to be heat tolerant have improved productivity and TR in response to HS compared to those deemed susceptible (Garner et al., 2016). Thus, improved thermotolerance does not necessarily need to be accompanied by blunted production.

In pigs, thermoregulatory and production efficiency responses to HS also appear to be controlled by different genomic regions, suggesting that maintenance of production efficiency during HS is a complex, multifactorial trait (Kim et al., 2017). Indeed, selecting animals with a “tolerant” phenotype based separately on thermoregulatory capacity or production efficiency may not ultimately increase the swine industry’s resilience to HS. This is especially true for single, economically important variables, such as feed efficiency, because the single trait only represents one component of an organism, which could potentially undermine other variables or “robustness” in response to environmental insults

(Friggens et al., 2017).

Importantly, results from this study encompass responses only to acute HS and may differ between pre- and postpubertal pigs. The types of HS that plague production agriculture are typically more chronic and episodic throughout the summer. Thus, future studies are needed to evaluate phenotypic responses to chronic and repeated HS, and also to consider periods of recovery following HS, in order to fully understand the complexity of the

HS response.

Conclusion

In the current study, we explicated that thermoregulatory ability does not predict productivity during acute HS in pigs. Interestingly, however, individual capacity to thermoregulate during HS can be predicted based on TN thermoregulatory parameters (e.g., 60

TN TR). In addition, ΔFI% did not confer a proportional relationship to ΔBW% during an acute heat challenge. Importantly, the variation of multiple biological factors, phenotypic and genotypic, need to be taken into account to address the complex multifactorial nature of HS tolerance and to ultimately bolster the pork industry’s resilience to HS.

Acknowledgements

The authors would like to acknowledge the assistance of Theresa Johnson,

Candice Hager, Matt Garrick, Sarah Edwards, Christy Calderwood, Kelsi Young, Kenton

Doty, Brittney Yehling, Marley Dobyns, Sterling Schnepf, Erin Nolan, Alexis Patinos, Briar

Tenold, Morgan Vanderlind, and Chris Almond for assistance in collecting temperature and production data.

Declaration of Interest

This project was supported by the National Pork Board. Any opinion, findings, conclusions or recommendations expressed in this publication are those of the authors and do not necessarily reflect the view of the National Pork Board. No conflicts of interest, financial or otherwise are declared by the author (s).

Literature Cited

Baumgard, L. H., and R. P. Rhoads. 2013. Effects of Heat Stress on Postabsorptive Metabolism and Energetics. Annu. Rev. Anim. Biosci. 1: 311-337. doi: 10.1146/annurev-animal-031412-103644

Bianca, W. 1976. The signifiance of meterology in animal production. Int. J. Biometeorol. 20:139-156. doi: 10.1007/BF01553047

Blaxter, K. L. 1989. Energy metabolism in animals and man. Cambridge University Press, Cambridge England; New York. 61

Bloemhof, S., E. H. van der Waaij, J. W. Merks, and E. F. Knol. 2008. Sow line differences in heat stress tolerance expressed in reproductive performance traits. J. Anim. Sci. 86:3330-3337. doi: 10.2527/jas.2008-0862

Bloemhof, S., A. Kause, E. F. Knol, J. A. M. Van Arendonk, and I. Misztal. 2012. Heat stress effects on farrowing rate in sows: Genetic paramter estimation using within-line and crossbred models. J. Anim. Sci. 90:2109-2119. doi: 10.2527/jas.2011-4650

Boddicker, R. L., J. T. Seibert, J. S. Johnson, S. C. Pearce, J. T. Selsby, N. K. Gabler, M. C. Lucy, T. J. Safranski, R. P. Rhoads, L. H. Baumgard, and J. W. Ross. 2014. Gestational heat stress alters postnatal offspring body composition indices and metabolic parameters in pigs. PLoS ONE 9: e110859. doi: 10.1371/journal.pone.0110859

Bohmanova, J., I. Misztal, S. Tsurauta, H. D. Norman, and T. J. Lawlor. 2005. National Genetic Evaluation of Milk Yield for Heat Tolerance of United State Holsteins. Interbull Bulletin 33:160-162.

Brand, M. D. 2005. The efficiency and plasticity of mitochondrial energy transduction. Biochem. Soc. Trans. 33:897-904. doi: 10.1042/BST20050897

Brown-Brandl, T. M., J. A. Nienaber, H. Xin, R. S. Gates. 2004. A literature review of swine heat production. Trans. ASAE 47:259-270.

Christopherson, R. J. and P. M. Kennedy. 1983. Effect of the thermal environment on digestion in ruminants. Can. J. Anim. Sci. 63:477-496. doi: 10.4141/cjas83-058

Collin, A., J. van Milgen, S. Dubois, and J. Noblet. 2001. Effect of high temperature and feeding level on energy utilization in piglets. J. Anim. Sci. 79:1849-1857. doi: 10.2527/2001.7971849x

Cosgrove, S. J., R. Mosenthin, R. J. Christopherson, B. T. Li, and W. C. Sauer. 2002. Thermogenic and digestive responses of ad libitum fed pigs to α2-adrenergic stimulation with guanfacin in different thermal environments. Can. J. Anim. Sci. 82: 79-86. doi: 10.4141/A01-026

D'Allaire, S., R. Drolet, and D. Brodeur. 1996. Sow mortality associated with high ambient temperatures. Can. Vet. J. 37:237-239.

DeShazer, J. A., G. L. Hahn, and H. Xin 2009. Chapter 1: Basic principles of the thermal environment and livestock energetics. In: J. A. Deshazer, editor, Livestock energetics and thermal environment management. Am. Soc. Agric. Biol. Eng., St. Joseph, MI. p. 181-209.

Dikmen, S., J. B. Cole, D. J. Null, and P. J. Hansen. 2013. Genome-Wide Association Mapping for Identification of Quantitative Trait Loci for Rectal Temperature during Heat Stress in Holstein Cattle. PLoS ONE 8(7):e69202. doi: 10.1371/journal.pone.0069202

Finocchiaro, R., J. B. C. H. M. van Kaam, M.T. Sardina, and I. Misztal. 2005. Effect of heat stress on production in Mediterranean dairy sheep. Ital. J. Anim. Sci. 4(Suppl. 2):70- 72. doi: 10.3168/jds.S0022-0302(05)72860-5

Friggens, N. C., F. Blanc, D. P. Berry, L. Puillet. 2017. Review: Deciphering animal robustness. A synthesis to facilitate its use in livestock breeding and management. Animal. 1-15. doi: 10.1017/S175173111700088X

Garner, J. B., M. L. Douglas, S. R. Williams, W. J. Wales, L. C. Marett, T. T. Nguyen, C. M. Reich, B. J. Hayes. 2016. Genomic selection improves heat tolerance in dairy cattle. Sci. Rep. 6:34114. doi: 10.1038/srep34114

Graves, K. L., J. T. Seibert, A. F. Keating, L. H. Baumgard, and J. W. Ross. 2017. Characterization of the heat stress response in gilts: II. Assessing repeatability and association with fertility. J. Anim. Sci. 96: 2419-2426. doi: 10.1093/jas/skx037

Hoffmann, I. 2010. Climate change and the characterization, breeding and conservation of animal genetic resources. Anim. Genet. 41 Suppl 1:32-46. doi: 10.1111/j.1365- 2052.2010.02043.x

Intergovernmental Panel on Climate Change (IPCC). 2007. Climate change 2007: The physical science basis. Contribution of Working Group I to the Fourth Assessment Report of the Intergovernmental Panel on Climate Change. Cambridge University Press, Cambridge; New York.

Kim, K. -S., J. T. Seibert, Z. Edea, K. L. Graves, E. -S. Kim, A. F. Keating, L. H. Baumgard, J. W. Ross, and M.F. Rothschild. 2017. Characterization of the heat stress response in gilts: III. Identification of underlying genomic control. J. Anim. Sci. 96: 2074-2085. doi: 10.1093/jas/sky131

Nienaber, J. A., and G. L. Hahn. 2007. Livestock production system management responses to thermal challenges. Int. J. Biometeorol. 52:149-157. doi: 10.1007/s00484-007- 0103-x

NRC. 2012. Nutrient Requirements of Swine. 11th rev. ed. Natl. Acad. Press, Washington, DC. doi: 10.17226/13298

Pearce, S. C., N. K. Gabler, J. W. Ross, J. Escobar, J. F. Patience, R. P. Rhoads, and L. H. Baumgard. 2013. The effects of heat stress and plane of nutrition on metabolism in growing pigs. J. Anim. Sci. 91:2108-2118. doi: 10.2527/jas.2012-5738

Pearce, S. C., M. V. Sanz Fernandez, J. Torrison, M. E. Wilson, L. H. Baumgard, and N. K. Gabler. 2015. Dietary organic zinc attenuates heat stress-induced changes in pig intestinal integrity and metabolism. J. Anim. Sci. 93:4702-4713. doi: 10.2527/jas.2015-9018

Pollmann, D. S. Seasonal effects on sow herds: industry experience and management strategies. 2010. J. Anim. Sci. 88(Suppl. 3):9. (Abstr).

Prunier, A., M. Messias de Braganca, and J. Le Dividich. 1997. Influence of high ambient temperature on performance of reproductive sows. Livest. Prod. Sci. 52:123-133. doi: 10.1016/S0301-6226(97)00137-1

Ravagnolo, O., and I. Misztal. 2000. Genetic Component of Heat Stress in Dairy Cattle, Parameter Estimation. J. Dairy Sci. 83:2126-2130. doi: 10.3168/jds.S0022- 0302(00)75095-8

Renaudeau, D., J. L. Gourdine, and N. R. St-Pierre. 2011. A meta-analysis of the effects of high ambient temperature on growth performance of growing-finishing pigs. J. Anim. Sci. 89:2220-2230. doi: 10.2527/jas.2010-3329

Sanz Fernandez, M. V., J. S. Johnson, M. Abuajamieh, S. K. Stoakes, J. T. Seibert, L. Cox, S. Kahl, T. H. Elsasser, J. W. Ross, S. C. Isom, R. P. Rhoads, and L. H. Baumgard. 2015. Effects of heat stress on carbohydrate and lipid metabolism in growing pigs. Physiol. Rep. 3:e12315. doi: 10.14814/phy2.12315

St-Pierre, N. R., B. Cobanov, and G. Schnitkey. 2003. Economic Losses from Heat Stress by US Livestock Industries. J. Dairy Sci. 86:E52-E77. doi: 10.3168/jds.S0022- 0302(03)74040-5

West, J. W. 2003. Effects of heat-stress on production in dairy cattle. J. Dairy Sci. 86:2131- 2144. doi: 10.3168/jds.S0022-0302(03)73803-X

Wynne, K., S. Stanley, B. McGowan, and S. Bloom. 2005. Appetite control. J. Endocrinol. 184:291-318. doi: 10.1677/joe.1.05866

Tables and Figures

Table 2.1. Ingredients and chemical composition of diet for growing pigs (as-fed basis) Ingredients ---- % ---- Corn 63.5 Soybean meal 14.3 Dried distillers grains 20.0 45-30 vitamin and mineral premix1 1.9 L-lysine 0.3 Calculated chemical composition % DM 87.87 Crude protein 18.91 Crude fat 3.51 Crude fiber 3.12 Ash 4.08 11.11% Limestone, 0.43% Salt, 0.21% Dried distillers grains, 0.13% Vitamin and Trace Mineral (Provided 8,382 IU vitamin A, 1,537 IU vitamin D, 45 IU vitamin E, 2 IU vitamin K, 21 mg niacin, 17 mg pantothenic acid, 4 mg riboflavin, 5 µg choline, 0.4 µg folic acid, 28 µg vitamin B12, 2 µg biotin, 246 PPM zinc, 118 PPM manganese, 321 PPM iron, 45 PPM copper, 2 PPM iodine, 0.6 PPM selenium per kg of diet), 0.02% Rono M 10,000

Table 2.2. Effects of environment on thermoregulatory indices and production variables in growing pigs Parameter Period 11 Period 22 SEM P Rectal temperature, °C 39.03 39.83 0.03 <0.01 Skin temperature, °C 36.63 42.27 0.11 <0.01 Respiration rate, bpm3 39.0 100.2 1.8 <0.01 Feed intake, kg 2.48 1.77 0.06 <0.01 BW 84.4 83.3 0.7 <0.01 1Thermoneutral (TN) conditions 2Heat stress (HS) conditions 3Breaths per minute

Table 2.3. Effects of retrospective classification and environment on thermoregulatory indices and production variables for growing pigs during the study TOL1 SUS2 P Parameter Period 13 Period 24 Period 1 Period 2 SEM Class5 Period Class*Period6 7 a c b d TR , °C 38.87 39.38 39.19 40.24 0.04 <0.01 <0.01 <0.01 8 HS TR m - 0.07 - 0.12 0.01 <0.01 - - 9 TS , °C 36.31 41.98 36.98 42.58 0.18 <0.01 <0.01 0.80 RR10, bpm 37.7a 93.3b 40.9a 108.7c 2.3 <0.01 <0.01 <0.01 FI11, kg 2.42 1.85 2.53 1.74 0.09 0.43 <0.01 0.18 BW 83.7 83.1 84.7 83.2 0.6 0.03 0.41 0.51 1Tolerant (n = 50) 2Susceptible (n = 50) 3Thermoneutral (TN) conditions 4Heat stress (HS) conditions 5 Classification of TOL or SUS; based on HS rectal temperature (TR; i.e. higher and lower HS TR were considered reflective of HS susceptibility or tolerance, respectively) 6Classification by period interaction 7Rectal temperature 8 HS TR slope (°C/h) 9Skin temperature 10Respiration rate, breaths per minute 11Feed intake a-dMeans with different letters differ (P ≤ 0.05) 67

Figure 2.1. Box and whisker plots of rectal temperature (TR; A), skin temperature (TS; B), and respiration rate (RR; breaths per minute [bpm]; C) distributions during period (P) 1 (24-hour thermoneutral [TN] conditions) and subsequently P2 (24-hour heat stress [HS]). Considerable variation exists between animals with respect to their thermoregulatory response during HS as compared to TN conditions. Whiskers denote the minimum and maximum value for each P. The bottom and top boundaries of each box represents the first and third quartiles, respectively, while the inner middle line marks the median during each P. 68

150 A r = -0.16; P = 0.03

100

% 0 ΔFI -50

-100

-150 38.5 39.0 39.5 40.0 40.5 41.0 41.5

HS TR, °C

B 3 r = -0.23; P < 0.01 2 1

0 -1 BW% Δ -2 -3 -4 -5 38.5 39.0 39.5 40.0 40.5 41.0 41.5

HS TR, °C

Figure 2.2. Relationships between rectal temperature (TR) and the change (Δ) in feed intake (FI) of pigs during period (P) 2 (heat stress [HS] conditions; A) and of HS TR with the ΔBW (B). The ΔFI and ΔBW variables were determined by subtracting the P2 measurement from P1 (thermoneutral [TN] conditions). 69

Figure 2.3. Correlations between average rectal temperature (TR) and skin temperature (TS) of pigs during period (P) 2 (heat stress [HS] conditions; A) and of TR with respiration rate (RR; B) during P2. Just as with TR, all TS and RR data collected 4h after HS initiation in P2 were condensed into single averages for each experimental unit. 70

Figure 2.4. Correlations between average rectal temperature (TR) of pigs during period (P) 1 (thermoneutral [TN] conditions) and P2 (heat stress [HS] conditions; A) and of TN TR with change (Δ) in TR (B). All TN TR’s recorded during P1 were condensed into a single average to represent each individual pig’s basal core temperature. Rectal temperatures recorded 4h after HS initiation were condensed into an average to represent core body temperature during P2. The difference in core body temperature (ΔTR) was calculated by subtracting TN TR from the HS TR. 71

3 r = 0.37; P < 0.01 2 1 0 -1 BW% Δ -2 -3 -4 -5 -100 -50 0 50 100 ΔFI%

Figure 2.5. Correlations between the percent change (Δ) in feed intake (FI) and the percent ΔBW.

Figure 2.6. Thermoregulatory ability of pigs retrospectively classified as tolerant (TOL) and susceptible (SUS) to HS during period (P) 1 (thermoneutral [TN] conditions) and P2 (heat stress [HS] conditions). Each gilt’s tolerance to the heat load was classified on the basis of HS rectal temperature (TR; i.e. higher and lower HS TR were considered markers of HS susceptibility or tolerance, respectively). For each of the five replications, the 10 most TOL (n = 50) and SUS (n = 50) were identified and allocated to their respective retrospective a-d treatments for TR (A) the percent change in BW (ΔBW; B) data analysis. Values with differing superscripts denote differences (P ≤ 0.05) between treatments. 73

CHAPTER 3. EFFECTS OF HEAT STRESS AND INSULIN SENSITIZERS ON PIG ADIPOSE TISSUE

Modified from a paper published in 2018 by the Journal of Animal Science 96: 510-520

J.T. Seibert*, M. Abuajamieh*, M.V. Sanz Fernandez*, J.S. Johnson*, S.K. Kvidera*, E. A.

Horst*, E. J. Mayorga*, S. Lei*, J.F. Patience*, J.W. Ross*, R.P. Rhoads†, R.C. Johnson‡, S.M.

Lonergan*, J.W. Perfield II§,#, and L.H. Baumgard*

*Department of Animal Science, Iowa State University, Ames, IA, 50011 †Department of Animal and Poultry Sciences, Virginia Tech, Blacksburg, VA, 24061 ‡Smithfield Farmland, Denison, IA, 51442 §Departments of Nutrition and Exercise Physiology, & Food Science, University of Missouri, Columbia, MO 65211 #Current Address: Lilly Research Labs, Eli Lilly and Company, Indianapolis, IN 46285

Corresponding author: Lance H. Baumgard, PhD Department of Animal Science Iowa State University Ames, IA, 50011, United States Phone number: (515) 294-3615 Fax number: (515) 294-3795 E-mail: [email protected]

Authors contributions:

J.T.S conducted the experiment, interpreted the data, drafted the manuscript, and prepared the figures. M.A., M.V.S.F, J.S.J. and S.K.S helped conduct the live phase of the experiment and provided intellectual contributions. S.L. performed fatty acid composition analysis.

J.W.R, R.P.R., R.C.J., S.M.L., J.W.P. provided intellectual contributions key to the success 74 of the experiment. J.T.S. and L.H.B. designed the experiment. All authors edited the manuscript.

Results described here within were supported by the National Pork Board and Agriculture and Food Research Initiative Competitive Grant no. 2011-67003-30007 from the USDA

National Institute of Food and Agriculture.

Abstract

Heat stress (HS) negatively impacts several swine production variables, including carcass fat quality and quantity. Pigs reared in HS have more adipose tissue than energetically predicted, explainable, in part, by HS-induced hyperinsulinemia. Study objectives were to evaluate insulin’s role in altering fat characteristics during HS via feeding insulin-sensitizing compounds. Forty crossbred barrows (113 ± 9 kg BW) were randomly assigned to 1 of 5 environment by diet treatments: 1) thermoneutral (TN) fed ad libitum

(TNAL), 2) TN and pair-fed (TNPF), 3) HS fed ad libitum (HSAL), 4) HS fed ad libitum with sterculic oil (SO) supplementation (HSSO; 13 g/d), and 5) HS fed ad libitum with dietary chromium (Cr) supplementation (HSCr; 0.5 mg/d; Kemin Industries, Des Moines,

IA). The study consisted of 3 experimental periods (P). During P0 (2 d), all pigs were exposed to TN conditions (23 ± 3°C, 68 ± 10% RH) and fed ad libitum. During P1 (7 d), all pigs received their respective dietary supplements, were maintained in TN conditions, and fed ad libitum. During P2 (21 d), HSAL, HSSO, and HSCr pigs were fed ad libitum and exposed to cyclical HS conditions (28 to 33°C, 58 ± 10% RH). The TNAL and TNPF pigs remained in TN conditions and were fed ad libitum or pair-fed to their HSAL counterparts.

Rectal temperature (TR), respiration rate (RR), and skin temperature (TS) were obtained daily 75

at 0600 and 1800 h. At 1800 h, HS exposed pigs had increased TR, RR, and TS relative to

TNAL controls (1.13°C, 48 bpm, and 3.51°C, respectively; P < 0.01). During wk 2 and 3 of

P2, HSSO pigs had increased 1800 h TR relative to HSAL and HSCr (~0.40 and ~0.42°C, respectively; P ≤ 0.05). Heat stress decreased ADFI and ADG compared to TNAL pigs (2.24 vs. 3.28 and 0.63 vs. 1.09 kg/d, respectively; P < 0.01) and neither variable was affected by

SO or Cr supplementation. Heat stress increased or tended to increase moisture content of abdominal (7.7 vs. 5.9%; P = 0.07) and inner s.c. (11.4 vs. 9.8%; P < 0.05) adipose depots compared to TNAL controls. Interestingly, TNPF pigs also had increased adipose tissue moisture content and this was most pronounced in the outer s.c. depot (15.0 vs. 12.2%; P <

0.01) compared to TNAL pigs. Heat stress had little or no effect on fatty acid composition of abdominal, inner, and outer s.c. adipose tissue depots. In summary, the negative effects of HS on fat quality do not appear to be fatty acid composition related, but may be explained by increased adipose tissue moisture content.

Keywords: adipose tissue, fatty acid composition, heat stress, insulin

Introduction

Heat stress (HS) impedes efficient pork production by reducing feed intake, altering metabolism, and ultimately compromising the animal’s ability to express its genetic potential for maximum growth (Baumgard and Rhoads, 2013). Paradoxically, animals, including pigs, reared during HS accumulate more carcass fat than their feed intake predicts (Close et al.,

1971; Collin et al., 2001; Pearce et al., 2013), likely stemming from HS-induced hyperinsulinemia (Baumgard and Rhoads, 2013). However, less is known about how HS influences adipose fatty acid (FA) characteristics. Animals strategically modify FA saturation 76 in an attempt to maintain proper membrane fluidity in response to differing ambient temperatures, a phenomenon referred to as homeoviscous adaptation (Hazel, 1995). The impact of HS on farm animal FA saturation is relevant because it affects meat processing, product quality, and shelf life (Wood et al., 2008).

Cells regulate FA saturation primarily through stearoyl-CoA desaturase (SCD) for which activity and quantity is regulated by insulin (Dobrzyn et al., 2010). Insulin is also a potent adipogenic signal, a proliferative process that can decrease overall adipocyte size and, consequently, enhance fat pliability (Mendizabal et al., 2004). Therefore, we hypothesized altered carcass fat characteristics could be mediated by HS-induced hyperinsulinemia.

Through its interaction with chromodulin, chromium (Cr) potentiates insulin action (Chen et al., 2006; Vincent, 2013) and sterculic oil (SO) also improves insulin sensitivity (Ortinau et al., 2012, 2013), although SO also inhibits SCD activity by binding the enzyme’s active site

(Corl et al., 2001). To investigate insulin’s potential role in the aforementioned parameters, we evaluated the ability of these insulin-sensitizing compounds (Cr and SO) to alter the impact of HS on carcass fat characteristics during a 21-day HS challenge.

Materials and Methods

Animals and Experimental Design

All procedures were approved by the Iowa State University Institutional Animal Care and Use Committee. Forty crossbred barrows (113 ± 9 kg BW) were randomly assigned to 1 of 5 diet by environmental treatments: 1) thermoneutral (TN) conditions and ad libitum fed

(TNAL; n = 8), 2) TN pair-fed (TNPF; n = 8), 3) HS conditions and ad libitum fed (HSAL; n

= 8), 4) HS and ad libitum fed a diet with SO supplementation (HSSO; 13 g/d; n = 8), or 5)

HS ad libitum fed a diet with Cr supplementation (HSCr; 0.5 mg/d, KemTRACE chromium 77 propionate, Kemin Industries, Des Moines, IA; n = 8). Pigs were housed in individual pens

(57 x 221 cm; 24 pens/room) at the Iowa State University Swine Nutrition Farm research facility (Ames, IA). Each pen was equipped with a stainless steel feeder and a nipple drinker.

Water was provided ad libitum during the entire experiment.

All pigs were fed a standard diet consisting mainly of corn and soybean meal formulated to meet or exceed nutrient requirements for energy, amino acids, protein, minerals, and vitamins. (NRC, 2012; Table 3.1). Three dietary supplements were formulated and mixed according to the following specifications: 1) a control supplement consisting of 30 g of a palatable carrier (cookie dough, Do-Biz Foods, LLC, Ames, IA), 2) a homogenized sample of seeds from the Sterculia foetida tree with the palatable carrier (13 g of sterculic seeds and 30 g of cookie dough/d), and 3) a Cr supplement with the palatable carrier (0.5 mg

Cr and 30 g cookie dough/d). Sterculic seeds were obtained from the Montgomery Botanical

Center (Miami, FL), stripped of their seed coat, and minced into < 0.5 cm pieces. The palatable carrier was a strategy to ensure supplement consumption. The SO dose was selected on a metabolic BW basis based on previous rodent reports (Ortinau et al., 2012,

2013). Each supplement was administered per os once daily at 0600 h.

This study was divided into three experimental periods (P): P0, P1, and P2. Period 0

(2 d in length) served as an acclimation P in which all pigs were housed individually in TN conditions (23 ± 3°C, 68 ± 10% relative humidity [RH]) with a 12:12 h light-dark cycle and fed ad libitum. During P1 (7 d), pigs received their respective dietary supplements while in

TN conditions and fed ad libitum. During P2 (21 d), HSAL, HSSO, and HSCr pigs were fed ad libitum and exposed to cyclical HS conditions with ambient temperatures ranging from

33°C (0800 to 1800 h; 56 ± 8% RH) to 28°C (1800 to 0800 h; 60 ± 10% RH). The TNAL 78 and TNPF pigs remained in TN conditions and were fed ad libitum or pair-fed to the HSAL counterparts to eliminate the confounding effect of dissimilar feed intake, respectively. Daily feed intake in P1 was averaged for each HSAL pig and used as a baseline; the decrease in intake during P2 was then calculated as the percentage of ADFI reduction relative to P1 for each d of HS exposure. The percentage of ADFI reduction was averaged for all HSAL pigs per d of heat exposure and applied individually to the baseline of each pig in the TNPF treatment as we have previously described (Sanz Fernandez et al., 2015a; 2015b). The calculated amount of feed was evenly distributed and offered to the TNPF pigs three times daily (~0600, 1200, and 1800 h) in an attempt to minimize gorging induced post-prandial shifts in metabolism. Ambient temperature was controlled but humidity was not governed and both parameters were recorded every 30 min by four data loggers (Lascar EL-USB-2-

LCD, Erie, PA) distributed evenly in each room.

Production and Thermoregulatory Measurements

Daily feed intake was measured during P1 and P2 as feed disappearance. Body weights were obtained at the beginning and the end of P1 and on d 7, 14, and 21 of P2.

Rectal temperature (TR) was measured with a calibrated digital thermometer (ReliOn,

Waukegan, IL, USA), skin temperature (TS) was measured using a calibrated infrared thermometer (IRT207: The Heat Seeker 8:1 Mid-Range Infrared Thermometer, General

Tools, New York, NY), and respiration rate (RR) was determined by counting flank movements during a 15 s interval and multiplied by 4 to obtain breaths/min. All thermal indices were recorded twice daily (0600 and 1800 h) and condensed into weekly AM and PM averages.

Blood Sampling and Analysis

Blood was obtained via jugular venipuncture (10 mL; BD® vacutainers; Franklin

Lakes, NJ; K3EDTA; EDTA) at 0600 h (following thermoregulation measurements and prior to feeding) on d 1 of P1 (before dietary treatment initiation) and d 7 of P1, and at 0600 h on d

8, 15, and 21 of P2. Plasma samples were harvested by centrifugation at 4°C and 2500 x g, aliquoted and stored at -80°C until further analysis. Plasma glucose was measured enzymatically using a commercially available kit (Wako Chemicals USA, Richmond, VA); the intra- and inter-assay coefficients of variation were 13.7 and 10.0%, respectively. An

ELISA kit was used to determine plasma insulin (Mercodia Porcine Insulin ELISA;

Mercodia AB; Uppsala, Sweden); the intra- and inter-assay CVs were 5.8 and 5.5%. Both assays were conducted following the manufacturer’s instructions and were read using a microplate photometer (Hycult Biotech, Uden, Netherlands).

Tissue Collection and Fatty Acid Composition Analysis

At the conclusion of the experiment, pigs were euthanized via captive bolt followed by exsanguination. Abdominal visceral fat as well as inner and outer s.c. adipose from the nape of the neck were immediately collected, snap frozen in liquid nitrogen, and stored at -

80°C until analysis. Back fat thickness at the nape of the neck (above cervical vertebrae) was measured to the nearest 0.1 cm using a ruler.

Lipids from abdominal, inner s.c., and outer s.c. adipose depots (nape of the neck) were extracted and FA methyl esters were prepared and quantified by gas chromatography.

Wet tissue lipid extraction was performed as previously described (Madron et al., 2002) and

FA methyl esters were prepared by transmethylation (Christie, 1982) with modifications

(Chouinard et al., 1999). Fatty acid methyl esters were quantified by a gas chromatograph 80

(Varian GC system 3900, Agilent Technologies, Santa Clara, CA) equipped with a flame- ionization detector and an Agilent DB-23 cyanopropyl capillary column (60 m × 0.25 mm i.d. with 0.15-µm thickness, Agilent Technologies, Santa Clara, CA). Initial oven temperature (50°C) was held for 1 min then ramped at 25°C/min to 175°C and thereafter ramped at 4°C/min to 230°C, where it was held for 8 min. Injector and detector temperatures were maintained at 240°C, and the split ratio was 100:1. Helium carrier gas flow rate through the column was 2 mL/min. Peaks in the chromatogram were identified and quantified using pure methyl ester standards gas liquid chromatography (GLC) 68D and GLC461.

Chromatogram analysis was carried out using Varian Star Chromatography Workstation

Version 5.52. Iodine value was calculated using the following equation as previously described (Kellner et al., 2016):

!"#$%& ()*+& = (.16: 1 × 0.95) + (.18: 1 × 0.86) + (.18: 2 × 1.732)

+ (.18: 3 × 2.616) + (.20: 1 × 0.785) + (.22: 1 × 0.723)

Stearoyl-CoA desaturase indices for palmitic acid (C16:0), stearic acid (C18:0), and total were calculated using the following equations:

(.14: 1 + .16: 1 + .18: 1) =.> $%#&? = (.14: 1 + .16: 1 + .18: 1 + .14: 0 + .16: 0 + .18: 0)

Adipose Moisture Content and Morphology Analysis

Adipose samples were weighed and then dried at 37.7°C (Precision: Division of

Jouan Inc., Winchester, VA) to a constant weight (96 h) to determine moisture content. To determine adipocyte cell size, frozen adipose tissue samples were sent to the University of

Iowa Histology Research Laboratory for sectioning and hematoxylin and eosin staining.

Microscopy was carried out using a microscope (DMI3000 B Inverted Microscope; Leica, 81

Bannockburn, IL) with an attached 12-bit QICAM Fast 1394 camera (QImaging, Surrey, BC) to obtain four images per section using Q Capture Pro software (Surrey, BC, Canada). Raw images were converted to solid contrasting colors using Open Lab software (Perkin Elmer,

Waltham, MA) and area was calculated using Image Pro Plus software (MediaCybernetics,

Rockville, MD). All area measurements were condensed into single averages for each adipose depot per experimental unit.

Statistical Analysis

All data were analyzed using PROC MIXED (version 9.3, SAS Institute Inc., Cary,

NC). Period 2 thermoregulatory and production data were analyzed with an autoregressive covariance structure with wk of experiment as the repeated effect. The model included treatment, wk, and their interaction as fixed effects; BW recorded at the beginning of P1

(prior to dietary treatment initiation) was used as a covariate. Plasma insulin was analyzed with a spatial power law covariance structure with d of the experiment as the repeated effect and treatment, d, and their interaction as fixed effects; insulin levels from the plasma sample obtained at 0600 h of P1D1 (prior to dietary treatment initiation) were used as a covariate.

Adipose tissue moisture content, back fat thickness, adipocyte area, and adipose FA content were analyzed using BW recorded at the beginning of P1 as a covariate. Preplanned orthogonal contrasts were conducted to evaluate differences among environmental treatments

(i.e. TNAL vs. HS [HSAL, HSSO, and HSCr], TNPF vs. HS, and TNAL vs. TNPF pigs) and dietary supplements (i.e. HSAL vs. HSSO, and HSAL vs. HSCr pigs). Data are reported as 82

LSmeans and statistical significance (P ≤ 0.05) and tendency thresholds (0.05 < P ≤ 0.10) were utilized for interpretation.

Results

Thermoregulatory, Production, and Blood Indices

During P1, no treatment differences were detected for any thermoregulatory measurements. Although AM TR was not different (Fig. 3.1A) during P2, HS increased AM

RR and TS (20 bpm and 1.91°C, respectively; P < 0.01; Table 3.2) and PM TR (Fig. 3.1B),

RR, and TS (1.13°C, 48 bpm, and 3.51°C, respectively; P < 0.01; Table 3.2) relative to

TNAL controls. While AM and PM RR and TR were not different, AM and PM TS (1.49 and

1.18°C, respectively; P < 0.01) were decreased in TNPF compared to TNAL controls during

P2. Heat-stressed pigs fed SO had increased PM TR relative to HSAL during wk 2 and 3 of

P2 (0.36°C and 0.44°C, respectively; P ≤ 0.05; Fig. 3.1B), but Cr did not affect thermal indices.

No dietary treatment differences were detected for ADFI, BW, or ADG during P1.

During P2, HS decreased ADFI compared to TNAL pigs (32%; P < 0.01; Table 3.3).

Similarly, ADG and final BW were decreased in HS treatments compared to TNAL pigs (42 and 7%, respectively; P < 0.01; Table 3.3); neither ADFI nor ADG were influenced by SO or

Cr supplementation (Table 3.3). By experimental design, TNPF pigs had a similar magnitude and pattern of reduced ADFI and their ADG, BW and G:F variables did not differ from their

HS counterparts. There was also no overall treatment effect on G:F (P = 0.20; Table 3.3).

During P1, no dietary treatment differences in circulating insulin or glucose were detected. Relative to TNAL pigs during P2, HS decreased circulating insulin (25%; P <

0.01); however, circulating insulin increased in HS pigs relative to TNPF controls (33%; P =

0.02; Fig. 3.2A). Circulating insulin was not influenced by SO (P = 0.12), but increased with 83

Cr supplementation relative to HSAL controls (0.13 vs. 0.10 µg/L; P = 0.05; Fig. 3.2A).

During P2, circulating glucose increased with HS compared to TNAL pigs (13%; P < 0.01), but it did not differ from TNPF controls (P = 0.24; Fig. 3.2B). Sterculic oil supplementation did not alter circulating glucose relative to HSAL pigs (P = 0.33), but plasma glucose decreased with dietary Cr (8%; P = 0.01; Fig. 3.2B).

Fatty Acid Composition

There were marginal environmental effects on fatty acid composition in the three adipose depots evaluated (Table 3.4, 3.5, and 3.6). The primary dietary effect on fatty acid composition was the decrease in SCD products (C14:1, C16:1 and C18:1) in all three depots from the HSSO fed pigs. Consequently, the total saturated fatty acid and monounsaturated fatty acid content increased and decreased, respectively, in each of the adipose depots from the SO fed pigs (Table 3.7).

Back Fat Thickness, Adipocyte Area, and Adipose Moisture Content

Back fat thickness was not influenced by diet or environment (P = 0.87; Table 3.7).

However, abdominal adipocyte size was decreased in TNPF pigs relative to TNAL controls

(26%; P = 0.02; Table 3.7). In HSAL pigs, abdominal adipoycte size was increased compared to TNPF controls (37%; P = 0.01), but it did not differ compared to TNAL fed pigs (P = 0.74; Table 3.7). Sterculic oil and Cr supplementation decreased abdominal adipocyte size relative to HSAL (22 and 24%, respectively; P ≤ 0.04; Table 3.7). No differences in adipocyte size were detected at inner and outer s.c. adipose depots (Table 3.7).

Regardless of SO and Cr supplementation, HS increased moisture content in the abdominal (30%; P = 0.07) and inner s.c. (16%; P = 0.05; Table 3.7 and Fig. 3.3) adipose 84 depots relative to TNAL controls. Similarly, TNPF pigs had increased moisture content in the outer s.c. (23%; P = 0.01; Table 3.7 and Fig. 3.3) adipose depot relative to TNAL controls; however, moisture content in outer s.c. adipose tissue was decreased in HS treatments compared to TNPF (11%; P = 0.04; Table 3.7 and Fig. 3.3).

Discussion

Despite aggressive heat abatement strategies, HS remains a major economic burden to the U.S. swine industry with an estimated $900 million in annual losses during the warm summer months (Pollmann, 2010). Sources of reduced revenue include slower growth rates, inefficient facility utilization, increased health care costs, inconsistent market weights, mortality, and altered carcass composition (Baumgard and Rhoads, 2013; Ross et al., 2017).

In addition, post-harvest adipose tissue is softer (also referred to as “flimsy fat”) from pigs marketed during the summer, and this creates processing and handling complications (Dr. R.

Johnson, Smithfield Farmland, Denison, IA; personal communication).

In this experiment, pigs allocated to the three HS treatments experienced a significant heat load, which was reflected by marked thermoregulatory responses (Table 3.2).

Interestingly, supplementing SO increased TR relative to the other two HS treatments during the last two weeks of P2. The reduced desaturase index at all adipose depots in HSSO pigs suggested SO supplementation attenuated SCD activity, which is responsible for inserting a double bond at the 9th position of myristic, palmitic, and stearic acids. Importantly, this enzyme has also been implicated in thermoregulation as SCD knockout mice have impaired thermoregulatory capacity and become critically hypothermic when housed in cold environments (Lee et al., 2004). Disrupting SCD with dietary SO also impaired 85 thermoregulation in this experiment and identifying whether or not SCD could be manipulated to help pigs maintain a healthy body temperature during HS is of interest.

Heat stress markedly decreased ADFI, ADG, and final BW. Interestingly, while ADG and G:F from wk 2 to 3 (data not shown) plateaued in pigs from the HSAL and HSSO treatments, both parameters continued to increase (30% from wk 2 to 3) in HSCr pigs.

Improvements in ADG and G:F have been observed in pigs fed Cr (Lindemann et al., 1995;

Hung et al., 2010; Sales and Jancik, 2011; Mayorga et al., 2016). While evaluating the effect of dietary supplements on traditional production parameters was not our primary objective, it would be interesting to investigate ADG and G:F over a longer duration to determine if HSCr pigs maintained their improvements in these economically important phenotypes.

The experiment’s primary objectives were to investigate the effects of HS and insulin sensitization on carcass fat characteristics. Developmentally, the inner portion of s.c. adipose is the most recently synthesized (Fortin 1986), so we separated this particular depot into outer (older) and inner (newest) sections because we hypothesized that the inner would be most responsive to a 21-day environmental and dietary intervention. Despite inconsistencies with respect to how HS influences adipose FA composition (Kloareg et al., 2005; White et al., 2008; Kellner et al., 2016), the potential for heat-induced FA unsaturation via increased

SCD activity represents a possible underlying cause for the aforementioned “flimsy fat” phenotype. The very name suggests there would be an overall increase in FA unsaturation, which would reduce firmness due to reduced melting points of MUFA and PUFA.

Furthermore, we hypothesized that the heat-induced increased circulating insulin would upregulate SCD abundance and activity, a scenario which should promote FA desaturation.

However, cellular membranes generally become more fluid-like at higher ambient 86 temperatures (Hazel, 1995) and therefore it would be biologically advantageous to employ mechanisms to increase FA saturation in order to maintain membrane integrity and plasticity.

In agreement with this homeoviscous adaptation concept, adipose from HSAL had decreased

MUFAs, but numerically increased SFA in all adipose depots compared to TNAL. Heat- induced increases in saturated FA content in pigs has also been observed elsewhere

(Lefaucheur et al., 1991; Katsumata et al., 1995; Kouba et al., 1999; Kloareg et al., 2005).

Furthermore, pigs reared in colder conditions exhibit a higher degree of FA unsaturation

(MacGrath et al., 1968; Fuller et al., 1974; Le Dividich et al., 1987; Lefaucheur et al., 1991).

Therefore, the summer affiliated “flimsy fat” phenotype occuring as a result of increased desaturation appears unlikely as the data reported herein conforms with the homeoviscous adaptation concept.

We also evaluated adipocyte size because it was thought to be associated with adipose tissue firmness (Mendizabal et al., 2004). Due to insulin’s ability to stimulate adipocyte proliferation (Geloen et al., 1989) we hypothesized heat-induced insulin response would result in more but smaller adipocytes. Although HS increased circulating insulin, it had little or no effect on adipocyte size, but feed restriction and SO and Cr supplementation decreased adipocyte size in the abdominal adipose depot, while no differences were detected at inner or outer s.c. adipose tissue locations. Previous studies have reported HS-induced increases in back fat adipocyte diameter in pigs (Rinaldo and Le Dividich, 1991) and cold stress-induced decreases in epididymal adipocyte diameter in rats (Cherqui et al., 1979).

Differences in experimental design (e.g., pattern, extent, adipose depot type, and magnitude of HS) may contribute to our lack of observed changes. 87

Interestingly, we observed HS-induced increased moisture content at all adipose depots relative to TNAL controls. Although the exact mechanisms for this are not clear, water content has been shown to be inversely related to s.c. adipose tissue firmness in boars and barrows (Wood et al., 1985), thus increased moisture content may at least partly contribute to soft carcass fat phenotypes observed during the summer months. This may be partially due to decreased feed intake, as pair-fed animals also had increased moisture content in outer s.c. adipose tissue relative to TNAL controls. Furthermore, obese humans administered a low-calorie diet have increased water content of abdominal adipose tissue

(Laaksonen et al., 2003). However, abdominal and s.c. adipose tissue water content was not affected by nutrient restriction in the current study, indicating a potential interaction of feed intake, environment, and adipose depot location. Further investigation into mechanisms and biological reasons for increased adipose water content during HS and nutrient restriction warrant additional research. Increased adipose tissue blood flow could also represent a explanation for increased adipocyte moisture content as obese humans have lower adipose tissue perfusion compared to lean individuals (Blaak et al., 1995) and this is improved with low calorie diet administration and weight loss (Blaak et al., 1995; Barbe et al., 1997).

Regardless of the potential mechanism, increased carcass adipose moisture content is not trivial as it directly affects adipose firmness and, ultimately, carcass processing. Whether or not increased adipose moisture content as a result of HS can be prevented will require a more comprehensive investigation.

Conclusion

In summary, despite the occurrence of the “flimsy fat” phenotype in response to HS, the overall degree of FA unsaturation was not increased in HS pigs at abdominal, inner, and outer s.c. adipose depots. Adipocyte size was not affected by HS, but was impacted by 88 nutrient restriction as well as both Cr and SO. Interestingly, all HS treatments exhibited increased adipose moisture content at all three adipose depots, but this may be partly attributed to reduced ADFI. Whether or not increased moisture content is linked to the altered carcass fat quality observed during the summer months is of scientific and practical interest, but will require more detailed investigation

Acknowledgements

The authors would like to acknowledge the assistance of Matthew Garrick, Sarah

Edwards, Christy Calderwood, Kelsi Young, Chris Almond, Tyler Leete, Kenton Doty, and

Brittney Yehling in collecting temperature and production data. The authors would also like to thank Kemin Industries for providing the Cr and to Montgomery Botanical Center for providing the raw sterculic seeds.

Literature Cited

Barbe, P., V. Stich, J. Galitzky, M. Kunesova, V. Hainer, M. Lafontan, and M. Berlan. 1997. In vivo increase in beta-adrenergic lipolytic response in subcutaneous adipose tissue of obese subjects submitted to a hypocaloric diet. J. Clin. Endocrinol. Metab. 82:63- 69.

Baumgard, L. H. and R. P. Rhoads. 2013. Effects of heat stress on postabsorptive metabolism and energetics. Annu. Rev. Anim. Biosci. 1:311-337.

Blaak, E. E., M. A. van Baak, G. J. Kemerink, M. T. Pakbiers, G. A. Heidendal, and W. H. Saris. 1995. Beta-adrenergic stimulation and abdominal subcutaneous fat blood flow in lean, obese, and reduced-obese subjects. Metabolism. 44:183-187.

Chen, G., P. Liu, G. R. Pattar, L. Tackett, P. Bhonagiri, A. B. Strawbridge, and J. S. Elmendorf. 2006. Chromium activates glucose transporter 4 trafficking and enhances insulin-stimulated glucose transport in 3T3-L1 adipocytes via a cholesterol- dependent mechanism. Mol. Endocrinol. 20:857-870.

Cherqui, G., M. Cadot, C. Senault, and R. Portet. 1979. Cellularity and composition of epididymal adipose tissue from cold-acclimatized rats. Experientia. 35:1353-1354. 89

Chouinard, P. Y., L. Corneau, A. Saebo, and D. E. Bauman. 1999. Milk yield and composition during abomasal infusion of conjugated linoleic acids in dairy cows. J. Dairy Sci. 82:2737-2745.

Christie, W. W. 1982. A simple procedure for rapid transmethylation of glycerolipids and cholesteryl esters. J. Lipid Res. 23:1072-1075.

Close, W. H., L. E. Mount, and I. B. Start. 1971. The influence of environmental temperature and plane of nutrition on heat losses from groups of growing pigs. Anim. Prod. 13:285–294.

Collin, A., J. van Milgen, S. Dubois, and J. Noblet. 2001. Effect of high temperature and feeding level on energy utilization in piglets. J. Anim. Sci. 79:1849–1857.

Corl, B. A., L. H. Baumgard, D. A. Dwyer, J. M. Griinari, B. S. Phillips, and D. E. Bauman. 2001. The role of Δ 9-desaturase in the production of cis-9, trans-11 CLA. J. Nutr. Biochem. 12:622-630.

Dobrzyn, P., M. Jazurek, and A. Dobrzyn. 2010. Stearoyl-CoA desaturase and insulin signaling--what is the molecular switch? Biochim. Biophys. Acta. 1797:1189-1194.

Fortin, A. 1986. Development of backfat and individual fat layers in the pig and its relationship with carcass lean. Meat Sci. 18:255-270.

Fuller, M. F., W. R. Duncan, and A. W. Boyne. 1974. Effect of environmental temperature on the degree of unsaturation of depot fats of pigs given different amounts of food. J. Sci. Food Agric. 25:205-210.

Geloen, A., A. J. Collet, G. Guay, and L. J. Bukowiecki. 1989. Insulin stimulates in vivo cell proliferation in white adipose tissue. Am. J. Physiol. 256:C190-196.

Hazel, J. R. 1995. Thermal adaptation in biological membranes: is homeoviscous adaptation the explanation? Annu. Rev. Physiol. 57:19-42.

Hung, T. Y., B. J. Leury, T. F. Lien, J. J. Lu, and F. R. Dunshea. 2010. Potential of nano- chromium to improve body composition and performance of pigs. In: Proc. 14th Anim. Sci. Congr. Pington, Taiwan. p:108-112.

Katsumata, M., H. Hirose, Y. Kaji, and M. Saitoh. 1995. Influence of a high ambient temperature and dietary fat supplementation on fatty acid composition of depot fats in finishing pigs. J. Anim. Sci. Technol. 66:225-231.

Kellner, T. A., L. H. Baumgard, K. J. Prusa, N. K. Gabler, and J. F. Patience. 2016. Does heat stress alter the pig's response to dietary fat? J. Anim. Sci. 94:4688-4703.

Kloareg, M., L. Le Bellego, J. Mourot, J. Noblet, and J. van Milgen. 2005. Deposition of dietary fatty acids and of de novo synthesised fatty acids in growing pigs: effects of high ambient temperature and feeding restriction. Br. J. Nutr. 93:803-811.

Kouba, M., D. Hermier, and J. Le Dividich. 1999. Influence of a high ambient temperature on stearoyl-CoA-desaturase activity in the growing pig. Comp. Biochem. Physiol. B. Biochem. Mol. Biol. 124:7-13.

Laaksonen, D. E., J. Nuutinen, T. Lahtinen, A. Rissanen, and L. K. Niskanen. 2003. Changes in abdominal subcutaneous fat water content with rapid weight loss and long-term weight maintenance in abdominally obese men and women. Int. J. Obes. Relat. Metab. Disord. 27:677-683.

Le Dividich, J., J. Noblet, and T. Bikawa. 1987. Effect of environmental temperature and dietary energy concentration on the performance and carcass characteristics of growing-finishing pigs fed to equal rate of gain. Livest. Prod. Sci. 17:235-246.

Lee, S. H., A. Dobrzyn, P. Dobrzyn, S. M. Rahman, M. Miyazaki, and J. M. Ntambi. 2004. Lack of stearoyl-CoA desaturase 1 upregulates basal thermogenesis but causes hypothermia in a cold environment. J. Lipid Res. 45:1674-1682.

Lefaucheur, L., J. Le Dividich, J. Mourot, G. Monin, P. Ecolan, and D. Krauss, 1991. Influence of environmental temperature on growth, muscle and adipose tissue metabolism, and meat quality in swine. J. Anim. Sci. 69:2844-2854.

Lindemann, M. D., C. M. Wood, A. F. Harper, E. T. Kornegay, and R. A. Anderson. 1995. Dietary chromium picolinate additions improve gain:feed and carcass characteristics in growing finishing pigs and increase litter size in reproducing sows. J. Anim. Sci. 73:457-465.

MacGrath, W. S., Jr., G. W. Noot, R. L. Gilbreath, and H. Fisher. 1968. Influence of environmental temperature and dietary fat on backfat composition of swine. J. Nutr. 96:461-466.

Madron, M. S., D. G. Peterson, D. A. Dwyer, B. A. Corl, L. H. Baumgard, D. H. Beermann, and D. E. Bauman. 2002. Effect of extruded full-fat soybeans on conjugated linoleic acid content of muscle fat in beef steers. J. Anim. Sci. 80:1135-1143.

Mayorga, E. J., S. K. Stoakes, J. Seibert, E. A. Horst, M. Abuajamieh, S. Lei, L. Ochoa, B. Kremer and L. H. Baumgard. 2016. Effects of dietary chromium propionate during heat stress on finishing pigs. J. Anim. Sci. 94(E-Suppl. 1):334. (Abstr.).

Mendizabal, J. A., M. Thériez, P. Bas, J. Normand, B. Aurousseau, and A. Purroy. 2004. Fat firmness of subcutaneous adipose tissue in intensively reared lambs. Small Rumin. Res. 53:173-180.

NRC. 2012. Nutrient Requirements of Swine. 11th rev. ed. Natl. Acad. Press, Washington, DC.

Ortinau, L. C., R. T. Pickering, K. J. Nickelson, K. L. Stromsdorfer, C. Y. Naik, R. A. Haynes, D. E. Bauman, R. S. Rector, K. L. Fritsche, and J. W. Perfield. 2012. Sterculic oil, a natural SCD1 inhibitor, improves glucose tolerance in obese ob/ob Mice. ISRN Endocrinol 2012:947323.

Ortinau, L. C., K. J. Nickelson, K. L. Stromsdorfer, C. Y. Naik, R. T. Pickering, R. A. Haynes, K. L. Fritsche, and J. W. Perfield 2013. Sterculic oil, a natural inhibitor of SCD1, improves the metabolic state of obese OLETF rats. Obesity (Silver Spring) 21:344-352.

Pollmann, D. S. Seasonal effects on sow herds: industry experience and management strategies. 2010. J. Anim. Sci. 88(Suppl. 3):9. (Abstr).

Rinaldo, D., and J. Le Dividich. 1991. Effects of warm exposure on adipose tissue and muscle metabolism in growing pigs. Comp. Biochem. Physiol. A. Comp. Physiol. 100:995-1002.

Ross, J. W., J. T. Seibert, M. K. Adur, M. R. Romoser, L. H. Baumgard, and A. F. Keating. 2017. Reproductive consequences of heat stress in pigs. Mol. Reprod. Dev. 84:934- 945.

Sales, J. and F. Jancik. 2011. Effects of dietary chromium supplementation on performance, carcass characteristics, and meat quality of growing-finishing swine: a meta-analysis. J. Anim. Sci. 89:4054-4057.

Vincent, J. B. 2013. The bioinorganic chemistry of chromium. John Wiley, Chichester, West Sussex England.

White, H. M., B. T. Richert, A. P. Schinckel, J. R. Burgess, S. S. Donkin, and M. A. Latour. 2008. Effects of temperature stress on growth performance and bacon quality in grow-finish pigs housed at two densities. J. Anim. Sci. 86: 1789-1798.

Wood, J. D., M. Enser, A. V. Fisher, G. R. Nute, P. R. Sheard, R. I. Richardson, S. I Hughes, and F. M. Wittington. 2008. Fat deposition, fatty acid composition and meat quality. A review. Meat Sci. 78:343-358.

Wood, J. D., R. C. D. Jones, J. A. Bayntuna, and E. Dransfield. 1985. Backfat quality in boars and barrows 90 kg live weight. Anim. Prod. 40:481-487. 93

Tables and Figures

Table 3.1. Ingredients and chemical composition of diet for growing pigs (as-fed basis) Ingredients ---- % ---- Corn 73.77 Soybean meal 9.36 Dried distillers grains 15.00 45-30 vitamin and mineral premix1 1.65 L-lysine HCL 0.22 Calculated chemical composition % DM 87.3 Crude protein 17.45 Crude fat 3.25 ADF 4.42 NDF 12.06 Ash 3.75 10.97% Limestone, 0.37% Salt, 0.18% Dried distillers grains, 0.11% Vitamin and Trace Mineral (Provided 7,279 IU vitamin A, 1,335 IU vitamin D, 39 IU vitamin E, 2 IU vitamin K, 19 mg niacin, 15 mg pantothenic acid, 4 mg riboflavin, 4 mg choline, 0.4 µg folic acid, 24 µg vitamin B12, 1 µg biotin, 214 ppm zinc, 103 ppm manganese, 278 ppm iron, 39 ppm copper, 2 ppm iodine, 0.5 ppm selenium per kg of diet), 0.02% Rono M 10,000

Table 3.2 Effects of period 2 treatment on body temperature indices. a-dValues with differing superscripts denote differences (P ≤ 0.05) between treatments

P Treatments1 TNAL TNPF TNAL HSAL HSAL Parameter TNAL TNPF HSAL HSSO HSCr SEM Trt2 Wk Trt*Wk3 vs. HS4 vs. HS vs. TNPF vs. HSSO vs. HSCr 5 c TR AM , °C 39.69 39.62 39.73 39.74 39.69 0.07 0.82 <0.01 0.15 0.73 0.27 0.53 0.89 0.72 6 a b c c c TS AM , °C 30.95 29.46 32.91 32.96 32.70 0.38 <0.01 0.88 <0.01 <0.01 <0.01 <0.01 0.92 0.69 RR AM7, bpm 57.54a 50.93a 76.26b 77.13b 77.87b 2.41 <0.01 0.01 0.18 <0.01 <0.01 0.18 0.86 0.74 8 a a b c b TR PM , °C 39.91 39.75 40.85 41.18 40.84 0.12 <0.01 <0.01 0.05 <0.01 <0.01 0.37 0.07 0.94 9 a b c c c TS PM , °C 33.13 31.95 36.13 36.13 35.90 0.22 <0.01 <0.01 <0.01 <0.01 <0.01 <0.01 0.99 0.47 RR PM10, bpm 59.28a 53.34a 102.74b 107.46b 103.36b 3.60 <0.01 <0.01 <0.01 <0.01 <0.01 0.25 0.36 0.90 1Treatments: TNAL = thermoneutral (TN) ad libitum; TNPF = TN pair-fed; HSAL = heat stress (HS) ad libitum; HSSO = HS sterculic oil; HSCr = HS chromium 2Treatment 3Treatment by wk interaction 4All HS treatments 5 Rectal temperature (TR) at 0600 6 Skin temperature (TS) at 0600 7Respiration rate (RR) in breaths per minute (bpm) at 0600 8 TR at 1800 9 TS at 1800 10RR in bpm at 1800 94

Table 3.3. Effects of period 2 treatment on production parameters. a-dValues with differing superscripts denote differences (P ≤ 0.05) between treatments

P Treatments1 TNAL TNPF TNAL HSAL HSAL Parameter TNAL TNPF HSAL HSSO HSCr SEM Trt2 Wk Trt*Wk3 vs. HS4 vs. HS vs. TNPF vs. HSSO vs. HSCr ADFI, kg/d 3.28a 002.46b 002.29c 002.31c 002.11c 0.14 <0.01 <0.01 0.53 <0.01 0.20 <0.01 0.90 0.39 FBW5, kg 143.5a 134.9b 134.2b 133.9b 133.3b 1.8 <0.01 - - <0.01 0.60 <0.01 0.91 0.75 ADG, kg/d 1.09a 0.62b 000.62b 000.65b 000.62b 0.05 <0.01 <0.01 0.12 <0.01 0.85 <0.01 0.78 0.99 G:F 0.33 0.25 000.26 000.27 000.30 0.03 0.20 <0.01 0.08 0.07 0.39 0.03 0.94 0.39 1Treatments: TNAL = thermoneutral (TN) ad libitum; TNPF = TN pair-fed; HSAL = heat stress (HS) ad libitum; HSSO = HS sterculic oil; HSCr = HS chromium 2Treatment 3Treatment by wk interaction 4All HS treatments 5Final BW

Table 3.4. Effects of period 2 treatment on fatty acid (FA) profile of abdominal adipose tissue. a-dValues with differing superscripts denote differences (P ≤ 0.05) between treatments P Treatments1 TNAL TNPF TNAL HSAL HSAL FA TNAL TNPF HSAL HSSO HSCr SEM Trt2 vs. HS3 vs. HS vs. TNPF vs. HSSO vs. HSCr IV4 54.8000 55.5200 56.4600 52.4400 56.3600 1.57 0.40 0.88 0.80 0.74 0.09 0.97 16:1/16:0 0.06bc 0.06c0 0.05ab 0.04a0 0.06bc <0.01 <0.01 0.03 0.01 0.74 0.12 0.29 18:1/18:0 1.90b0 1.79b0 1.63b0 1.26a0 1.77b0 0.12 <0.01 0.02 0.10 0.53 0.04 0.42 SCDI (14,16,18)5 0.45b0 0.44b0 0.41b0 0.36a0 0.43b0 0.01 <0.01 0.01 0.03 0.69 0.02 0.49 PUFA/SFA6 0.2800 0.3000 0.3300 0.3200 0.3200 0.02 0.58 0.13 0.48 0.50 0.65 0.67 Omega-3/omega-6 0.0400 0.0400 0.0400 0.0400 0.0400 <0.01 0.87 0.82 0.35 0.54 0.51 0.55 MUFA/SFA7 0.82b0 0.79b0 0.72b0 0.61a0 0.76b0 0.05 0.03 0.03 0.08 0.66 0.10 0.52 UFA/SFA8 1.1000 1.1000 1.0500 0.9200 1.0800 0.06 0.30 0.27 0.31 0.95 0.17 0.76 ––––––––––– g/100 g fatty acids ––––––––––– 12:0 0.0700 0.0700 0.0800 0.0800 0.0700 0.01 0.25 0.17 0.28 0.83 0.81 0.24 14:0 1.2800 1.2800 1.4000 1.4000 1.4600 0.07 0.29 0.09 0.09 0.99 0.99 0.59 15:0 0.0300 0.0300 0.0500 0.0500 0.0300 0.01 0.16 0.39 0.95 0.95 0.61 0.12 16:0 25.8500 25.4600 25.9200 25.7700 26.3700 0.61 0.88 0.82 0.44 0.65 0.87 0.61 16:1 1.51b0 1.53b0 1.30b0 1.02a0 1.46b0 0.09 <0.01 0.02 0.01 0.88 0.04 0.20

17:0 0.24a0 0.25a0 0.34b0 0.32b0 0.28ab 0.02 0.03 0.02 0.03 0.89 0.62 0.12 96

17:1 0.1300 0.1600 0.1600 0.1400 0.1600 0.02 0.67 0.35 0.66 0.27 0.36 0.72 18:0 19.00a0 19.31a0 19.86a0 24.40b0 19.01a0 0.89 <0.01 0.05 0.10 0.81 <0.01 0.51 18:1 c9 35.72c0 34.38bc 31.98b0 28.10a0 33.48bc 1.07 <0.01 <0.01 0.02 0.38 0.02 0.33 18:2 n6 12.0600 13.1900 14.8300 14.6300 13.9700 0.74 0.08 <0.01 0.14 0.28 0.86 0.42 18:3 n3 0.4400 0.4800 0.5400 0.5500 0.5200 0.03 0.10 0.01 0.15 0.39 0.92 0.60 20:0 0.28bc 0.31c0 0.26ab 0.28bc 0.23a0 0.01 <0.01 0.16 <0.01 0.07 0.42 0.07 20:1 0.62b0 0.58b0 0.58b0 0.47a0 0.56b0 0.03 0.05 0.03 0.24 0.41 0.03 0.64 20:2 n6 0.4700 0.4900 0.5300 0.5300 0.4900 0.03 0.67 0.25 0.53 0.69 0.90 0.39 20:3 n6 0.0600 0.0500 0.0700 0.0800 0.0600 0.01 0.43 0.35 0.14 0.56 0.32 0.75 22:1 0.2000 0.2400 0.2700 0.2500 0.2500 0.02 0.15 0.02 0.53 0.18 0.37 0.44 Unidentified9 2.17b0 2.60c0 2.02ab 1.99ab 1.67a0 0.16 <0.01 0.13 <0.01 0.06 0.90 0.13 Chain length <16 1.3500 1.3400 1.5000 1.4900 1.4900 0.08 0.38 0.12 0.10 0.95 0.96 0.92 Chain length >16 69.1100 69.0600 69.2700 69.7400 69.0200 0.65 0.94 0.76 0.71 0.96 0.62 0.79 SFA10 46.65a0 46.53a0 47.79a0 52.26b0 47.38a0 1.40 0.05 0.14 0.12 0.95 0.04 0.84 MUFA11 38.15c0 36.80bc 34.24b0 29.97a0 35.91bc 1.14 <0.01 <0.01 0.02 0.41 0.02 0.31 PUFA12 13.0300 14.0700 15.9600 15.7800 15.0500 0.80 0.07 <0.01 0.11 0.36 0.88 0.43

Table 3.4 continued P Treatments1 TNAL TNPF TNAL HSAL vs. HSAL vs. FA TNAL TNPF HSAL HSSO HSCr SEM Trt2 vs. HS3 vs. HS vs. TNPF HSSO HSCr Omega-3 FA 0.4400 0.4800 0.5400 0.5500 0.5200 0.03 0.12 0.02 0.15 0.42 0.92 0.60 Omega-6 FA 12.5900 13.6500 15.4200 15.2400 14.5300 0.76 0.07 <0.01 0.12 0.33 0.88 0.42 1Treatments: TNAL = thermoneutral (TN) ad libitum; TNPF = TN pair-fed; HSAL = heat stress (HS) ad libitum; HSSO = HS sterculic oil; HSCr = HS chromium 2Treatment 9Total unidentified peaks 3All HS treatments 10Total saturated FA 4Iodine value 11Total monounsaturated FA 5Stearoyl-CoAdesaturase index 12Total polyunsaturated FA 6Polyunsaturated to saturated FA ratio 7Monounsaturated to saturated FA ratio 8Unsaturated to saturated FA ratio

Table 3.5 Effect of period 2 treatment on fatty acid (FA) profile of inner s.c. adipose tissue. a-dValues with differing superscripts denote differences (P ≤ 0.05) between treatments P Treatments1 TNAL TNPF TNAL HSAL HSAL FA TNAL TNPF HSAL HSSO HSCr SEM Trt2 vs. HS3 vs. HS vs. TNPF vs. HSSO vs. HSCr IV4 68.6500 68.0000 68.0500 64.3200 69.3000 <1.40 <0.13 0.38 0.63 0.74 0.07 0.53 16:1/16:0 0.08bc 0.08c 0.07ab 0.06a0 0.07abc <0.01 <0.02 0.04 0.01 0.58 0.29 0.65 18:1/18:0 3.20c0 2.98bc 2.62b0 1.98a0 2.84bc <0.17 <0.01 <0.01 0.02 0.37 0.01 0.38 SCDI (14,16,18)5 0.54c0 0.53bc 0.51b0 0.46a0 0.52bc <0.01 <0.01 <0.01 0.02 0.40 <0.01 0.62 PUFA/SFA6 0.5500 0.5500 0.5500 0.5000 0.5600 <0.03 <0.62 0.72 0.81 0.93 0.23 0.78 Omega-3/omega-6 0.04 0.0400 0.0400 0.0400 0.0400 <0.01 <0.38 0.21 0.26 0.92 0.71 0.18 MUFA/SFA7 1.21c 1.14bc 1.06b0 0.87a0 1.09b0 <0.04 <0.01 <0.01 0.01 0.26 0.01 0.63 UFA/SFA8 1.76b 1.69b0 1.61b0 1.37a0 1.65b0 <0.07 <0.01 0.01 <0.0700 0.44 0.02 0.66 ––––––––––– g/100 g fatty acids ––––––––––– 12:0 0.0600 0.0600 0.0600 0.0700 0.0600 <0.01 <0.16 0.20 0.32 0.69 0.03 0.37 14:0 1.0400 1.1100 1.0800 1.1800 1.2000 <0.06 <0.27 0.09 0.46 0.43 0.22 0.17 15:0 0.0500 0.0500 0.0500 0.0500 0.0600 <0.01 <0.13 0.31 0.03 0.58 0.89 0.07 16:0 21.2300 21.2500 21.2900 22.3400 21.9100 <0.56 <0.54 0.35 0.37 0.98 0.19 0.44 16:1 1.6300 1.7300 1.4000 1.2700 1.5100 <0.12 <0.09 0.10 0.02 0.58 0.46 0.54

17:0 0.3300 0.3300 0.4100 0.4000 0.3900 <0.03 <0.10 0.04 0.02 0.89 0.90 0.70 98

17:1 0.2500 0.2600 0.2700 0.2200 0.2500 <0.02 <0.40 0.78 0.54 0.79 0.06 0.38 18:0 12.58a0 12.93a0 14.31a0 17.48b0 13.28a0 <0.83 <0.01 0.01 0.03 0.76 0.01 0.38 18:1 c9 39.59c0 37.81bc 36.50b0 33.01a0 37.35b0 <0.73 <0.01 <0.01 0.01 0.09 <0.01 0.42 18:2 n6 17.5400 17.9400 18.7500 18.5000 19.0600 <0.68 <0.54 0.13 0.31 0.68 0.80 0.75 18:3 n3 0.7200 0.7500 0.7800 0.7600 0.7600 <0.03 <0.75 0.20 0.71 0.46 0.70 0.65 20:0 0.19ab 0.23d0 0.22cd 0.21bc 0.17a0 <0.01 <0.01 0.34 0.01 0.01 0.46 <0.01 20:1 0.78c0 0.73bc 0.75c0 0.59a0 0.67ab <0.03 <0.01 <0.01 0.13 0.22 <0.01 0.10 20:2 n6 0.8200 0.8200 0.8400 0.7700 0.8300 <0.04 <0.72 0.91 0.86 0.96 0.19 0.80 20:3 n6 0.14b0 0.11a0 0.15b0 0.13b0 0.15b0 <0.01 <0.01 0.77 <0.01 0.01 0.26 0.93 22:1 0.2900 0.3200 0.3400 0.3200 0.3200 <0.02 <0.76 0.22 0.86 0.38 0.61 0.64 Unidentified9 2.96b0 3.64c0 2.82b0 2.75b0 2.13a0 <0.22 <0.01 0.14 <0.01 0.04 0.81 0.04 Chain length <16 1.0900 1.1900 1.1900 1.2600 1.2500 <0.06 <0.29 0.04 0.54 0.24 0.40 0.53 Chain length >16 73.0900 72.1900 73.3000 72.3800 73.2100 <0.64 <0.64 0.86 0.31 0.33 0.32 0.92 SFA10 35.34a0 35.90a0 37.41a0 41.69b0 36.98a0 <1.22 <0.01 0.02 0.06 0.75 0.02 0.81 MUFA11 42.48c0 40.86bc 39.26b0 35.41a0 40.09b0 <0.82 <0.01 <0.01 0.01 0.17 <0.01 0.48 PUFA12 19.2100 19.6000 20.5100 20.1600 20.7900 <0.75 <0.58 0.16 0.32 0.72 0.74 0.80

Table 3.5 continued P Treatments1 TNAL TNPF TNAL HSAL vs. HSAL FA TNAL TNPF HSAL HSSO HSCr SEM Trt2 vs. HS3 vs. HS vs. TNPF HSSO vs. HSCr Omega-3 FA 0.7400 0.7500 0.7800 0.7600 0.7600 <0.03 <0.94 0.47 0.72 0.77 0.72 0.67 Omega-6 FA 18.4800 18.8500 19.7400 19.4000 20.0300 <0.72 <0.56 0.15 0.31 0.72 0.75 0.78 1Treatments: TNAL = thermoneutral (TN) ad libitum; TNPF = TN pair-fed; HSAL = heat stress (HS) ad libitum; HSSO = HS sterculic oil; HSCr = HS chromium 2Treatment 9Total unidentified peaks 3All HS treatments 10Total saturated FA 4Iodine value 11Total monounsaturated FA 5Stearoyl-CoAdesaturase index 12Total polyunsaturated FA 6Polyunsaturated to saturated FA ratio 7Monounsaturated to saturated FA ratio 8Unsaturated to saturated FA ratio

Table 3.6. Effects of period 2 treatment on fatty acid (FA) profile of outer s.c. adipose tissue. a-dValues with differing superscripts denote differences (P ≤ 0.05) between treatments P Treatments1 TNAL TNPF TNAL HSAL HSAL FA TNAL TNPF HSAL HSSO HSCr SEM Trt2 vs. HS3 vs. HS vs. TNPF vs. HSSO vs. HSCr IV4 71.12 70.5600 71.5200 69.9900 71.5300 1.05 0.81 0.93 0.71 0.71 0.31 0.99 16:1/16:0 0.10c 0.10bc 0.08ab 0.08a0 0.08ab 0.01 0.05 0.01 0.02 0.83 0.61 0.81 18:1/18:0 3.93c 3.52bc 3.18b0 2.60a0 3.37b0 0.20 <0.01 <0.01 0.05 0.15 0.05 0.50 SCDI (14,16,18)5 0.57c 0.55bc 0.54b0 0.51a0 0.56bc 0.01 <0.01 <0.01 0.15 0.12 0.01 0.17 PUFA/SFA6 0.61 0.6200 0.6400 0.6100 0.6300 0.03 0.96 0.79 0.89 0.92 0.49 0.89 Omega-3/omega-6 0.04a 0.04c0 0.04bc 0.04ab 0.04a0 0.00 0.01 0.03 0.07 <0.01 0.20 0.06 MUFA/SFA7 1.35c 1.24b0 1.20b0 1.06a0 1.28bc 0.04 <0.01 0.01 0.20 0.10 0.03 0.21 UFA/SFA8 1.96b 1.86b0 1.83b0 1.67a0 1.90b0 0.07 0.05 0.05 0.45 0.31 0.10 0.46 ––––––––––– g/100 g fatty acids ––––––––––– 12:0 0.06 0.0600 0.0500 0.0500 0.0500 0.01 0.62 0.60 0.14 0.65 0.74 0.57 14:0 1.06 1.1400 1.0900 1.1300 1.0300 0.06 0.68 0.82 0.40 0.38 0.63 0.51 15:0 0.06 0.0500 0.0600 0.0600 0.0500 0.01 0.58 0.75 0.46 0.49 0.82 0.20 16:0 20.83 20.7500 20.4100 20.8500 19.9100 0.43 0.51 0.39 0.48 0.90 0.48 0.42 16:1 1.99b 1.97b0 1.64a0 1.57a0 1.63a0 0.12 0.05 0.01 0.02 0.89 0.70 0.97 17:0 0.33a 0.34a0 0.42b0 0.42b0 0.38ab 0.02 0.03 0.01 0.02 0.77 0.89 0.30 100 00 00 00 00 17:1 0.29 0.31 0.31 0.29 0.29 0.02 0.95 0.73 0.75 0.59 0.64 0.57 18:0 10.68a 11.17ab 12.14ab 13.83c0 11.80ab 0.56 0.01 0.01 0.04 0.55 0.04 0.68 18:1 c9 40.64d 38.32bc 37.87b0 35.72a0 39.44cd 0.51 <0.01 <0.01 0.30 <0.01 0.01 0.04 18:2 n6 18.23 18.9500 19.8300 20.1400 19.1600 0.59 0.19 0.04 0.27 0.40 0.72 0.42 18:3 n3 0.73 0.7900 0.8300 0.8200 0.7800 0.03 0.07 0.01 0.57 0.09 0.79 0.16 20:0 0.17 0.2000 0.1900 0.1500 0.1600 0.02 0.55 0.82 0.20 0.40 0.27 0.39 20:1 0.74b 0.75b0 0.75b0 0.62a0 0.75b0 0.03 0.03 0.32 0.21 0.81 0.01 0.99 20:2 n6 0.83 0.8500 0.8900 0.8100 0.9000 0.03 0.22 0.37 0.58 0.78 0.06 0.93 20:3 n6 0.14 0.1200 0.1500 0.1300 0.1400 0.01 0.20 0.83 0.03 0.12 0.27 0.54 22:1 0.33 0.3500 0.3900 0.3500 0.3500 0.02 0.41 0.22 0.49 0.63 0.19 0.19 Unidentified9 3.13 3.9100 3.0300 3.0600 3.2200 0.24 0.09 0.92 0.01 0.03 0.93 0.59 Chain length <16 1.10 1.2400 1.1700 1.2200 1.0900 0.06 0.36 0.41 0.27 0.12 0.60 0.40 Chain length >16 72.95 72.1400 73.7600 73.3000 74.1600 0.59 0.16 0.25 0.02 0.33 0.59 0.63 SFA10 33.04a 33.69a0 34.32a0 36.48b0 33.35a0 0.78 0.03 0.08 0.27 0.56 0.06 0.39 MUFA11 43.92d 41.69bc 40.95b0 38.55a0 42.46c0 0.54 <0.01 <0.01 0.11 0.01 <0.01 0.06 PUFA12 19.91 20.7100 21.7100 21.9000 20.9800 0.64 0.20 0.04 0.28 0.38 0.83 0.42

Table 3.6 continued P Treatments1 TNAL TNPF TNAL HSAL vs. HSAL FA TNAL TNPF HSAL HSSO HSCr SEM Trt2 vs. HS3 vs. HS vs. TNPF HSSO vs. HSCr Omega-3 FA 0.73 0.8000 0.8300 0.8200 0.7800 0.03 0.07 0.01 0.72 0.06 0.79 0.16 Omega-6 FA 19.19 19.9100 20.8700 21.0800 20.2000 0.61 0.21 0.04 0.26 0.41 0.81 0.44 1Treatments: TNAL = thermoneutral (TN) ad libitum; TNPF = TN pair-fed; HSAL = heat stress (HS) ad libitum; HSSO = HS sterculic oil; HSCr = HS chromium 2Treatment 9Total unidentified peaks 3All HS treatments 10Total saturated FA 4Iodine value 11Total monounsaturated FA 5Stearoyl-CoAdesaturase index 12Total polyunsaturated FA 6Polyunsaturated to saturated FA ratio 7Monounsaturated to saturated FA ratio 8Unsaturated to saturated FA ratio

101

Table 3.7. Effects of period 2 treatment on backfat thickness and fatty acid (FA) profile, adipocyte area, and moisture content of abdominal, inner, and outer s.c. a-d adipose tissue depots during the study. Values with differing superscripts denote differences (P ≤ 0.05) between treatments P Parameter Treatments1 SEM Trt2 TNAL TNPF TNAL HSAL HSAL TNAL TNPF HSAL HSSO HSCr vs. HS3 vs. HS vs. TNPF vs. HSSO vs. HSCr Back fat thickness (cm) 3.61 3.73 3.97 3.72 3.63 0.26 0.88 0.60 0.87 0.76 0.50 0.39 Abdominal Adipocyte area (µm2) 3299bc 2431a 3796c 2970ab 2888ab 266 0.01 0.74 0.01 0.02 0.04 0.02 Moisture content (%) 5.90 7.33 7.67 8.76 6.57 0.80 0.15 0.07 0.72 0.23 0.34 0.34 IV4 54.80 55.52 56.46 52.44 56.36 1.57 0.40 0.88 0.80 0.74 0.09 0.97 SCDI (14,16,18)5 0.45b 0.44b 0.41b 0.36a 0.43b 0.01 <0.01 0.01 0.03 0.69 0.02 0.49 ––––––––––––––– g/100 g fatty acids ––––––––––––––– Chain length <16 01.35 1.34 1.50 1.49 1.49 0.08 0.38 0.12 0.10 0.95 0.96 0.92 C16:0 & C16:1 27.37 26.99 27.21 26.78 27.83 0.64 0.82 0.90 0.70 0.67 0.64 0.49 Chain length >16 69.11 69.06 69.27 69.74 69.02 0.65 0.94 0.76 0.71 0.96 0.62 0.79 SFA6 46.65a 46.53a 47.79a 52.26b 47.38a 1.40 0.05 0.14 0.12 0.95 0.04 0.84 MUFA7 38.15c 36.80bc 34.24b 29.97a 35.91bc 1.14 <0.01 <0.01 0.02 0.41 0.02 0.31 8 PUFA 13.03 14.07 15.96 15.78 15.05 0.80 0.07 <0.01 0.11 0.36 0.88 0.43 Inner s.c. Adipocyte area (µm2) 2328 2046 2255 2019 2235 183 0.71 0.46 0.55 0.27 0.39 0.94 Moisture content (%) 9.8 11.03 10.81 11.90 11.39 0.67 0.27 0.05 0.66 0.20 0.26 0.54 IV 68.65 68.00 68.05 64.32 69.30 1.40 0.13 0.38 0.63 0.74 0.07 0.53 SCDI (14,16,18) 0.54c 0.53bc 0.51b 0.46a 0.52bc 0.01 <0.01 <0.01 0.02 0.40 <0.01 0.62 ––––––––––––––– g/100 g fatty acids ––––––––––––––– 102 Chain length <16 1.09 1.19 1.19 1.26 1.25 0.06 0.29 0.04 0.54 0.24 0.40 0.53

C16:0 & C16:1 22.86 22.98 22.68 23.61 23.41 0.61 0.80 0.60 0.72 0.89 0.29 0.40 Chain length >16 73.09 72.19 73.30 72.38 73.21 0.64 0.64 0.86 0.31 0.33 0.32 0.92 SFA 35.34a 35.90a 37.41a 41.69b 36.98a 1.22 0.01 0.02 0.06 0.75 0.02 0.81 MUFA 42.48c 40.86bc 39.26b 35.41a 40.09b 0.82 <0.01 <0.01 0.01 0.17 <0.01 0.48 PUFA 19.21 19.60 20.51 20.16 20.79 0.75 0.58 0.16 0.32 0.72 0.74 0.80 Outer s.c. Adipocyte area (µm2) 2289 2143 2049 2136 1978 91 0.18 0.03 0.46 0.29 0.47 0.63 Moisture content (%) 12.15 14.98 13.05 13.09 13.85 0.68 0.07 0.14 0.04 0.01 0.96 0.41 IV 71.12 70.56 71.52 69.99 71.53 1.05 0.81 0.93 0.71 0.71 0.31 0.99 SCDI (14,16,18) 0.57c 0.55bc 0.54b 0.51a 0.56bc 0.01 <0.01 <0.01 0.15 0.12 0.01 0.17 ––––––––––––––– g/100 g fatty acids ––––––––––––––– Chain length <16 1.10 1.24 1.17 1.22 1.09 0.06 0.36 0.41 0.27 0.12 0.60 0.40 C16:0 & C16:1 22.82 22.72 22.05 22.42 21.53 0.49 0.35 0.16 0.21 0.88 0.59 0.46 Chain length >16 72.95 72.14 73.76 73.30 74.16 0.59 0.16 0.25 0.02 0.33 0.59 0.63 SFA 33.04a 33.69a 34.32a 36.48b 33.35a 0.78 0.03 0.08 0.27 0.56 0.06 0.39 MUFA 43.92d 41.69bc 40.95b 38.55a 42.46c 0.54 <0.01 <0.01 0.11 0.01 <0.01 0.06 PUFA 19.91 20.71 21.71 21.90 20.98 0.64 0.20 0.04 0.28 0.38 0.83 0.42 1Treatments: TNAL = thermoneutral (TN) ad libitum; TNPF = TN pair-fed; 5Stearoyl-CoA desaturase index HSAL = heat stress (HS) ad libitum; HSSO = HS sterculic oil; HSCr = HS 6Total saturated FA chromium 7Total monounsaturated FA 2Treatment 9Total polyunsaturated FA 3All HS treatments 4Iodine value 103

Figure 3.1. Effect of thermoneutral (TN) ad libitum (TNAL), TN pair-fed (TNPF), heat stress (HS) ad libitum (HSAL), HS sterculic oil (HSSO), and HS chromium (HSCr) on AM rectal temperature (TR; A) and PM TR (B). Error bars represent SE for each wk during the study. The dashed line separates period (P) 1 from P2. 104

Figure 3.2. Effect of thermoneutral (TN) ad libitum (TNAL), TN pair-fed (TNPF), heat stress (HS) ad libitum (HSAL), HS sterculic oil (HSSO), and HS chromium (HSCr) on circulating insulin (A) and glucose (B) levels during period 2. Error bars represent SE for each treatment during the study. a-dValues with differing superscripts denote differences (P ≤ 0.05) between treatments. 105

Figure 3.3. Effect of thermoneutral (TN) ad libitum (TNAL), TN pair-fed (TNPF), and all heat stress (HS) treatments combined on adipose tissue moisture content during period 2. Error bars represent SE for each treatment. 106

CHAPTER 4. DIFFERENTIATING BETWEEN THE EFFECTS OF HEAT STRESS AND LIPOPOLYSACCHARIDE ON PORCINE OVARIAN HEAT SHOCK PROTEIN EXPRESSION

Modified from a manuscript formatted for submission to Biology of Reproduction

J.T. Seibert*, M.K. Adur*, R.B. Schultz*, P.Q. Thomas*, K.L. Clark*, Z.E. Kiefer*, A.F.

Keating*, L.H. Baumgard*, and J.W. Ross*

*Department of Animal Science, Iowa State University, Ames, IA, 50011

Authors contributions:

J.T.S. designed experimental assays, purified mRNA and protein samples, optimized and conducted western blots and qRT-PCR reactions, interpreted the data, drafted the manuscript, and prepared the figures. M.K.A and P.Q.T. helped perform and interpret some of the western blot data. R.B.S., Z.E.K., and K.L.C. helped collect tissue samples, purify mRNA samples, and/or provided intellectual contributions. J.W.R., A.F.K., J.T.S. and L.H.B. contributed to the experimental design of the study. All authors edited the manuscript. 107

Results described here within were supported by the National Pork Board, the Iowa Pork

Producers Association, and Agriculture and Food Research Initiative Competitive Grant no. 2011-67003-30007 from the USDA National Institute of Food and Agriculture.

Abstract

Heat stress (HS) negatively affects both human and farm-animal health and undermines efficiency in a variety of economically important agricultural variables, including reproduction. Heat stress impairs the intestinal barrier, allowing for translocation of the resident microflora and endotoxins, such as lipopolysaccharide

(LPS), from the gastrointestinal lumen into systemic circulation. While much is known about the cellular function of heat shock proteins (HSP), the in vivo ovarian HSP response to stressful stimuli, such as HS or LPS, remains ill-defined. The purpose of this study was to compare the effects of HS or LPS on ovarian HSP expression in pigs. We hypothesized that ovarian HSPs are responsive to both HS and LPS. This study consisted of three independent live-phase experiments. In all experiments, altrenogest (15 mg/d) was administered per os for estrus synchronization (14 d) prior to treatment. In the first experiment, gilts were exposed to cyclical HS (31.1 ± 1.4°C; n = 6) or thermoneutral

(TN; 20.3 ± 0.5°C; n = 6) conditions immediately following altrenogest withdrawal for 5 d during follicular development. Similarly, in experiment two, gilts were subjected to repeated (4x/d) saline (CON; n = 3) or LPS (0.1 μg/kg BW; n = 6) i.v. infusion immediately following altrenogest withdrawal for 5 d. In the third experiment, gilts were subjected to TN (20 ± 1°C; n = 7) or cyclical HS (31 to 35°C; n = 7) conditions 2 d post estrus (dpe) until 12 dpe during the luteal phase. Ovarian samples were preserved for analysis post-euthanasia. While there were marginal differences in transcript abundances 108 of the assessed HSP in each experiment, the protein abundance of specific HSP were influenced by stressors during the follicular or luteal phases. Heat stress during the follicular phase (Exp. 1) tended to increase whole-ovarian protein abundance of

HSP90AA1 (70%, P = 0.07), HSF1 (53%; P = 0.05), HSPA1A (101%; P = 0.10),

HSPD1 (95%; P < 0.01), and HSPB1 (101%; P = 0.01) compared to TN controls; while

HS decreased HSP90AB1 (45%, P = 0.01). Exposure to LPS during the follicular phase

(Exp. 2) increased ovarian HSP90AA1 (61%, P = 0.01), HSPA1A (86%; P = 0.01) and tended to increase HSF1 (200%; P = 0.08) and HSPB1 (57%; P = 0.09) compared to

CON gilts, while HSP90AB1 and HSPD1 were not influenced by LPS. Heat stress during the luteal phase (Exp. 3) increased abundance of HSPB1 (175%; P = 0.03) in corpora lutea (CL), but only numerically increased HSF1 (P = 0.17) and HSPD1 (P = 0.17).

Additionally, HS decreased CL HSP90AB1 (53%, P = 0.02), but HSP90AA1 (P = 0.42) and HSPA1A (P = 0.51) were not influenced by HS. Additionally, while neither HS nor

LPS during the follicular phase influenced hypoxia-related markers in the whole ovary,

HS during the luteal phase numerically decreased CL HIF1A transcript (7.24 fold; P =

0.11) and VEGFA protein (38%; P = 0.14) abundances and tended to reduce G6PD transcript abundance (5.8 fold; P = 0.09). Thus, these data support that HS and LPS similarly regulate expression of specific ovarian HSP and that HS can influence hypoxia- related markers in the CL. The similarities between HS and LPS infusion suggest intraovarian signaling during HS may partially result from compromised intestinal integrity.

Keywords: heat stress, pigs, ovary, corpus luteum 109

Introduction

Heat stress (HS) is a major environmental issue, claiming thousands of lives in recent years world-wide [1, 2]. In addition to its impact on human health, HS threatens food security for the growing world population [3, 4], undermining efficiency at every stage of livestock production [5, 6]. Many species, including pigs, limit their feed intake during HS to minimize the thermic effect of feeding, resulting in reduced growth [5, 7-9].

However, despite the reduced nutrient intake, heat-stressed animals actually retain more adipose tissue as compared to thermoneutral counterparts on the same plane of nutrient intake [5, 8, 10, 11]. While this contradicts what is energetically predicted, hyperthermia also results in increased circulating insulin [11], stemming from increased secretion from pancreatic beta cells [12]. Additionally, whole-body insulin sensitivity is increased [13], which appears to be explained primarily by increased glucose utilization by the immune system [14]. Immunoactivation is in response to compromised intestinal integrity and concomitant increase in circulating lipopolysaccharide (LPS) during HS [15].

Collectively, the aforementioned immune activation and endocrine changes contributes to the overall metabolic dysfunction during HS [5].

Hyperthermia, similar to other hyperinsulinemic conditions (e.g., polycystic ovarian syndrome and obesity), can compromise many aspects of female reproduction and reduce fertility [6, 16]. Moreover, just as with HS [17-19], LPS exposure can also alter gonadotropin secretion [20] and impair embryonic survival [21]. In the ovary specifically, HS and LPS both influence the transcript abundance of steroidogenic enzymes [16, 22, 23] and disrupt in vitro gamete development [24, 25]. Interestingly, however, while in vivo LPS exposure has pro-apoptotic effects in the ovary [26, 27], in vivo HS induces ovarian autophagy and promotes anti-apoptotic signaling [28]. Thus, the 110 stressors have similarities and differences with respect to molecular mechanisms underlying how each contributes to female infertility.

Originally discovered in Drosophila for their involvement in the heat shock response (HSR) [29], heat shock proteins (HSP) are stress responsive proteins that maintain intracellular homeostasis [30]. Heat shock factor 1 (HSF1) is a critical transcription factor in the HSR that acts via binding to regulatory regions containing heat shock elements, ultimately promoting the transcription of the affiliated HSP gene [30,

31]. In response to a stress stimulus, HSF1 self-oligomerizes and translocates to the nucleus, enabling DNA-binding [32, 33]. Heat shock protein family A (Hsp70) member

1A (HSPA1A) and HSP family B (small) member 1 (HSPB1) are two stress-induced proteins involved in protein refolding, the latter specifically involved in maintaining the cytoskeleton [34, 35]. In the mitochondria, HSP family D (HSP60) member 1 (HSPD1) is important for protein import and stress-responsive protein refolding [36, 37]. Heat shock protein 90 chaperones have a variety of functions with two major isoforms, HSP 90 alpha family class A member 1 (HSP90AA1) and HSP 90 alpha family class B member 1

(HSP90AB1), reported to have differing responses to thermal stimuli [38]. Interestingly, specific HSP have been implicated in cytokine production via immune signaling pathways [39, 40]. Additionally, HSP have also been implicated in normal reproductive processes, including estrogen receptor alpha regulation [41], luteolysis [42-44], and control of CYP19A1 activity [45]. Collectively, HSPs are involved not only in cellular homeostasis, but also have roles in other biological processes.

Although HSP are involved in cellular pathways responsive to HS and LPS, their involvement in ovarian stress response pathways are ill-defined. Furthermore, it remains 111 uncertain as to whether the direct and/or indirect effects (e.g., HS-induced compromised intestinal integrity) of HS influence the in vivo ovarian HSP response and how that corresponds with altered ovarian function and concomitant impaired fertility. We hypothesized that the ovarian HSP response is similarly regulated during in vivo HS or

LPS challenges in the pig ovary. Thus, our objective herein was to investigate and compare the effects of HS or LPS exposure during specific phases of the estrous cycle on the ovarian HSP response.

Materials and Methods

Animals and Experimental Design

This study consisted of three independent live-phase experiments (Figure 4.1), all of which have been described previously [28, 46, 47]. In all experiments, altrenogest (15 mg/d; Matrix; Merck Animal Health; Summit, NJ) was administered per os for estrus synchronization 14 d prior to treatment. In the first experiment [28], postpubertal gilts

(126.0 ± 21.6 kg) were exposed to thermoneutral (TN; 20.3 ± 0.5°C; n = 6) or cyclical

HS (25.4 to 31.1°C ± 1.4°C; n = 6) conditions for 5 d immediately following altrenogest withdrawal, during follicular development. Similarly, in experiment two [46], gilts (163 ±

3 kg) were subjected to 5 d of repeated (4x/d) saline (CON; n = 3) or LPS (0.1 μg/kg

BW; from E. coli O55:B5, L2880; n = 6) via i.v. injections immediately following altrenogest withdrawal. In the third experiment [47], gilts (167±10 kg) were subjected to

TN (20 ± 1°C; n = 7) or cyclical HS (31 to 35°C; n = 7) conditions 2 d post estrus (dpe) until 12 dpe during the luteal phase. At the end of each experiment, all animals were euthanized using captive bolt technique followed by exsanguination. Whole ovaries and 112 corpora lutea (CL) were removed immediately following exsanguination, flash frozen in liquid nitrogen, and stored at -80°C until utilized for nucleic acid and protein extraction.

Quantitative One-step RT-PCR

Frozen whole-ovarian (Exp. 1 and 2) or CL (Exp. 3) tissues from each animal were homogenized in QIAzol Lysis Reagent using the TissueLyser II and total RNA isolated using the miRNeasy Mini kit (Qiagen, Hilden, Germany) according to the manufacturer’s protocol. Quantitative RT-PCR analysis of transcripts of interest from both tissues was conducted using the QuantiTect SYBR Green RT-PCR Kit (Qiagen,

Hilden, Germany) and measured on an Eppendorf Mastercycler (Eppendorf, Hamburg,

Germany). All primer sequences utilized for quantitative analysis for each target gene are presented in Table 4.1. RNA input (10 ng) was DNase treated according to the manufacturer’s instructions (Ambion, AM1906) and assayed for each sample in triplicate.

Thermal cycling conditions for SYBR Green detection were 50°C for 30 min (reverse transcription), 95°C for 15 min, followed by 40 repetitive cycles of denaturation at 94°C for 15 s, annealing at 57 to 60°C (depending on the primer set) for 30 s, and extension at

72°C for 30 s followed by fluorescent data acquisition. Melting curve analysis was subsequently conducted following the completion of the PCR protocol. Additionally, to determine possible genomic DNA contamination, a pooled sample was assayed in the absence of reverse transcription. Glyceraldehyde-3-phosphate dehydrogenase (GAPDH) mRNA was included as an endogenous normalization control to correct for loading discrepancies. Relative quantification of target gene expression was evaluated with the comparative cycle threshold (Ct) method. The ΔCt value was determined by subtracting the target Ct of each sample from its respective GAPDH Ct value. Calculation of ΔΔCt 113

involved using the single greatest sample ΔCt (the sample with the lowest expression) as an arbitrary constant to subtract from all other sample ΔCt values [48]. Relative differences for each sample were calculated assuming an amplification efficiency of 2 during the geometric region of amplification and applying the equation 2ΔΔCt. Statistical analysis for each assay was performed on the ΔCt values.

Western Blot

Whole-ovarian (Exp. 1 and 2) or CL (Exp. 3) tissues were homogenized in tissue lysis buffer (1% Triton-x-100, 50 mM HEPES, 150 mM NaCl, 10% glycerol, 50 mM

NaF, 2 mM EDTA, 1% SDS), incubated on ice for 30 min, and centrifuged at 10,621 × g at 4°C for 15 min. The resulting supernatant was centrifuged again at 10,621 × g at 4°C for 15 min. The protein concentration of the clarified supernatant was determined using

Pierce® BCA Protein Assay Kit (Thermo Scientific, Rockford, IL) and quantified using a microplate photometer (Hycult Biotech, Uden, Netherlands). Samples were denatured in loading buffer (50 mM Tris-HCl pH 6.8, 2% sodium dodecyl sulfate [SDS], 10% glycerol, 1% 2-mercaptoethanol, 12.5 mM EDTA, 0.02% bromophenol blue) at 95°C for

5 min, placed immediately on ice for 1 min, then stored at -80°C until downstream use.

Protein samples were loaded into a 4-20% gradient Tris glycine gel (Lonza PAGEr®

Gold Precast Gels) with 50 µg per lane. The BioRad Mini PROTEAN Tetra System was used to run the gel at 60 volts for 30 minutes followed by 120 volts for 90 minutes at room temperature (RT) in running buffer (25 mM Tris base, 246 mM glycine, 3 mM

SDS). After size separation, the proteins were transferred to a nitrocellulose membrane for 1.5 h at 100 volts at 4°C in transfer buffer (25 mM Tris base, 246 mM glycine, 1 mM sodium dodecyl sulfate, 20% ethanol). Equal protein loading and transfer efficiency was 114 confirmed by Ponceau S staining of the nitrocellulose membranes and then washed with shaking in TBST three times for 10 minutes at RT. Membranes were blocked for 1 h with shaking at RT in 5% milk in Tris-buffered saline (TBS) with 0.1% Tween-20 (TBST).

Membranes were incubated overnight with primary antibodies specific for HSP90AA1

(Cell Signaling, 8165; 1:1000), HSP90AB1 (Cell Signaling, 7411; 1:1000), HSF1 (Cell

Signaling, 12972; 1:1000), HSPA1A (Novus Biologicals, NB110-96427; 1:5000),

HSPD1 (Novus Biologicals, NBP2-25149; 1:10000), HSPB1 (Novus Biologicals,

NB110-96427; 1:5000), and vascular endothelial growth factor A (VEGFA; Novus

Biologicals, NBP2-25149; 1:1000) at 4°C in 5% milk TBST. Negative controls were also evaluated using a pooled sample, representing a portion of all protein samples utilized in the western blot analysis, in which the membrane was incubated with rabbit IgG (Cell

Signaling, 2729; 1:1000), mouse IgG (Cell Signaling, 5415; 1:1000),or no primary antibody. Membranes were washed with shaking in TBST three times for 10 min at RT and incubated with the appropriate secondary antibody for 1 h at RT and washed three times for 10 min each in TBST at RT. Horseradish peroxidase substrate (Millipore,

Billerica, MA) was added to the membrane for 40 s in the dark. Membrane images were captured using a ChemiImager 5500 (Alpha Innotech, San Leandro, CA) with Alpha Ease

FC software (version 3.03 Alpha Innotech) and densitometry was used to quantify protein band intensities corresponding to the primary antibody. Densitometric analysis was also conducted with Ponceau S staining for each lane on every blot using Image Studio™ Lite

(Li-Core®), confirming equal protein loading and also used for normalization of specific 115 proteins of interest for the respective sample. Technical replications were also performed for each protein of interest.

Statistical Analysis

Western blot and qRT-PCR data were analyzed using PROC TTEST in SAS

University Edition software, version 9.4 (SAS Institute Inc., Cary, NC). Data are reported as means and statistical significance (P ≤ 0.05) and tendency thresholds (0.05 < P ≤ 0.10) were utilized for interpretation.

Results

HS or LPS have Marginal Effects on Ovarian HSP Transcript Abundance

Quantitative RT-PCR was conducted to analyze changes in mRNA transcript abundance of several HSP genes (Figure 4.2). There was a tendency for whole-ovarian

HSPB8 to be increased (1.67-fold; P = 0.09; Figure 4.2C) in response to LPS exposure during the follicular phase and numerical reductions for whole-ovarian HSPB1 (1.50- fold; P = 0.17; Figure 4.2A) and CL HSPA1A (3.50-fold; P = 0.13; Figure 4.2E) in response to HS during the follicular phase or luteal phase, respectively. The transcript abundance of all other assessed HSP genes were not influenced by the stressor in each experiment (P ≥ 0.27; Figure 4.2A, 4.2C, and 4.2E)

The Protein Abundance of Ovarian HSPs are Selectively Influenced by HS and LPS

Despite little influence on transcript abundance of HSP-related genes, HS during the follicular phase tended to increase whole-ovarian HSP90AA1 (70%, P = 0.07; Figure

4.3B) and HSPA1A (101%, P = 0.10; Figure 4.3E) protein abundance. A similar pattern 116 of upregulation was observed in response to LPS during the follicular phase as

HSP90AA1 (61%, P = 0.01; Figure 4.4B) and HSPA1A (86%, P = 0.01; Figure 4.4E) protein abundances were increased in the whole ovary. However, the protein abundances of CL HSP90AA1 (P = 0.42; Figure 4.5B) and HSPA1A (P = 0.51; Figure 4.5E) were not influenced by HS during the luteal phase. Thus, whole-ovarian HSP90AA1 and

HSPA1A are responsive to HS or LPS during the follicular phase, but not influenced by

HS in the CL during the luteal phase.

Interestingly, the protein abundance of HSP90AB1 was decreased in response to

HS in the whole ovary (45%, P = 0.01; Figure 4.3C) and CL (53%, P = 0.02; Figure

4.5C) during the follicular and luteal phases, respectively. Inversely, HS increased whole- ovarian HSPD1 during the follicular phase (95%, P < 0.01; Figure 4.3F) and numerically increased CL HSPD1 during the luteal phase (69%, P = 0.17; Figure 4.5F). Unlike HS,

LPS exposure during the follicular phase did not alter the protein abundance of whole- ovarian HSP90AB1 (P = 0.86; Figure 4.4C) or HSPD1 (P = 0.89; Figure 4.4F).

Collectively, HSP90AB1 and HSPD1 are altered in response to HS, but not influenced by

LPS exposure.

The protein abundance of whole-ovarian HSF1 was increased in response to HS

(53%, P = 0.05; Figure 4.3D) and tended to be increased due to LPS (200%, P = 0.08;

Figure 4.4D) during the follicular phase. During the luteal phase, HS numerically increased HSF1 protein abundance in the CL (47%, P = 0.17; Figure 4.5D). Whole- ovarian HSPB1 protein abundance was increased due to HS (101%; P = 0.01; Figure

4.3G) and tended to increase in response to LPS (57%, P = 0.09; Figure 4.4G) during the follicular phase. Heat stress also increased CL HSPB1 protein abundance during the 117 luteal phase (175%; P = 0.03; Figure 4.5G). Taken together, ovarian HSF1 and HSPB1 are increased in response to HS or LPS exposure.

Whole-ovarian Hypoxia Markers are not Influenced by HS or LPS during the Follicular Phase, but are Reduced in the CL during HS in the Luteal Phase

We assessed the transcript abundance of five genes involved in hypoxia signaling.

The mRNA abundance of whole-ovarian hypoxia-inducible factor 1-alpha (HIF1A), a major coordinator of the cellular hypoxia response, was unaffected by HS (P = 0.56;

Figure 4.2B) or LPS (P = 0.43; Figure 4.2D) exposure during the follicular phase.

However, CL HIF1A transcript abundance was numerically reduced by HS during the luteal phase (7.24-fold; P = 0.11; Figure 4.2F). Similarly, while HS (P = 0.51) or LPS (P

= 0.69) during the follicular phase did not impact whole-ovarian glucose-6-phosphate dehydrogenase (G6PD) transcript abundance, HS during the luteal phase tended to reduce

CL G6PD transcript abundance (5.8-fold; P = 0.09; Figure 4.2F). The transcript abundance of all other assessed HIF1A target genes (VEGFA; lactate dehydrogenase A,

LDHA; and solute carrier family 2 member 1, SLC2A1) were not altered in any of the experiments (P ≥ 0.32; Figure 4.2B, 4.2D, and 4.2F).

The protein abundance of VEGFA was also investigated and, similar to the transcript response, HS (P = 0.20; Figure 4.6B) or LPS (P = 0.63; Figure 4.6D) during the follicular phase did not influence whole-ovarian VEGFA. However, CL VEGFA 118 protein abundance was numerically reduced by HS during the luteal phase (38%; P =

0.14; Figure 4.6F).

Discussion

Heat stress results in female infertility through alterations to cyclicity, pregnancy, ovarian function, as well as direct effects on the developing offspring [6]. Many of these reproductive consequences imposed by the abiotic stressor are also recapitulated through direct exposures to LPS [16]. However, HS also indirectly results in endotoxemia via vascular-mediated mechanisms leading to a hypoxic environment in the gastrointestinal tract, ultimately compromising intestinal barrier function [15, 49-53]. Thus, it remains unclear whether direct and/or indirect effects of HS mediate alterations to ovarian intracellular signaling pathways and little is known regarding the molecular stress response machinery in the ovary during hyperthermia. Interestingly, while marginal effects were observed for transcript levels, HS or LPS had a major impact on the protein abundances of the assessed HSPs in this study. Thus, intraovarian HSP-related pathways during HS may partially be the result of compromised intestinal integrity and subsequently increased circulating LPS.

There were differential effects on specific HSP with regard to the type of stress stimulus during the follicular phase. While HS or LPS increased whole-ovarian

HSP90AA1, HSPA1A, HSPB1, and HSF1 protein abundances, HSP90AB1 and HSPD1 were only responsive to HS and not influenced by LPS. Nonovarian tissues have previously demonstrated increased HSP90AA1 and HSP90AB1 transcript abundance in response to in vivo LPS exposure [54], which may indicate overall differences in tissue responses to LPS. Regarding thermal stimuli, HSP90AA1 is inducible [55] and 119

HSP90AB1 is unresponsive [56], which may be due to the constitutive expression of the latter [57-59]. Although HSP90AA1 was increased, HS unexpectedly decreased the protein abundance of HSP90AB1. Reasons for this discrepancy are unclear, but could be due to differences between in vitro and in vivo experiments, cell types, and/or species- specific responses to HS.

We have previously shown that the protein abundance of whole-ovarian TLR4, a major receptor involved in LPS stimulation of innate immune pathways [60], is increased in response to HS or LPS during the follicular phase [46, 61]. Interestingly, HSP90AA1 is implicated in the LPS response and subsequent cytokine production [39]. Whether

HSP90AA1 responds to LPS independently or in concert with ovarian TLR4, as well as the possible involvement of other HSPs [39, 40], in ovarian immune-related pathways during HS is of further interest.

Additionally, HS decreased HSP90AB1 and increased HSPB1 protein abundance in the whole ovary and CL during the follicular phase and luteal phase, respectively.

Heat-induced numerical increases for HSF1 (P = 0.17) and HSPD1 (P = 0.17) were also observed in the CL. Interestingly, and perhaps insightful to their biological roles,

HSP90AA1 and HSPA1A were only altered by HS in the ovary during the follicular phase and not in the CL. Cell-specific responses to HS regarding ovarian HSP machinery have been demonstrated in pigs [62], thus the possibility that HS-induced alterations to

HSP90AA1 and HSPA1A occurring in different ovarian cell types during the luteal phase is plausible. Interestingly, however, the abundance of certain HSP transcripts and proteins in porcine cumulus cells are not influenced by in vitro or in vivo HS [62], suggesting specific HSPs could become active after granulosa and thecal cell differentiation based 120 on our CL results. If and how the assessed HSPs in this study become active during luteinization and their involvement in CL regression in the pig will need to be addressed in future studies [42-44].

Alterations to the assessed ovarian HSPs, abundance and/or post-translational modifications, could have critical consequences during HS. The cytoskeleton is a well- known responder to HS and increases in HSPB1 and HSF1 may indicate a response to stabilize cellular structures during HS [34, 63-65]. The HS-induced increases to HSPD1 may indicate possible mitochondrial specific responses as HSPD1 is a critical component for protein import and stress responses within mitochondria [36, 37] and HS has been shown to negatively affect mitochondrial function (e.g., altered morphology, oxidative stress, cytochrome c release) [64, 66, 67]. If localized to the oocyte specifically, this could result in energetic problems for the developing gamete, considering the high lipid content and the emphasized use of mitochondrial β oxidation in the porcine oocyte [68].

Additionally, HSP90AA1 and HSP90AB1 are highly abundant in the murine oocyte [69].

Selective chemical inhibition of the latter alters HSP protein abundance, including HSF1, in cancer cells [70] and has been implicated in ovarian autoimmunity in women with premature ovarian failure [71]. Further studies could establish specific roles of ovarian

HSPs and their association with HS-induced infertility.

We investigated the impacts of HS and LPS on ovarian hypoxia-related markers because ovarian blood flow is reduced in heat-stressed chickens and rabbits during HS

[72, 73], likely due to blood flow redistribution to the periphery [53]. Composed of alpha and beta subunits, HIF1 is a major active transcription factor that binds to hypoxia response elements of target genes [74-76]. Low oxygen availability at the cellular level 121 results in activation of genes involved in angiogenesis (e.g., VEGFA), metabolic shift from oxidative phosphorylation to glycolysis (e.g., SLC2A1 a.k.a. GLUT1, LDHA), and the pentose phosphate pathway (e.g., G6PD) [77, 78].

During hypoxia, ubiquitin-mediated degradation of HIF1A is halted and can be accompanied by increases in HIF1A transcript abundance [74, 79, 80]. In this study, HS nor LPS exposure during the follicular phase altered ovarian HIF1A transcript abundance.

We also did not observe transcript differences in the assessed hypoxia-responsive genes

(VEGFA, SLC2A1, LDHA, and G6PD) as well as VEGFA protein abundance. Cell- specific response differences could explain the lack of influence from either stress exposure, which may have been masked when assessing the transcript or protein levels on a whole-ovarian basis. This notion is supported by tissue and cell line differences in response to hypoxia [81-83] as well as our results pertaining to the CL. Regardless, additional investigation evaluating multiple time points and cell specificity is needed to further elucidate the effect of HS on the ovarian hypoxia response in pigs.

Following ovulation, the granulosa and thecal cells differentiate into luteal cells and contribute to CL formation, which is critical for progesterone production and maintenance of pregnancy in the pig. During the preovulatory-to-CL transition, there are changes in hypoxia signaling and proangiogenic factors, accompanied by an eventual increase in vascularization and blood flow to the ruptured follicle as the CL forms [84,

85]. After CL formation, HIF1A mRNA undergoes alterations in abundance as the luteal phase progresses [86-89] and we hypothesized that it would be altered during hyperthermia, considering HS-induced ovarian and systemic blood flow alterations [53,

72, 73]. Unexpectedly, HS resulted in numerical reductions in HIF1A and VEGFA 122 transcript and protein abundances, respectively, and G6PD tended to be reduced in the

CL at the time of peak progesterone production. Importantly, the presence of developing conceptuses at day 12 of the cycle could significantly alter the endocrine signals influencing the CL response to HS [90]; suggesting that HS could affect the CL differently in pregnant gilts 12 days after the estrus onset. How these reductions in hypoxia markers, as well as the HSPs expressed in the CL, relate to the previously observed HS-induced decrease in CL size [47], and how HS affects the function and development of CL in later luteal stages, is of further interest.

Conclusion

In summary, this study demonstrated similar and differential ovarian protein expression of specific HSPs as a result of HS or LPS, suggesting the posit that the intraovarian HSP response during HS stems from HS-induced compromised intestinal integrity should be investigated. Additionally, while we did not detect differences in the whole ovary during the follicular phase, hypoxia-related transcript and protein abundances were numerically influenced by HS in the CL during the luteal phase. It is important to note that the tissue collection was at the end of the respective treatment in each experiment, thus temporal impacts on the endpoints measured herein could have been missed and this is an area for further examination. How the observe alterations to

HSPs and hypoxia markers correspond to altered ovarian function, CL size reductions, and infertility during HS warrants further investigation.

123

Acknowledgements

The authors would like to thank Emily L. Fett for her design with Figure 1 in this manuscript.

Declaration of Interest

This project was supported by the Iowa Pork Producers Association, the National

Pork Board, and the United States Department of Agriculture. Any opinion, findings, conclusions or recommendations expressed in this publication are those of the authors and do not necessarily reflect the view of the Iowa Pork Producers Association, the

National Pork Board, or the United States Department of Agriculture. No conflicts of interest, financial or otherwise are declared by the author (s).

Literature Cited

1. Kosatsky T. The 2003 European heat waves. Euro Surveill 2005; 10:3-4.

2. Russo S, Sillmann J, Fischer EM. Top ten European heatwaves since 1950 and their occurrence in the coming decades. Environ. Res. Lett. 2015; 10.

3. World Bank. World Development Report 2008: Agriculture for Development. In. Washington, D.C.; 2007.

4. United Nations. World population projected to reach 9.8 billion in 2050, and 11.2 billion in 2100. In. New York, NY; 2017: United Nations Department of Economic and Social Affairs.

5. Baumgard LH, Rhoads RP. Effects of Heat Stress on Postabsorptive Metabolism and Energetics. Annual Review of Animal Biosciences 2013; 1:311-337.

6. Ross JW, Hale BJ, Seibert JT, Romoser MR, Adur MK, Keating AF, Baumgard LH. Physiological mechanisms through which heat stress compromises reproduction in pigs. Mol Reprod Dev 2017; 84:934-945.

7. Bianca W. The signifiance of meterology in animal production. Int J Biometeorol 1976; 20:139-156. 124

8. Collin A, van Milgen J, Dubois S, Noblet J. Effect of high temperature and feeding level on energy utilization in piglets. J Anim Sci 2001; 79:1849-1857.

9. Blaxter KL. Energy metabolism in animals and man. Cambridge, England: Cambridge University Press; 1989.

10. Close WH, Mount LE, Start IB. The influence of environmental temperature and plane of nutrition on heat losses from groups of growing pigs. Animal Production 1971; 13:285-294.

11. Pearce SC, Gabler NK, Ross JW, Escobar J, Patience JF, Rhoads RP, Baumgard LH. The effects of heat stress and plane of nutrition on metabolism in growing pigs. J Anim Sci 2013; 91:2108-2118.

12. Sanz Fernandez MV, Johnson JS, Abuajamieh M, Stoakes SK, Seibert JT, Cox L, Kahl S, Elsasser TH, Ross JW, Isom SC, Rhoads RP, Baumgard LH. Effects of heat stress on carbohydrate and lipid metabolism in growing pigs. Physiol Rep 2015; 3.

13. Sanz Fernandez MV, Stoakes SK, Abuajamieh M, Seibert JT, Johnson JS, Horst EA, Rhoads RP, Baumgard LH. Heat stress increases insulin sensitivity in pigs. Physiol Rep 2015; 3.

14. Kvidera SK, Horst EA, Mayorga EJ, Sanz-Fernandez MV, Abuajamieh M, Baumgard LH. Estimating glucose requirements of an activated immune system in growing pigs. J Anim Sci 2017; 95:5020-5029.

15. Pearce SC, Mani V, Weber TE, Rhoads RP, Patience JF, Baumgard LH, Gabler NK. Heat stress and reduced plane of nutrition decreases intestinal integrity and function in pigs. J Anim Sci 2013; 91:5183-5193.

16. Bidne KL, Dickson MJ, Ross JW, Baumgard LH, Keating AF. Disruption of female reproductive function by endotoxins. Reproduction 2018; 155:R169-R181

17. Tompkins EC, Heidenreich CJ, Stob M. Effect of Post-Breeding Thermal Stress on Embryonic Mortality in Swine. J Anim Sci 1967; 26:377-380.

18. Omtvedt IT, Nelson RE, Edwards RL, Stephens DF, Turman EJ. Influence of heat stress during early, mid and late pregnancy of gilts. J Anim Sci 1971; 32:312-317.

19. Madan ML, Johnson HD. Environmental heat effects on bovine luteinizing hormone. J Dairy Sci 1973; 56:1420-1423.

20. Shakil T, Snell A, Whitehead SA. Effects of lipopolysaccharide and cyclosporin on the endocrine control of ovarian function. J Reprod Fertil 1994; 100:57-64. 125

21. Deb K, Chaturvedi MM, Jaiswal YK. A 'minimum dose' of lipopolysaccharide required for implantation failure: assessment of its effect on the maternal reproductive organs and interleukin-1alpha expression in the mouse. Reproduction 2004; 128:87-97.

22. Herath S, Williams EJ, Lilly ST, Gilbert RO, Dobson H, Bryant CE, Sheldon IM. Ovarian follicular cells have innate immune capabilities that modulate their endocrine function. Reproduction 2007; 134:683-693.

23. Nteeba J, Sanz-Fernandez MV, Rhoads RP, Baumgard LH, Ross JW, Keating AF. Heat Stress Alters Ovarian Insulin-Mediated Phosphatidylinositol-3 Kinase and Steroidogenic Signaling in Gilt Ovaries. Biol Reprod 2015; 92:148.

24. Tseng JK, Tang PC, Ju JC. In vitro thermal stress induces apoptosis and reduces development of porcine parthenotes. Theriogenology 2006; 66:1073-1082.

25. Bromfield JJ, Sheldon IM. Lipopolysaccharide initiates inflammation in bovine granulosa cells via the TLR4 pathway and perturbs oocyte meiotic progression in vitro. Endocrinology 2011; 152:5029-5040.

26. Besnard N, Horne EA, Whitehead SA. Prolactin and lipopolysaccharide treatment increased apoptosis and atresia in rat ovarian follicles. Acta Physiol Scand 2001; 172:17-25.

27. Perez GI, Kujjo LL, Bosu WT. Endotoxin-induced apoptosis in ovarian follicles is partially blocked by 2-methylthioATP or 2-chloroATP. Mol Reprod Dev 1996; 44:360-369.

28. Hale BJ, Hager CL, Seibert JT, Selsby JT, Baumgard LH, Keating AF, Ross JW. Heat stress induces autophagy in pig ovaries during follicular development. Biol Reprod 2017; 97:426-437.

29. Tissieres A, Mitchell HK, Tracy UM. Protein synthesis in salivary glands of Drosophila melanogaster: relation to chromosome puffs. J Mol Biol 1974; 84:389-398.

30. Fulda S, Gorman AM, Hori O, Samali A. Cellular stress responses: cell survival and cell death. Int J Cell Biol 2010; 2010:214074.

31. Tetievsky A, Horowitz M. Posttranslational modification in histones underlie heat acclimation-mediated cytoprotective memory. J Appl Physiol 2010; 109:1552- 1561.

126

32. Sarge KD, Murphy SP, Morimoto RI. Activation of heat shock gene transcription by heat shock factor 1 involves oligomerization, acquisition of DNA-binding activity, and nuclear localization and can occur in the absence of stress. Mol Cell Biol 1993; 13:1392-1407.

33. Westwood JT, Wu C. Activation of Drosophila heat shock factor: conformational change associated with a monomer-to-trimer transition. Mol Cell Biol 1993; 13:3481-3486.

34. Bakthisaran R, Tangirala R, Rao Ch M. Small heat shock proteins: Role in cellular functions and pathology. Biochim Biophys Acta 2015; 1854:291-319.

35. Seo JH, Park JH, Lee EJ, Vo TT, Choi H, Kim JY, Jang JK, Wee HJ, Lee HS, Jang SH, Park ZY, Jeong J, et al. ARD1-mediated Hsp70 acetylation balances stress-induced protein refolding and degradation. Nat Commun 2016; 7:12882.

36. Martin J, Horwich AL, Hartl FU. Prevention of protein denaturation under heat stress by the chaperonin Hsp60. Science 1992; 258:995-998.

37. Levy-Rimler G, Viitanen P, Weiss C, Sharkia R, Greenberg A, Niv A, Lustig A, Delarea Y, Azem A. The effect of nucleotides and mitochondrial chaperonin 10 on the structure and chaperone activity of mitochondrial chaperonin 60. Eur J Biochem 2001; 268:3465-3472.

38. Schopf FH, Biebl MM, Buchner J. The HSP90 chaperone machinery. Nat Rev Mol Cell Biol 2017; 18:345-360.

39. Triantafilou K, Triantafilou M, Dedrick RL. A CD14-independent LPS receptor cluster. Nat Immunol 2001; 2:338-345.

40. Kol A, Lichtman AH, Finberg RW, Libby P, Kurt-Jones EA. Cutting edge: heat shock protein (HSP) 60 activates the innate immune response: CD14 is an essential receptor for HSP60 activation of mononuclear cells. J Immunol 2000; 164:13-17.

41. Dhamad AE, Zhou Z, Zhou J, Du Y. Systematic Proteomic Identification of the Heat Shock Proteins (Hsp) that Interact with Estrogen Receptor Alpha (ERalpha) and Biochemical Characterization of the ERalpha-Hsp70 Interaction. PLoS ONE 2016; 11:e0160312.

42. Khanna A, Aten RF, Behrman HR. Physiological and pharmacological inhibitors of luteinizing hormone-dependent steroidogenesis induce heat shock protein-70 in rat luteal cells. Endocrinology 1995; 136:1775-1781.

43. Khanna A, Aten RF, Behrman HR. Heat shock protein-70 induction mediates luteal regression in the rat. Mol Endocrinol 1995; 9:1431-1440. 127

44. Kim AH, Khanna A, Aten RF, Olive DL, Behrman HR. Cytokine induction of heat shock protein in human granulosa-luteal cells. Mol Hum Reprod 1996; 2:549-554.

45. Driancourt MA, Guet P, Reynaud K, Chadli A, Catelli MG. Presence of an aromatase inhibitor, possibly heat shock protein 90, in dominant follicles of cattle. J Reprod Fertil 1999; 115:45-58.

46. Bidne KL, Kvidera SS, Ross JW, Baumgard LH, Keating AF. Impact of repeated lipopolysaccharide administration on ovarian signaling during the follicular phase of the estrous cycle in post-pubertal pigs. J Anim Sci 2018; 96:3622-3634.

47. Bidne KL. Investigating the ovarian response to endotoxemia. Ames, IA: Iowa State University; 2017.

48. Ashworth MD, Ross JW, Hu J, White FJ, Stein DR, Desilva U, Johnson GA, Spencer TE, Geisert RD. Expression of porcine endometrial prostaglandin synthase during the estrous cycle and early pregnancy, and following endocrine disruption of pregnancy. Biol Reprod 2006; 74:1007-1015.

49. Pearce SC, Mani V, Boddicker RL, Johnson JS, Weber TE, Ross JW, Rhoads RP, Baumgard LH, Gabler NK. Heat stress reduces intestinal barrier integrity and favors intestinal glucose transport in growing pigs. PLoS ONE 2013; 8:e70215.

50. Hall DM, Buettner GR, Oberley LW, Xu L, Matthes RD, Gisolfi CV. Mechanisms of circulatory and intestinal barrier dysfunction during whole body hyperthermia. Am J Physiol Heart Circ Physiol 2001; 280:H509-521.

51. Pearce SC, Sanz-Fernandez MV, Hollis JH, Baumgard LH, Gabler NK. Short- term exposure to heat stress attenuates appetite and intestinal integrity in growing pigs. J Anim Sci 2014; 92:5444-5454.

52. Dokladny K, Moseley PL, Ma TY. Physiologically relevant increase in temperature causes an increase in intestinal epithelial tight junction permeability. Am J Physiol Gastrointest Liver Physiol 2006; 290:G204-212.

53. Hall DM, Baumgardner KR, Oberley TD, Gisolfi CV. Splanchnic tissues undergo hypoxic stress during whole body hyperthermia. Am J Physiol 1999; 276:G1195- 1203.

54. Kaucsar T, Bodor C, Godo M, Szalay C, Revesz C, Nemeth Z, Mozes M, Szenasi G, Rosivall L, Soti C, Hamar P. LPS-induced delayed preconditioning is mediated by Hsp90 and involves the heat shock response in mouse kidney. PLoS ONE 2014; 9:e92004.

128

55. Zhang SL, Yu J, Cheng XK, Ding L, Heng FY, Wu NH, Shen YF. Regulation of human hsp90alpha gene expression. FEBS Lett 1999; 444:130-135.

56. Meng X, Jerome V, Devin J, Baulieu EE, Catelli MG. Cloning of chicken hsp90 beta: the only vertebrate hsp90 insensitive to heat shock. Biochem Biophys Res Commun 1993; 190:630-636.

57. Csermely P, Schnaider T, Soti C, Prohaszka Z, Nardai G. The 90-kDa molecular chaperone family: structure, function, and clinical applications. A comprehensive review. Pharmacol Ther 1998; 79:129-168.

58. Sreedhar AS, Kalmar E, Csermely P, Shen YF. Hsp90 isoforms: functions, expression and clinical importance. FEBS Lett 2004; 562:11-15.

59. Schopf FH, Biebl MM, Buchner J. The HSP90 chaperone machinery. Nat Rev Mol Cell Biol 2017; 18:345-360.

60. Hoshino K, Takeuchi O, Kawai T, Sanjo H, Ogawa T, Takeda Y, Takeda K, Akira S. Cutting edge: Toll-like receptor 4 (TLR4)-deficient mice are hyporesponsive to lipopolysaccharide: evidence for TLR4 as the Lps gene product. J Immunol 1999; 162:3749-3752.

61. Dickson MJ, Hager CL, Al-Shaibi A, Thomas PQ, Baumgard LH, Ross JW, Keating AF. Impact of heat stress during the follicular phase on porcine ovarian steroidogenic and phosphatidylinositol-3 signaling. J Anim Sci 2018; 96:2162- 2174.

62. Pennarossa G, Maffei S, Rahman MM, Berruti G, Brevini TA, Gandolfi F. Characterization of the constitutive pig ovary heat shock chaperone machinery and its response to acute thermal stress or to seasonal variations. Biol Reprod 2012; 87:119.

63. Baird NA, Douglas PM, Simic MS, Grant AR, Moresco JJ, Wolff SC, Yates JR, 3rd, Manning G, Dillin A. HSF-1-mediated cytoskeletal integrity determines thermotolerance and life span. Science 2014; 346:360-363.

64. Welch WJ, Suhan JP. Morphological study of the mammalian stress response: characterization of changes in cytoplasmic organelles, cytoskeleton, and nucleoli, and appearance of intranuclear actin filaments in rat fibroblasts after heat-shock treatment. J Cell Biol 1985; 101:1198-1211.

65. Gavrilova LP, Korpacheva, II, Semushina SG, Iashin VA. [Heat shock induces simultaneous rearrangements of all known cytoskeletal filaments in normal interphase fibroblasts]. Tsitologiia 2012; 54:837-846.

129

66. Davidson JF, Schiestl RH. Mitochondrial respiratory electron carriers are involved in oxidative stress during heat stress in Saccharomyces cerevisiae. Mol Cell Biol 2001; 21:8483-8489.

67. Qian L, Song X, Ren H, Gong J, Cheng S. Mitochondrial mechanism of heat stress-induced injury in rat cardiomyocyte. Cell Stress Chaperones 2004; 9:281- 293.

68. Prates EG, Nunes JT, Pereira RM. A role of lipid metabolism during cumulus- oocyte complex maturation: impact of lipid modulators to improve embryo production. Mediators Inflamm 2014; 2014:692067.

69. Zhang P, Ni X, Guo Y, Guo X, Wang Y, Zhou Z, Huo R, Sha J. Proteomic-based identification of maternal proteins in mature mouse oocytes. BMC Genomics 2009; 10:348.

70. Khandelwal A, Kent CN, Balch M, Peng S, Mishra SJ, Deng J, Day VW, Liu W, Subramanian C, Cohen M, Holzbeierlein JM, Matts R, et al. Structure-guided design of an Hsp90beta N-terminal isoform-selective inhibitor. Nat Commun 2018; 9:425.

71. Pires ES, Khole VV. A block in the road to fertility: autoantibodies to heat-shock protein 90-beta in human ovarian autoimmunity. Fertil Steril 2009; 92:1395-1409.

72. Lublin A, Wolfenson D. Lactation and pregnancy effects on blood flow to mammary and reproductive systems in heat-stressed rabbits. Comp Biochem Physiol A Physiol 1996; 115:277-285.

73. Wolfenson D, Frei YF, Snapir N, Berman A. Heat stress effects on capillary blood flow and its redistribution in the laying hen. Pflugers Arch 1981; 390:86- 93.

74. Wang GL, Jiang BH, Rue EA, Semenza GL. Hypoxia-inducible factor 1 is a basic-helix-loop-helix-PAS heterodimer regulated by cellular O2 tension. Proc Natl Acad Sci U S A 1995; 92:5510-5514.

75. Jiang BH, Agani F, Passaniti A, Semenza GL. V-SRC induces expression of hypoxia-inducible factor 1 (HIF-1) and transcription of genes encoding vascular endothelial growth factor and enolase 1: involvement of HIF-1 in tumor progression. Cancer Res 1997; 57:5328-5335.

76. Jiang BH, Rue E, Wang GL, Roe R, Semenza GL. Dimerization, DNA binding, and transactivation properties of hypoxia-inducible factor 1. J Biol Chem 1996; 271:17771-17778.

130

77. Chi JT, Wang Z, Nuyten DS, Rodriguez EH, Schaner ME, Salim A, Wang Y, Kristensen GB, Helland A, Borresen-Dale AL, Giaccia A, Longaker MT, et al. Gene expression programs in response to hypoxia: cell type specificity and prognostic significance in human cancers. PLoS Med 2006; 3:e47.

78. Hakim AM, Moss G, Gollomp SM. The effect of hypoxia on the pentose phosphate pathway in brain. J Neurochem 1976; 26:683-688.

79. Ohh M, Park CW, Ivan M, Hoffman MA, Kim TY, Huang LE, Pavletich N, Chau V, Kaelin WG. Ubiquitination of hypoxia-inducible factor requires direct binding to the beta-domain of the von Hippel-Lindau protein. Nat Cell Biol 2000; 2:423- 427.

80. Jaakkola P, Mole DR, Tian YM, Wilson MI, Gielbert J, Gaskell SJ, von Kriegsheim A, Hebestreit HF, Mukherji M, Schofield CJ, Maxwell PH, Pugh CW, et al. Targeting of HIF-alpha to the von Hippel-Lindau ubiquitylation complex by O2-regulated prolyl hydroxylation. Science 2001; 292:468-472.

81. Ladoux A, Frelin C. Cardiac expressions of HIF-1 alpha and HLF/EPAS, two basic loop helix/PAS domain transcription factors involved in adaptative responses to hypoxic stresses. Biochem Biophys Res Commun 1997; 240:552- 556.

82. Minet E, Ernest I, Michel G, Roland I, Remacle J, Raes M, Michiels C. HIF1A gene transcription is dependent on a core promoter sequence encompassing activating and inhibiting sequences located upstream from the transcription initiation site and cis elements located within the 5'UTR. Biochem Biophys Res Commun 1999; 261:534-540.

83. Wiener CM, Booth G, Semenza GL. In vivo expression of mRNAs encoding hypoxia-inducible factor 1. Biochem Biophys Res Commun 1996; 225:485-488.

84. Robinson RS, Woad KJ, Hammond AJ, Laird M, Hunter MG, Mann GE. Angiogenesis and vascular function in the ovary. Reproduction 2009; 138:869- 881.

85. Ziecik AJ, Przygrodzka E, Jalali BM, Kaczmarek MM. Regulation of the porcine corpus luteum during pregnancy. Reproduction 2018; 156:R57-R67.

86. van den Driesche S, Myers M, Gay E, Thong KJ, Duncan WC. HCG up-regulates hypoxia inducible factor-1 alpha in luteinized granulosa cells: implications for the hormonal regulation of vascular endothelial growth factor A in the human corpus luteum. Mol Hum Reprod 2008; 14:455-464.

131

87. Berisha B, Schams D, Rodler D, Sinowatz F, Pfaffl MW. Expression pattern of HIF1alpha and vasohibins during follicle maturation and corpus luteum function in the bovine ovary. Reprod Domest Anim 2017; 52:130-139.

88. Boonyaprakob U, Gadsby JE, Hedgpeth V, Routh PA, Almond GW. Expression and localization of hypoxia inducible factor-1alpha mRNA in the porcine ovary. Can J Vet Res 2005; 69:215-222.

89. Przygrodzka E, Kaczmarek MM, Kaczynski P, Ziecik AJ. Steroid hormones, prostanoids, and angiogenic systems during rescue of the corpus luteum in pigs. Reproduction 2016; 151:135-147.

90. Bazer FW. Pregnancy recognition signaling mechanisms in ruminants and pigs. J Anim Sci Biotechnol 2013; 4:23. 132

Tables and Figures

Table 4.1. Primer information for quantitative amplification of each target gene Gene Description Primers, 5'-3' GAPDH Glyceraldehyde-3- F: TGGTGAAGGTCGGAGTGAAC phosphate dehydrogenase R: GAAGGGGTCATTGATGGCGA HSP90AA1 Heat shock protein 90 F: CTGGTCAAGAAATGCTTGGAG alpha family class A R: TGGTCCTGGGTCTCACCTGT member 1 HSP90AB1 Heat shock protein 90 F: AACACTGCGGTCAGGGTATC alpha family class B R: ACATTCCCTCTCCACACAGG member 1 HSF1 Heat shock transcription F: AGCTCAGTGACGTCGGAGAT factor 1 R: AGCATGGATTCCAAACTGCT HSPA1A Heat shock protein family F: TCAAGGGCAAGATCAGCGAG A (Hsp70) member 1A R: TCAAACTCGTCCTTCTCGGC HSPB1 Heat shock protein family F: TCGAAAATACACGCTGCCCC B (small) member 1 R: TTCCGGGCTTTTCCGACTTT HSPB8 Heat shock protein family F: GTCTGGCAAACACGAGGAGA B (small) member 8 R: TGGGGAAAGCGAGGCAAATA HIF1A Hypoxia inducible factor 1 F: ATACTCGGGGACAGATTCGC subunit alpha R: TCCTCACACGCAAATAGCTGA VEGFA Vascular endothelial F: TCACCAAGGCCAGCACATAG growth factor A R: GCCCACAGGGATTTTCTTGC G6PD Glucose-6-phosphate F: ATCATCATGGGAGCATCGGG dehydrogenase R: GTGGCTTTGAAGAACGGCTC LDHA Lactate dehydrogenase A F: ACTCTAGTGTGCCTGTATGGA R: CCAGGACGTGTAGCCTTTCA SLC2A1 Solute carrier family 2 F: TGAGCGTCATCTTCATCCCG member 1 R: GCTTCTTCAGCACGCTCTTG 133

Figure 4.1. Experimental scheme for the study. Gilts were subjected to thermoneutral (TN; n = 6) or heat stress (HS; n = 6) conditions during the follicular phase (5 d; top), saline (CON; n = 3) or lipopolysaccharide (LPS; n = 6) infusion via indwelling jugular catheter during the follicular phase (5 d; middle), or TN (n = 7) or HS (n = 7) during the luteal phase (12 d; bottom). Each live phase experiment was conducted independently [27, 45, 46]. In all experiments, altrenogest was administered per os to all gilts to facilitate estrus synchronization prior to thermal or IV treatments during the follicular phase; gilts underwent estrus detection (behavioral estrus) prior to thermal treatment during the luteal phase.

134

Figure 4.2. Effects of heat stress (HS) or lipopolysaccharide (LPS) on ovarian or CL transcript abundance of heat shock protein (HSPs) or hypoxia-responsive genes. Heat stress during the follicular phase (n = 6) had a marginal impact on the whole-ovarian transcript abundance of HSP-related genes (HSF1, HSP90AA1, HSP90AB1, HSPA1A, and HSPB8; P ≥ 0.35; A) and hypoxia-responsive genes (HIF1A, VEGFA, G6PD, LDHA, and SLC2A1; P ≥ 0.51; B) relative to thermoneutral (TN; n = 6) conditions; HSPB1 was numerically reduced (P = 0.17; A). Similarly, LPS during the follicular phase (n = 6) had little influence on whole-ovarian HSP transcript abundance (P ≥ 0.51; C) relative to saline- infused controls (CON; n = 3), but HSPB8 tended to be increased due to LPS (P = 0.09; C). The transcript abundance of whole-ovarian hypoxia-responsive genes was not influenced by LPS (P ≥ 0.43; D). Heat stress during the luteal phase (n = 7) had a similar influence on corpora lutea (CL) transcript abundance of most HSP-related genes (P ≥ 0.26; E) and only numerically increased HSPA1A transcript abundance (P = 0.13; E) relative to TN conditions (n = 7). Additionally, there was a tendency (P = 0.09) and a numerical (P = 0.11) reduction of CL G6PD and HIF1A, respectively, in response to HS during the luteal phase (F); the CL transcript abundance of other hypoxia-responsive genes were not altered by HS during the luteal phase (P ≥ 0.32; F). Relative differences of normalized transcript abundance are represented as fold change between the treatments in each experiment. The # symbol indicates a statistical tendency (0.05 < P ≤ 0.10). 135

Figure 4.3. Heat stress (HS) during the follicular phase alters heat shock protein (HSP) abundance in the ovary. Gilts either underwent five days of cyclical HS (n = 6) or TN (n = 6) conditions during the follicular phase after estrus synchronization. Western blotting of whole ovarian protein lysates for each gilt with antibodies directed toward HSPs (A) demonstrated a tendency for increased protein abundance due to HS for HSP90AA1 (B); HS decreased the protein abundance of HSP90AB1 (C), but increased and tended to increase HSF1 (D) and HSPA1A (E) protein abundance, respectively. Heat stress increased the protein abundance of HSPD1 (F) and HSPB1 (G). Protein band intensities corresponding to the primary antibody were normalized to Ponceau S staining for each respective lane following densitometric analysis. 136

Figure 4.4. Lipopolysaccharide (LPS) during the follicular phase alters heat shock protein (HSP) abundance in the ovary. Gilts were infused with either saline (CON; n = 3) or LPS (n = 6) via indwelling jugular catheter for five days during the follicular phase after estrus synchronization. Western blotting of whole ovarian protein lysates for each gilt with antibodies directed toward HSPs (A) demonstrated increased protein abundance due to LPS for HSP90AA1 (B), but HSP90AB1 (C) was unaltered. Lipopolysaccharide tended to increase HSF1 (D) and increased HSPA1A (E). The protein abundance of HSPD1 (F) and HSPB1 (G) was not influenced and tended to be increased by LPS, respectively. Protein band intensities corresponding to the primary antibody were normalized to Ponceau S staining for each respective lane following densitometric analysis.

137

Figure 4.5. Heat stress (HS) during the luteal phase alters heat shock protein (HSP) abundance in the corpus luteum (CL). Gilts either underwent twelve days of cyclical HS (n = 7) or TN (n = 7) conditions during the luteal phase after estrus synchronization and subsequent estrus detection. Western blotting of CL protein lysates for each gilt with antibodies directed toward HSPs (A) demonstrated that HS did not influence HSP90A1A (B), but reduced HSP90AB1 (C) protein abundance. Heat stress numerically increased HSF1 (D), but did not alter HSPA1A (E). The protein abundance of HSPD1 (F) and HSPB1 (G) was numerically increased and increased due to HS, respectively. Protein band intensities corresponding to the primary antibody were normalized to Ponceau S staining for each respective lane following densitometric analysis. 138

Figure 4.6. Effects of heat stress (HS) or lipopolysaccharide (LPS) on ovarian VEGFA protein abundance. The protein abundance of VEGFA was not influenced by HS (n = 6) compared to thermoneutral (TN, n = 6; A and B) or LPS (n = 6) compared to saline infusion (CON, n = 3; C and D) during the follicular phase. Heat stress (n = 7) during the luteal phase numerically reduced corpora lutea (CL) VEGFA protein abundance compared to TN (n = 7; E and F). Protein band intensities corresponding to the primary antibody were normalized to Ponceau S staining for each respective lane following densitometric analysis. 139

CHAPTER 5. IMPACT OF PRENATAL AND POSTNATAL HEAT STRESS ON THE ADIPOSE, LIVER, AND MUSCLE-SPECIFIC TRANSCRIPTOME AND TARGETED DNA METHYLATION RESPONSES OF MUSCLE IN PIGS

Modified from a manuscript formatted for submission to PLOS ONE

J.T. Seibert1, J.E. Koltes1, R.L. Boddicker2, N.V.L. Serão1, M. Liu3, C. Du3, D. Nettleton3,

J.M. Reecy1, L.H. Baumgard1, and J.W. Ross1*

1Department of Animal Science, Iowa State University, Ames, IA 2Department of Laboratory Medicine and Pathology, Mayo Clinic, Rochester, MN 3Department of Statistics, Iowa State University, Ames, IA

Author contributions:

J.T.S. conducted DAVID and IPA analysis of results from RNA-Seq, qRT-PCR validation, extracted gDNA, transcription factor binding site prediction, western blotting, determined specific genomic loci to assess DNA methylation, interpreted data, prepared figures (except

Figure 1), and drafted the manuscript. R.L.B conducted RNA-Seq library construction and analysis (with J.E.K. and D.N.), created Figure 1. J.T.S and R.L.B purified RNA for qRT-

PCR. J.T.S and J.E.K conducted statistical analysis evaluating bisulfite sequencing data and relating gene expression with bisulfite sequencing. M.L., C.D., and D.N. provided intellectual contributions regarding statistical analysis of RNA-Seq analysis. N.V.L.S. and

J.M.R. provided intellectual contributions key to the success of the experiment. J.T.S., 140

L.H.B., and J.W.R. provided contributions to the experimental design of the study. All authors edited the manuscript.

Results described here within were supported by the National Pork Board, the Iowa Pork

Producers Association, and Agriculture and Food Research Initiative Competitive Grant no.

2011-67003-30007 from the USDA National Institute of Food and Agriculture.

Abstract

Heat stress (HS) experienced prenatally and/or postnatally alters metabolism and body composition in pigs. Experimental objectives were to identify the transcriptomic responses to prenatal and postnatal HS in tissues involved in metabolic regulation using next- generation RNA sequencing (RNA-Seq). We hypothesized the tissue transcriptional profile is associated with specific changes in DNA methylation. To investigate this, 14 first-parity pregnant crossbred gilts were exposed to one of four environmental treatments (either thermoneutral (TN) or HS conditions during the first and/or second half of gestation: TNTN, n = 4; TNHS, n = 3; HSTN, n = 3; HSHS, n = 4). The resulting 48 offspring were exposed to either constant TN (TNTN, n = 6; TNHS, n = 6; HSTN, n = 6; HSHS, n = 6) or HS (TNTN, n = 6; TNHS, n = 6; HSTN, n = 7; HSHS, n = 5) for five weeks, ending at 19 weeks of age.

Following postnatal treatments, RNA-Seq coupled with downstream bioinformatics analysis for biological themes was performed on adipose tissue, liver, and muscle as well as targeted bisulfite sequencing in muscle. In addition to differences in normalized methylation status of

CpGs within 4 kb of targeted differentially expressed genes, gene expression of PYGM and

HSF1 was associated with cumulative methylation of all CpGs assessed at both loci (P <

0.01; q < 0.01), indicating gene expression differences due to gestational and/or postnatal HS 141 may be mediated by an epigenetic mechanism. However, no relationships were detected when assessing the methylation status of each individual CpG with the respective gene’s transcript expression (P < 0.01; q ≥ 0.28). Postnatal abundance of specific proteins (HSF1,

HIF1A, and MEF2A) were influenced by gestational HS interacting with postnatal exposure to HS or sex effects (P ≤ 0.05). In summary, gestational HS altered the transcriptome in major metabolic tissues, which was associated with changes to DNA methylation at associated loci in muscle. Future studies are needed to evaluate the contribution of the identified cellular pathways, proteomic responses, and epigenetic regulation with the programmed phenotypes in gestationally heat-stressed pigs.

Keywords: transcriptome, DNA methylation, pigs

Introduction

Heat stress (HS) is a teratogen that poses health concerns to all mammals [1].

Importantly, while already a threat to human well-being [2] and animal agriculture [3], heat- related ailments may become more frequent if climate patterns continue to change as predicted [4, 5]. Pigs are particularly sensitive to HS because they lack functional sweat glands [6] and have a thick layer of subcutaneous adipose tissue, which can act as effective biological insulation [7]. Inadequate thermoregulatory capabilities are evident through more pronounced HS-induced reductions in feed intake and growth in recent years. This may stem from a possible negative impact of genetic selection for lean tissue accretion on pig thermal sensitivity [8] as synthesis and maintenance of skeletal muscle generates metabolic heat [9].

The postnatal HS response in pigs is characterized by reduced feed intake and weight gain, decreased reproductive performance, and altered body composition [8, 10-12]. A well- 142 documented adaptation to chronic HS is increased fat accumulation (especially compared to pair-fed thermoneutral controls), despite a corresponding reduction in feed intake [13-15].

Although the mechanisms that drive this shift in energetic partitioning are unknown, a corresponding increase in insulin has been shown in pigs [16] and other species [17, 18] that underscores the metabolic dysregulation, which occurs during HS. Interestingly, prenatal HS similarly alters future body composition (more adipose tissue and less skeletal muscle) and insulin homeostasis [19-21]. Additionally, intrauterine exposure to HS in pigs results in impaired thermoregulation [22] and elevated body temperature [23] relative to TN counterparts.

Throughout in utero development, dynamic changes occur in the epigenome via coordinated histone modifications and DNA methylation in spatial and temporal patterns needed for appropriate growth and differentiation [24]. Alterations to the epigenetic landscape during gestational development can have lasting implications on gene expression and eventually postnatal phenotypes [25, 26]. Importantly, heat-induced epigenetic modifications have been demonstrated in a variety of organisms, including plants [27], insects [28], avian species [29], and mammals [30, 31]. In terms of DNA methylation specifically, HS alters the methylation patterns of two paternally imprinted genes, H19 and

IGF2R, in murine embryos [30]. Interestingly, some quantitative trait loci associated with growth and development in pigs are paternally expressed [32]. Thus, changes to certain imprinted regions via aberrant DNA methylation may explain altered body compositions as a result of gestational HS. Similar to nutritional [26] or chemical [33] prenatal exposures, the aforementioned heat-induced postnatal phenotypes suggest underlying epigenetic 143 mechanisms are established prenatally, which are realized phenotypically during postnatal growth and adulthood.

Objectives of the present study were to compare the transcriptomes of tissues involved in metabolic regulation in response to prenatal and postnatal HS. We hypothesized altered gene expression patterns as a result of prenatal thermal exposure would be associated with specific DNA methylation changes. Understanding how prenatal HS alters the postnatal function of specific porcine tissues is essential to better comprehend the mechanisms underlying the phenotypic responses to HS.

Materials and Methods

Animals and Experimental Design

All experiments involving the use of animals were approved by the Institutional

Animal Care and Use Committee at Iowa State University. A split plot study design was employed to test the effect of gestational HS on the postnatal response to chronic exposure to

HS as previously described [19]. Briefly, 14 pregnant gilts were exposed to one of four different environmental conditions during gestation, consisting of thermoneutral (TN; 18 to

22°C) or HS (28 to 34°C) ambient diurnal temperature cycles. Gilts in TNTN (n = 4) and

HSHS (n = 4) treatment groups were exposed to TN or HS conditions, respectively, for the duration of gestation. The remaining thermal treatment groups represent HS conditions only during the first half (HSTN; n = 3) or second half (TNHS; n = 3) of gestation. Thermal treatments began on day six of pregnancy and the mid-pregnancy switch for TNHS and

HSTN groups occurred on day 55 of pregnancy. The resulting 48 offspring were then exposed to constant thermoneutral (TN; 21°C; 35-50% humidity; TNTN, n = 6; TNHS, n =

6; HSTN, n = 6; HSHS, n = 6) or HS (35°C; 24-43% humidity; TNTN, n = 6; TNHS, n = 6; 144

HSTN, n = 7; HSHS, n = 5) environments during postnatal life for 5 weeks beginning at 14 weeks of age. At the end of the postnatal environmental treatment, pigs were euthanized using a captive bolt followed by exsanguination. Adipose tissue, longissimus dorsi muscle, and liver samples were removed immediately following exsanguination and flash frozen in liquid nitrogen. Frozen tissues were stored at -80°C until utilized for nucleic acid and protein extraction.

RNA Extraction, RNA-Seq Library Preparation, Sequencing, and Alignment

Frozen adipose, muscle, and liver tissues from each animal were homogenized in liquid nitrogen and total RNA was isolated using Trizol reagent (Life Technologies,

Carlsbad, CA) according to the manufacturer’s protocol. RNA pellets were resuspended in nuclease free water. Total RNA was quantified by spectrophotometer (ND-100, NanoDrop

Technologies, Rockland, DE), and quality was determined by visualization of 18S and 28S ribosomal bands stained with SYBR® Safe DNA gel stain (Life Technologies, Carlsbad, CA) after electrophoresis on 2% agarose gels.

The TruSeq v2 kit (Illumina, San Diego, CA) was used to construct mRNA-Seq libraries from 2 μg total RNA per sample following manufacturer’s protocol. The 144 libraries (48 offspring × 3 tissues) were indexed with 24 unique indexes using the TruSeq v2 kit. Library quantity was assessed using a Qubit Fluorometer (Life Technologies, Carlsbad,

CA) and diluted to 2 nM for sequencing. Sixteen libraries were pooled per lane, and lane placement was constructed such that each lane consisted of at least one animal per treatment.

Each tissue was represented on each of two flow cells for sequencing. Next generation sequencing was used to generate 100 bp, paired end reads on an Illumina HiSeq 2000

(Illumina, San Diego, CA) at the Iowa State University DNA Facility. Paired-end reads were 145 mapped to the porcine genome (Sscrofa10.2) using Tophat [34]. Aligned reads were imported into SeqMonk, and read counts were quantified for each gene. Genes with an average of at least one uniquely mapped read across samples and a number of non-zero read counts at least as large as the number of combinations of gestational treatment, postnatal treatment, and sex were separately analyzed for differential expression.

Bioinformatic Analysis

Gene names were converted to human identifications using Clone/Gene ID Converter

[35] and then converted to Ensembl IDs using BioMart (http://www.ensembl.org/biomart).

Differentially expressed (q-value ≤ 0.2) gene lists (Supplementary Tables 5.1 to 5.3) for prenatal analyses in each tissue were then related to cellular pathways using the “Disease &

Functions” feature in Ingenuity Pathway Analysis (IPA; Ingenuity Systems, Redwood City,

CA, USA). Only biological functions demonstrating P < 0.05, calculated with Fisher’s exact test, were considered as significant associations between the expression pattern of the transcripts and cellular pathways. Database for Annotation, Visualization, and Integrated

Discovery (DAVID) version 6.8 [36, 37] was also utilized to annotate biological themes occurring in the same gene lists against the human genome.

Quantitative One-step RT-PCR

Quantitative RT-PCR analysis of transcripts of interest from adipose, muscle, and liver was conducted using the QuantiTect SYBR Green RT-PCR Kit (Qiagen, Hilden,

Germany) and measured on an Eppendorf Mastercycler (Eppendorf, Hamburg, Germany).

All primer sequences utilized for quantitative analysis for each target gene are presented in

Table 5.1. Fifty (adipose tissue) or 100 ng (liver and muscle) of starting RNA input was 146 assayed for each sample in triplicate. Thermal cycling conditions for SYBR Green detection were 50°C for 30 min, 95°C for 15 min, followed by 40 repetitive cycles of denaturation at

94°C for 15 s, annealing at 57°C for 30 s, and extension at 72°C for 30 s followed by fluorescent data acquisition. Melting curve analysis was subsequently conducted following the completion of the PCR protocol. Additionally, to confirm the lack of genomic DNA contamination, a pooled sample was assayed in the absence of reverse transcription.

Glyceraldehyde-3-phosphate dehydrogenase (GAPDH) mRNA was assayed as an endogenous normalization control to correct for loading discrepancies. Relative quantification of target gene expression was evaluated with the comparative cycles to threshold (Ct) method. The ΔCt value was determined by subtracting the target Ct of each sample from its respective GAPDH Ct value. Calculation of ΔΔCt involved utilizing the average ΔCt from the TNTN group to subtract from the other three gestational treatment groups. Relative mRNA units for each gestational treatment were calculated assuming that each cycle difference is equivalent to a 2-fold difference, and applying the equation 2ΔΔCt.

Statistical analysis for each assay was performed on the ΔCt values.

Bisulfite Sequencing of Targeted Genomic Regions

Genomic DNA from the longissimus dorsi muscle samples was purified using the

Quick-DNA Universal Kit (Zymo Research, Irvine, CA) and submitted for Targeted

Sequencing for DNA Methylation Analysis provided by Zymo Research (Irvine, CA).

Briefly, gDNA was bisulfite converted and subjected to PCR amplification of specific loci near genes selected following RNA-Seq analysis (Table A.1). The CpG Island predictor track from the UCSC Genome Browser [38] was used to identify genomic regions of interest upstream, within, or downstream of the selected genes. Following amplification, the resulting 147 amplicons were barcoded, purified, and then prepared for massively parallel sequencing using a MiSeq V2 300bp Reagent kit (Illumina, San Diego, CA) and paired-end sequencing protocol according to the manufacturer’s guidelines. Sequence reads were identified using standard Illumina base-calling software and then analyzed using a Zymo Research proprietary analysis pipeline.

Western Blot

Longissimus dorsi samples were homogenized in 10 mM phosphate buffer that contained 2% SDS with protease and phosphatase inhibitors (Halt™, Thermo Scientific,

Rockford, IL) and whole protein lysates were placed on ice for 30 min, followed by two rounds of centrifugation at 10,000 rpm for 15 min at 4°C. The supernatant was collected and protein concentration was determined using Pierce® BCA Protein Assay Kit (Thermo

Scientific, Rockford, IL) and quantified using a microplate photometer (Hycult Biotech,

Uden, Netherlands). Samples were denatured in loading buffer (50 mM Tris-HCl pH 6.8, 2% sodium dodecyl sulfate [SDS], 10% glycerol, 1% 2-mercaptoethanol, 12.5 mM EDTA,

0.02% bromophenol blue) at 95°C for 5 min, placed immediately on ice for 1 min, then stored at -80°C until downstream use. Protein samples were loaded into a 4-20% Tris glycine gel (Lonza PAGEr® Gold Precast Gels) with 50 µg per lane. The BioRad Mini PROTEAN

Tetra System was used to run the gel at 60 volts for 30 min followed by 120 volts for 90 min at room temperature (RT) in running buffer (25 mM Tris base, 246 mM glycine, 3 mM SDS).

After fractionation, the proteins were transferred to a nitrocellulose membrane for 1 h at 110 volts at 4°C in transfer buffer (25 mM Tris base, 246 mM glycine, 1 mM sodium dodecyl sulfate, 20% ethanol). Equal protein loading and transfer efficiency was confirmed by

Ponceau S staining of the nitrocellulose membranes and then washed with gentle agitation in 148

TBST three times for 10 min at RT. Membranes were blocked for 1 hr with gentle agitation at RT in 5% milk in Tris-buffered saline (TBS) with 0.1% Tween-20 (TBST). Membranes were then incubated with primary antibodies specific for HIF1A (Cell Signaling, 3716;

1:1000), HSF1 (Cell Signaling, 12972; 1:1000), HSPA1A (Novus Biologicals, NB110-

96427; 1:1000), MEF2A (Cell Signaling, 9736; 1:1000), and MYOD1 (Cell Signaling,

13812; 1:500) at 4°C in 5% milk TBST. Negative controls were also evaluated using a reference pooled sample, representing a portion of all protein samples utilized in the western blot analysis, in which rabbit IgG (Cell Signaling, 2729; 1:1000), mouse IgG (Cell Signaling,

5415; 1:1000), or no primary antibody input was incubated with the membrane in place of primary antibody. Following primary antibody incubation, the membranes were washed with gentle agitation in TBST three times for 10 min at RT. Secondary antibodies (goat anti- rabbit, Cell Signaling, 7074; 1:1000; horse anti-mouse, Cell Signaling, 7076; 1:1000) were incubated with the membrane for 1 hr at RT and then washed three times for 10 min each in

TBST at room temperature. Horseradish peroxidase substrate (Millipore, Billerica, MA) was added to the membrane for 40 sec in the dark. Membrane images were captured using a

ChemiImager 5500 (Alpha Innotech, San Leandro, CA) and Alpha Ease FC software

(version 3.03 Alpha Innotech). Densitometric analysis was also conducted with Ponceau S staining for each lane on every blot using Image Studio™ Lite (Li-Core®), confirming equal protein loading and also used for normalization of specific proteins of interest for the respective sample.

Statistical Analysis

The read counts for a given gene were modeled using a mixed negative binomial model with a gene-specific dispersion parameter. The read counts were modeled using an 149 offset normalization factor – determined for each sample by the log of the 0.75 quantile of read counts [39] – and the following gene-specific effects: random effects of sire and dam effects; fixed effects for flow cells and for main effects and all possible interactions among gestational treatment, postnatal treatment, and sex. Estimates of the fold change between treatment comparisons of interest were computed by exponentiating estimates of the appropriate effect differences. The significance of each fold change was evaluated from P- values computed using pseudo-likelihood techniques [40, 41] with the Kenward and Roger

[42] method for approximating denominator degrees of freedom, as implemented by SAS

PROC GLIMMIX [43]. For each comparison of interest, the number of genes with true null hypotheses among all genes tested was estimated [44] and used to estimate the number of genes with true null hypotheses among all genes tested, and this estimate was used to convert the P-values to q-values [45]. Genes with q-values no larger than 0.05 were declared differentially expressed to control false discovery rate at approximately 5%. However, for the sake of bioinformatic analysis, we included genes with q-values no larger than 0.20 for a more inclusive approach (Supplementary Tables 5.1 to 5.3).

Western blot, qRT-PCR, and methylation data were analyzed using PROC MIXED in

SAS, version 9.4 (SAS Institute Inc., Cary, NC). The model included gestational environment, postnatal environment, sex, and all two-way and three-way interactions as fixed effects. Dam nested in gestational environment and sire were used as random effects. To meet the objective of determining the impact of DNA methylation on gene expression, we used the following rationale: the methylation status for each CpG position and the cumulative methylation status for each targeted genomic region (i.e., all evaluated CpGs within a 4 kb proximity of the nearest gene) from the bisulfite sequencing data were analyzed using the 150 same model but also with total read counts for each CpG position used as a covariate. This model residual (the normalized methylation value) for each genomic region was then used to predict the gene expression (the pattern or direction of expression) based on log transformed normalized read counts (from the RNA-Seq analysis) of the respective gene after accounting for the fixed (gestational environment, postnatal environment, sex, and their interactions) and random (dam nested in gestational environment and sire) effects. Visual interpretation of the residual and predicted values from this model was used to assess constant variance of all error terms in the model (means of determining if the model was appropriate). Data are reported as LSmeans and statistical significance (P ≤ 0.05) and tendency thresholds (0.05 < P

≤ 0.10) were utilized for interpretation.

Results

Gene Expression

RNA Sequencing of Adipose, Liver, and Muscle

We used a high-throughput approach to evaluate the transcriptomic response in three different tissues from four gestational and two postnatal thermal treatments in pigs. RNA-Seq resulted in averages of approximately 17.4, 20.6, 16.6 million high-quality reads generated from adipose, liver, and LD, respectively, across all treatments. Of these generated reads, averages of 10,764,853 (61%), 15,232,098 (74%), and 11,176,587 (67%) in adipose, liver, and LD, respectively, were mapped to the pig genome (data not shown). At 19 weeks of postnatal age, we identified 20, 42, and 434 DEG in adipose, liver, and LD, respectively, as a result of prenatal HS exposure, with eight transcripts shared between liver and muscular tissue (Figure 5.1 and data not shown). Differentially expressed genes were identified for gestational treatment effects in offspring at 12 weeks of age following the acute postnatal 151 challenge (data not shown). Additionally, the postnatal effect of thermal stress was evaluated for acute (data not shown) and chronic (data not shown) thermal treatments exposure.

Gene Ontology and Bioinformatics

To better understand the biological processes affected by gestational HS in offspring adipose, liver, and muscle tissue, we conducted pathway analysis with DAVID using DEG (q

< 0.20; Supplementary Tables 5.1 to 5.3). In adipose tissue, genes associated with ubiquitin conjugation, ATP binding, cell cycle, intramembranous spaces, extracellular structures, and metal ion binding were identified (Table A.2). Genes associated with oxidative phosphorylation, phosphorylation, cytoskeleton, ATP binding, and blood vessel development were identified in liver (Table A.3). Similar to adipose and liver, genes pertaining to ATP binding were also found in muscle. Other critical processes in muscle appear to be those involved with intracellular membrane-enclosed lumen, kinase activity, metal ion binding, protein transport, SNARE interaction in vesicular transport, and phosphorylation (Table A.4).

Ingenuity Pathway Analysis was utilized to supplement the characterization of biological processes impacted by gestational HS in the three tissues and also to indicate the activity of specific pathways identified using the DEG list (Supplementary Tables 5.1 to 5.3).

The “Diseases & Functions” feature was used to relate transcript expression with biological functions. Positive and negative z-scores indicate increased and decreased functional activity, respectively, when compared to the TNTN group. Using the same gene list as was used for

DAVID, 1,396 and 621 genes were mapped in the IPA database for muscle and liver, respectively, while only 203 genes were mapped for adipose tissue. Pathways pertaining to apoptosis, autophagy, reactive oxygen species, protein metabolism, and cellular proliferation were influenced by gestational HS with many of these pathways showing the most 152 pronounced difference when comparing HSTN to TNTN (Figure A.1A) as compared to the other two gestational treatments (Table A.5). Many immune-related processes were represented by differentially expressed transcripts of the liver, including differentiation of mononuclear leukocyte cell lines, migration of leukocyte cell lines, and adhesion of macrophages (Table A.5). The common transcripts shared for all three immune-related functions identified in HSTN are represented in Figure A.1B.

Quantitative PCR Analysis and Comparison to RNA-Seq

Differences in mRNA abundance between the four gestational treatments were further quantified with the one-step method for quantitative RT-PCR. Transcripts selected to validate the sequencing results were chosen so that further analysis would provide novel data with regard to the impact of gestational HS on future development and would be physiologically relevant based on their expression profile and their association with biological themes as determined through our bioinformatic analysis. Quantitative RT-PCR was conducted to analyze changes in mRNA expression of transmembrane protein 63A (TMEM63A), immunoglobulin superfamily member 5 (IGSF5), integrin subunit beta 4 (ITGB4), phospholipase D family member 4 (PLD4), NME/NM23 nucleoside diphosphate kinase 3

(NME3), nucleotide binding protein 2 (NUBP2), and zinc finger DHHC-type containing 12

(ZDHHC12). Expression patterns for TMEM63A and IGSF5 were similar to those determined by RNA-Seq analysis, but only IGSF5 was significant (Table 5.2). The expression patterns for the other transcripts differed between the two techniques (Table 5.2).

153

Transcriptional Control

Since DNA methylation is implicated in transcriptional control of gene expression, we evaluated the effects of gestational HS on the methylation status of certain CpG loci, individually and as a defined group, and related the degree of methylation to the gene expression (log transformed normalized RNA-Seq read count) of the proximate gene. Muscle tissue was the primary focus of this analysis as previous research has demonstrated gestational HS alters future body composition, such that offspring had increased adipose accretion at the expense of lean tissue [19, 21]. Bisulfite sequencing was conducted for 1,017 different CpG sites at targeted genomic loci within 4kb (upstream, downstream, or within the gene body; see Table A.6 for exact coordinates) of 20 selected DEG following RNA-Seq analysis. We chose genes based on their relevance to thermal biology, magnitude of difference between the gestational treatments relative to the TNTN control, and the presence of relatively dense CpG regions as predicted by the CpG Islands predictor track in the UCSC

Genome Browser. The average read depth per CpG was ≥ 1767× (Table A.7).

A gestational by postnatal interaction was observed for cumulative methylation status of the 41 CpGs overlapping the PYGM genomic locus (P = 0.03; q = 0.10; Supplementary

Table 5.4). The interaction indicated that postnatal HS during weeks 14 to19 of age resulted in altered cumulative methylation status compared to TN counterparts differentially in pigs from dams subjected to HS during the first half or second half of gestation. In general, those receiving postnatal HS had decreased cumulative methylation compared to postnatal TN conditions, ranging from 31 to 94% decrease depending on the gestational treatment (Figure

5.2A). An additional interaction was observed between postnatal treatments and sex for cumulative methylation status of the CpGs overlapping the PYGM locus (P = 0.04; q = 0.12;

Supplementary Table 5.4). Essentially, there was a decrease in cumulative methylation status 154 in postnatal HS compared to TN conditions, but the reduction was more pronounced in females (86%) compared to males (28%; Figure 5.2B). Additionally, within the PYGM genic region, pigs from dams that experienced HS during the first half of gestation (HSTN and

HSHS) had numerically increased cumulative methylation compared to pigs from dams exposed to TN during the first half of gestation regardless of postnatal treatments (TNTN and

TNHS; 44%; P = 0.12; q = 0.31; Supplementary Table 5.5).

A tendency was observed for gestational treatment to affect cumulative methylation status of the 50 CpGs overlapping the HSF1 genomic locus (P = 0.09; q = 0.21;

Supplementary Table 5.4), indicating that cumulative methylation was decreased by 34 and

40% in HSTN and HSHS, respectively, as compared to TNTN. However, irrespective of postnatal treatment, prenatal HS during the first half of gestation (HSTN and HSHS) decreased (29%; P = 0.04; q = 0.18; Supplementary Table 5.5) cumulative methylation as compared to pigs exposed to TN during the first half of gestation (TNTN and TNHS; Figure

5.2C).

A postnatal by sex interaction was observed for cumulative methylation status of the

61 CpGs overlapping the HIF1A genomic locus (P = 0.03; q = 0.10; Supplementary Table

5.4). The interaction demonstrated chronic postnatal HS numerically reduced cumulative methylation status (63%; P = 0.18) compared to TN counterparts in females, but, in males, postnatal HS tended to increase cumulative methylation (275%; P = 0.08) compared to postnatal TN conditions (Figure 5.2D).

In conjunction with the effects observed for cumulative methylation within the aforementioned genes, the degree of cumulative methylation influenced gene expression, such that there were positive and negative relationships for PYGM and HSF1, respectively 155

(both P < 0.01; q < 0.01; Table 5.3). A weak relationship was also observed for HIF1A cumulative methylation status and gene expression (P = 0.12; q = 0.18; Table 5.3). While there were main and interactive effects observed for methylation status on an individual CpG basis (Supplementary Tables 5.6 and 5.7), no significant relationships were detected between the degree of methylation for any single CpG and transcript abundance of associated genes

(Table A.8).

Protein Expression

Western blots using total longissimus dorsi protein were conducted to ascertain if expression of select genes (from RNA-Seq analysis) were maintained at the level of translation (Figure 5.3A). A gestational treatment by sex interaction (P = 0.03; Figure 5.3B) was detected for HIF1A protein abundance whereby HIF1A abundance was increased for females from the HSTN (≥ 62%) gestational treatment relative to other gestation by sex combinations (P ≤ 0.04), but was not different from HSHS females (P = 0.15). Females from gilts exposed to HS during the first half of gestation (HSTN and HSHS) also had higher

HIF1A abundance (≥ 48%; P ≤ 0.02; Figure 5.3D) compared with all other gestational treatment by sex combinations. Offspring exposed to HS during the first half of gestation

(HSTN and HSHS) also had increased HIF1A abundance (56%) compared with those subjected to TN during the first half of gestation (P = 0.04; data not shown), regardless of sex.

Although gestational treatment tended to influence HSF1 abundance (P = 0.09), postnatal HS increased (19%) HSF1 compared with those assigned to the TN environment irrespective of prenatal treatment (P = 0.04; Figure 5.3C). When offspring were grouped based only on thermal environment during the first half of gestation, an interaction was 156 observed between pre- and postnatal treatments (P = 0.05). The protein abundance of HSF1 was increased due to TN prenatal conditions with postnatal HS (44%) and HS prenatal conditions regardless of postnatal environment (80%; Figure 5.3E) compared with pigs from pre- and postnatal TN conditions. Additionally, HSF1 protein abundance tended to be increased (25%; P = 0.10) for offspring exposed to HS during the first half of gestation, regardless of postnatal environment, compared to offspring subjected to TN during the first half of gestation with a HS postnatal environment (Figure 5.3E). Irrespective of postnatal thermal environment, HS during the first half of gestation increased (48%; P = 0.01) HSF1 abundance compared to pigs from dams exposed to TN conditions during the first half of gestation.

An interaction was observed between gestational treatment and sex for MEF2A protein abundance (P = 0.05; Figure 5.3F). The interaction demonstrated females subjected to HS during the first half of gestation (HSTN and HSHS) had increased (80%) MEF2A abundance relative to males with same prenatal treatment (P = 0.02). Additionally, females subjected to HS during the first half of gestation had increased (200%) MEF2A abundance compared to either sex subjected to TN conditions during the first half of gestation (P =

0.01). Regardless of sex, HS during the first half of gestation increased MEF2A abundance compared with pigs from dams that experienced TN during the first half of gestation (124%;

P = 0.05; data not shown). The abundance of MEF2A also tended to be higher (52%; P =

0.10; data not shown) in females compared to males (when considering TN or HS during first half of gestation). No gestational treatment, postnatal treatment, or sex differences were detected for MYOD1 (Figure 5.3G) or HSPA1A (Figure A.2) protein abundance.

Additionally, each protein was analyzed for the effects of gestational treatment, sex, and their 157 interactions using only offspring exposed to TN postnatal conditions (Tables A.9 and A.10) and the effects of postnatal treatment, sex, and their interactions using only offspring subjected to TN conditions throughout gestation (TNTN; Table A.11).

Discussion

Numerous models have demonstrated that maternal environments have lasting effects on offspring during postnatal life [19, 21, 26, 46, 47]. Studies investigating the effects of gestational HS in sheep have shown lower birth weights [48] and future obesity [47], which may partly be attributed to reduced blood flow to the placenta [49]. Similar to intrauterine growth restriction models in swine [50], our previous studies understanding the effects of gestational HS in pigs indicate increased adipose accretion and reduced lean tissue growth occurring at specific stages of development [19-21]. Additionally, pigs exposed to HS during the first half of gestation had increased subcutaneous fat thickness and circulating levels of the antilipolytic and lipogenic hormone insulin [19] representing another mechanism by which HS negatively impacts animal health and productivity [3]. Corresponding with these phenotypic changes in response to gestational HS, we observed transcriptional profile alterations in major metabolic tissues. Interestingly, our bioinformatics analysis identified metabolic and degradative processes as pathways influenced by gestational HS. However, future studies are needed to elucidate the specific molecular mechanisms involved and how they pertain to the postnatal phenotypes observed as a result of HS during intrauterine development.

Abiotic factors, such as HS, have previously been shown to impact the epigenome, corresponding with altered gene expression [28, 29, 31, 51, 52]. In pigs, specifically, the epigenetic landscape differs among distinct breeds for economically important traits [53] and 158

HS can alter genome-wide DNA methylation in muscle [54]. We were particularly interested if gestational and/or postnatal HS could influence the expression of transcripts through an epigenetic mechanism, so we conducted a targeted DNA methylation analysis. Interestingly, after accounting for false discoveries, gene expression was not statistically related to the methylation status for any individual CpG, but the cumulative methylation status of all CpGs within defined genomic regions did account for some differences in transcript abundance of

HSF1, PYGM, and, to a lesser extent, HIF1A, in the longissimus dorsi. While few studies have observed relationships between the methylation status of a single CpG site and expression of the corresponding gene [25, 55, 56], our results are corroborated by strong associations observed between the methylation of multiple CpGs and gene expression in previous studies [25, 57-59]. Additional investigation is needed to more fully elucidate the functionality of individual CpG methylation status as it relates to the regulation of HSF1 and

PYGM gene expression.

Intriguingly, we observed positive and negative relationships between the cumulative methylation status and gene expression of PYGM and HSF1, respectively. While commonly associated with repressive ability [25, 55, 58, 59], the effect of DNA methylation on transcriptional plasticity is unequivocally dependent on genomic context [60-63]. An example of genomic context dependency occurs in many species at the IGF2-H19 locus, where an imprinting control region (ICR) containing multiple CpGs is located between IGF2 and H19 and mediates the transcription of each gene in a parent-of-origin manner, which is dependent on the methylation status of each allele [63, 64]. Genomic studies have also observed differing effects of DNA methylation depending on whether the CpGs are located in a promoter element [25, 55, 58] or within the gene body [60, 62], typically associated with 159 negative and positive transcriptional regulation, respectively. Moreover, the regulatory effect of DNA methylation at a specific genomic element can have differing effects transcriptionally [58], but this may also depend on the simultaneous status of histone modifications [60, 61] and/or overall chromatin accessibility as assessed by DNase I hypersensitivity [60]. Collectively, the regulatory ability of DNA methylation is complex, and this may explain our observation of differential relationships between DNA methylation and HSF1 or PYGM transcript expression. Regardless of how specific epigenetic signatures influence transcriptional abundance, these data underscore the potential of DNA methylation as a mechanism through which in utero HS exposure alters the postnatal metabolic phenotype [19, 21]

Regardless of gestational treatment, transcript abundance was not predictive of HSF1,

MYOD1, MEF2A, and HIF1A protein abundance. Transcript and protein abundance of

HSF1 was increased in TNHS and HSHS relative to TNTN, although an inverse relationship existed in comparison of TNTN to HSTN, which was similar to patterns observed for

MYOD1. Myocyte enhancement factor 2A and HIF1A protein abundance had an inverse relationship with transcriptional abundance. Reasons for this discrepancy may pertain to differences in the transcriptional rate of the gene, alternative splicing, mRNA turnover, the translatability of the mRNA into functional protein, stability of the protein, and/or feedback loops in which protein abundance influences transcription. The samples used in this analysis represent a single time in the development of the phenotypes previously observed due to gestational HS. For example, programmed body composition alterations in other gestational

HS studies were only evident at later stages of growth [21] and not earlier phases [20]. Thus, 160 perhaps transcript and protein expression patterns would demonstrate different relationships depending on when the postnatal samples are collected.

As expected, HSF1 protein abundance in muscle was increased as a result of postnatal

HS. However, this postnatal effect may be limited to the pigs born from dams subjected to

TN conditions during the first half of gestation as HSF1 protein abundance was increased as a result of HS during the first half of gestation regardless of postnatal environment. The increased protein abundance of this HS-responsive transcription factor may be playing a critical role in heat-induced changes to the epigenome and may be critical in thermally programmed phenotypes [29, 31, 51, 65, 66]. Heat shock factor 1 has been directly implicated in altering histone-specific acetylation levels post-translationally [51, 65].

Additionally, heat-induced histone modifications can result in increased binding of HSF1 to heat-shock regulatory elements of heat-responsive genes, such as HSP70 [31, 67] and -90

[31], ultimately promoting transcription [67]. While not assessed in this study, HSF1 regulation of histone modifications clearly plays a critical role in thermal-responsive epigenetic modifications and, therefore, poses an important factor to evaluate in future studies.

Myocyte enhancer factor 2A is one of four proteins belonging to MEF2 family of

MADS-box transcription factors, which can interact with myogenic basic helix-loop-helix transcription factors, such as MyoD1, and regulate muscle development [68, 69]. In addition to their critical role in myogenesis, MEF2 protein activation is mediated via phosphorylation of p38 MAPK to a variety of environmental stressors [70] and is known to regulate the transcription of genes involved with HS, hypoxia, unfolded protein response, and oxidative damage [71]. Interestingly, MEF2A protein abundance was altered as a result of gestational 161

HS. However, this effect was influenced by sex as females exposed to HS during the first half of gestation had a more pronounced increased MEF2A compared to males of the same gestational conditions. Previous studies have demonstrated female-specific induction of

MEF2A gene expression in mouse and human muscle after exercise [72, 73], validating the sex-specific observation in this study. Specifically, the exercise-induced increase in MEF2A transcript abundance in cardiac tissue in female mice is estrogen receptor beta dependent, coinciding with a metabolic preference for non-glycolytic pathways compared to male counterparts [72, 73]. Although glycolytic metabolism is emphasized in skeletal muscle under HS [3], if and how MEF2A plays a role in HS-induced metabolic changes in muscle, and whether it involves calcineurin-mediated MEF2A activation [74, 75] is of further interest.

Unexpectedly, we observed a gestational effect on HIF1A protein levels in muscle such that there was an increase in HIF1A protein abundance in offspring, particularly in females, subjected to HS during the first half of gestation. Hypoxia inducible factor 1 alpha serves as a sensory protein within the cell, responding to hypoxia by translocating to the nucleus, binding to HIF1B, and positively regulating the transcription of hypoxic responsive genes, such as VEGF or GLUT1 [76, 77]. As part of its role during hypoxia, HIF1A coordinates a cellular metabolic shift from oxidative phosphorylation to glycolysis [77]. This increased glycolytic metabolism is what has been previously observed in porcine heat- stressed muscle [3, 78-80], which may serve as a mechanism to spare glucose for the immune system in response to HS-induced leaky gut [3]. Interestingly, heat shock factor gene expression is dependent on HIF1A during hypoxia [81], indicating crosstalk between hypoxia and heat shock response pathways and may be involved in the upregulation of HSF1 protein 162 abundance in this study. Additionally, in a muscle context, HIF1A negatively regulates gene expression of myogenic regulatory factors, inhibiting myoblast proliferation and differentiation and satellite cell differentiation during embryonic and postnatal developmental periods, respectively [82]. If and how gestational HS impacts myogenesis and muscle regeneration during intrauterine and postnatal developmental periods, respectively, and their contribution to the overall reduced lean tissue accretion postnatally is of obvious interest and will need to be addressed in future studies.

The fact HIF1A was increased in pigs exposed to HS during the first half of gestation suggests muscle tissue was programmed for hypoxic conditions, despite suspected maintenance of perfusion to muscle during HS [83]. On the other hand, HIF1A can be regulated in an oxygen-independent manner as overexpression of glycogen synthase kinase 3

β isoform or forkhead box (FOX)O4 has been shown to negatively regulate the protein levels of HIF1A via promoting ubiquitination and degradation [84]. Oxygen-independent stabilization of HIF1A protein has also been demonstrated with calcineurin activity, which dephosphorylates RACK1 preventing dimerization and subsequent recruitment for ubiquitin ligases and proteasomal degradation [85]. Calcineurin phosphatase activity requires the binding of calmodulin, which in turn relies on binding of calcium for its activity [85].

Interestingly, HS has been shown to increase free calcium from the sarcoplasmic reticulum via ryanodine receptor, resulting in eventual mitochondrial fragmentation and muscle dysfunction [86]. Heat stress has also altered the transcript abundance of genes involved in calcium balance in turkey breast muscle, including calsequestrin 1, which may suggest an attempt to resequester excess calcium from the sarcolemma back to the sarcoplasmic reticulum [87]. Moreover, genome-wide DNA methylation differences occur at genes 163 involved in calcium signaling in muscle of pigs exposed to HS [54]. Perhaps the increased functional activity of apoptotisis-related pathways observed with the IPA results in muscle could be partly due to calcium dysregulation [88]. Regardless of the potential mechanism, elucidating the mechanism regulating the protein abundance of HIF1A and its involvement in the gestationally programmed phenotypes observed from this study will require a more comprehensive investigation.

Conclusions

In summary, this study demonstrated differential gene expression in adipose, liver, and muscle, as a result of prenatal HS exposure. Catabolic processes and immune-related pathways were identified from DEG in muscle and liver, respectively, but whether they are linked to phenotypes from gestational HS will need to be addressed in future studies. While this effect may occur through many mechanisms, these data indicate that cumulative CpG methylation could alter cellular function in key metabolic tissues that contribute to differential gene expression. Additionally, we discovered multiple proteins, whose gestational HS-induced perturbation may contribute to altered pig growth and body composition postnatally. Further investigations are needed to continue expanding our understanding of how epigenetic regulation is impacted during gestational HS, influencing cellular pathways and biological processes translating into altered metabolic and body composition differences in pigs.

Literature Cited

1. Graham JM, Edwards MJ. Teratogen Update: Gestational Effects of Maternal Hyperthermia Due to Febrile Illnesses and Resultant Patterns of Defects in Humans. Teratology. 1998;58(5):209-21. 164

2. Mora C, Counsell CWW, Bielecki CR, Louis LV. Twenty-Seven Ways a Heat Wave Can Kill You: Deadly Heat in the Era of Climate Change. Circ Cardiovasc Qual Outcomes. 2017;10(11). doi: 10.1161/CIRCOUTCOMES.117.004233. PubMed PMID: 29122837.

3. Baumgard LH, Rhoads RP. Effects of Heat Stress on Postabsorptive Metabolism and Energetics. Annu Rev Anim Biosci. 2013;1:311-37. doi: 10.1146/annurev-animal-031412- 103644.

4. St-Pierre NR, Cobanov B, Schnitkey G. Economic Losses from Heat Stress by US Livestock Industries. J Dairy Sci. 2003;86(Supplement):E52-E77. doi: 10.3168/jds.S0022- 0302(03)74040-5.

5. IPCC. Climate Change in 2007-Impacts, Adaptation and Vulnerability. Cambridge, UK: Cambridge University Press; 2007.

6. Ingram DL. Stimulation of cutaneous glands in the pig. J Comp Pathol. 1967;77:93-8.

7. Curtis SE. Environmental Management in Animal Agriculture. Ames, IA: 1983.

8. Renaudeau D, Gourdine JL, St-Pierre NR. A meta-analysis of the effects of high ambient temperature on growth performance of growing-finishing pigs. J Anim Sci. 2011;89(7):2220-30. doi: 10.2527/jas.2010-3329.

9. Brown-Brandl TM, Eigenberg RA, Nienaber JA, Kachman SD. Thermoregulatory profile of a newer genetic line of pigs. Livest. Prod. Sci.. 2001;71:253-60.

10. Rhoads RP, Baumgard LH, Suagee JK. Metabolic Priorities During Heat Stress With An Emphasis on Skeletal Muscle. J Anim Sci. 2013. Epub 2013/02/15. doi: jas.2012-6120 [pii] 10.2527/jas.2012-6120. PubMed PMID: 23408824.

11. Tompkins EC, Heidenreich CJ, Stob M. Effect of Post-Breeding Thermal Stress on Embryonic Mortality in Swine. J Anim Sci. 1967;26(2):377-80. doi: 10.2134/jas1967.262377x.

12. Omtvedt IT, Nelson RE, Edwards RL, Stephens DF, Turman EJ. Influence of Heat Stress During Early, Mid and Late Pregnancy of Gilts. J Anim Sci. 1971;32:312-7.

13. Close WH, Mount LE, Start IB. The influence of environmental temperature and plane of nutrition on heat losses from groups of growing pigs. Anim Prod. 1971;13(2):285- 94.

14. Verstegen MWA, Brascamp EW, Hel WVD. Growing and fattening of pigs in relation to temperature of housing and feeding level. Can J Anim Sci. 1978;58(1-13).

165

15. Collin A, van Milgen J, Dubois S, Noblet J. Effect of high temperature and feeding level on energy utilization in piglets. J Anim Sci. 2001;79(7):1849-57. PubMed PMID: 11465372.

16. Pearce SC, Gabler NK, Ross JW, Escobar J, Patience JF, Rhoads RP, et al. The effects of heat stress and plane of nutrition on metabolism in growing pigs. J Anim Sci. 2013;91(5):2108-18. doi: 10.2527/jas.2012-5738.

17. Torlinska T, Banach R, Paluszak J, Gryczka-Dziadecka A. Hyperthermia effect on lipolytic processes in rat blood and adipose tissue. Acta Physiol Pol. 1987;38(4):361-6. PubMed PMID: 3330899.

18. O'Brien MD, Rhoads RP, Sanders SR, Duff GC, Baumgard LH. Metabolic adaptations to heat stress in growing cattle. Domest. Anim. Endocrinol. 2010;38(2):86-94. doi: 10.1016/j.domaniend.2009.08.005. PubMed PMID: 19783118.

19. Boddicker RL, Seibert JT, Johnson JS, Pearce SC, Selsby JT, Gabler NK, et al. Gestational Heat Stress Alters Postnatal Offspring Body Composition Indices and Metabolic Parameters in Pigs. PloS ONE. 2014;9(11). doi: 10.1371/journal.pone.0110859.

20. Johnson JS, Sanz Fernandez MV, Gutierrez NA, Patience JF, Ross JW, Gabler NK, et al. Effects of in utero heat stress on postnatal body composition in pigs: I. Growing phase. Journal of animal science. 2015;93(1):71-81. doi: 10.2527/jas.2014-8354. PubMed PMID: 25568358.

21. Johnson JS, Sanz Fernandez MV, Patience JF, Ross JW, Gabler NK, Lucy MC, et al. Effects of in utero heat stress on postnatal body composition in pigs: II. Finishing phase. J Anim Sci. 2015;93(1):82-92. doi: 10.2527/jas.2014-8355. PubMed PMID: 25568359.

22. Johnson JS, Boddicker RL, Sanz-Fernandez MV, Ross JW, Selsby JT, Lucy MC, et al. Effects of mammalian in utero heat stress on adolescent body temperature. Int J Hyperthermia. 2013;29(7):696-702. doi: 10.3109/02656736.2013.843723.

23. Johnson JS, Sanz Fernandez MV, Seibert JT, Ross JW, Lucy MC, Safranski TJ, et al. In utero heat stress increases postnatal core body temperature in pigs. J Anim Sci. 2015;93(9):4312-22. doi: 10.2527/jas.2015-9112. PubMed PMID: 26440331.

24. Reik W. Stability and flexibility of epigenetic gene regulation in mammalian development. Nature. 2007;447(7143):425-32.

25. Dolinoy DC, Weidman JR, Waterland RA, Jirtle RL. Maternal genistein alters coat color and protects Avy mouse offspring from obesity by modifying the fetal epigenome. Environ Health Perspect. 2006;114(4):567-72. PubMed PMID: 16581547; PubMed Central PMCID: PMC1440782.

166

26. Begum G, Davies A, Stevens A, Oliver M, Jaquiery A, Challis J, et al. Maternal Undernutrition Programs Tissue-Specific Epigenetic Changes in the Glucocorticoid Receptor in Adult Offspring. Endocrinology. 2013;154(12):4560-9. doi: 10.1210/en.2013-1693.

27. Pecinka A, Dinh HQ, Baubec T, Rosa M, Lettner N, Mittelsten Scheid O. Epigenetic regulation of repetitive elements is attenuated by prolonged heat stress in Arabidopsis. Plant Cell. 2010;22(9):3118-29. doi: 10.1105/tpc.110.078493. PubMed PMID: 20876829; PubMed Central PMCID: PMCPMC2965555.

28. Seong KH, Li D, Shimizu H, Nakamura R, Ishii S. Inheritance of stress-induced, ATF- 2-dependent epigenetic change. Cell. 2011;145(7):1049-61. doi: 10.1016/j.cell.2011.05.029.

29. Kisliouk T, Ziv M, Meiri N. Epigenetic control of translation regulation: alterations in histone H3 lysine 9 post-translation modifications are correlated with the expression of the translation initiation factor 2B (Eif2b5) during thermal control establishment. Dev Neurobiol. 2009;70(2):100-13. Epub 2009/12/02. doi: 10.1002/dneu.20763. PubMed PMID: 19950192.

30. Zhu JQ, Liu JH, Liang XW, Xu BZ, Hou Y, Zhao XX, et al. Heat stress causes aberrant DNA methylation of H19 and Igf-2r in mouse blastocysts. Mol Cells. 2008;25(2):211-5. PubMed PMID: 18413996.

31. Tetievsky A, Horowitz M. Posttranslational modification in histones underlie heat acclimation-mediated cytoprotective memory. J Appl Physiol. 2010;109:1552-61.

32. Thomsen H, Lee HK, Rothschild MF, Malek M, Dekkers JC. Characterization of quantitative trait loci for growth and meat quality in a cross between commercial breeds of swine. J Anim Sci. 2004;82(8):2213-28. doi: 10.2527/2004.8282213x. PubMed PMID: 15318717.

33. Li S, Washburn KA, Moore R, Uno T, Teng C, Newbold RR, et al. Developmental exposure to diethylstilbestrol elicits demethylation of estrogen-responsive lactoferrin gene in mouse uterus. Cancer Res. 1997;57(19):4356-9. PubMed PMID: 9331098.

34. Trapnell C, Roberts A, Goff L, Pertea G, Kim D, Kelley DR, et al. Differential gene and transcript expression analysis of RNA-seq experiments with TopHat and Cufflinks. Nat Protoc. 2012;7(3):562-78. doi: 10.1038/nprot.2012.016. PubMed PMID: 22383036; PubMed Central PMCID: PMCPMC3334321.

35. Alibes A, Yankilevich P, Canada A, Diaz-Uriarte R. IDconverter and IDClight: conversion and annotation of gene and protein IDs. BMC Bioinformatics. 2007;8:9. doi: 10.1186/1471-2105-8-9. PubMed PMID: 17214880; PubMed Central PMCID: PMCPMC1779800.

167

36. Huang da W, Sherman BT, Lempicki RA. Systematic and integrative analysis of large gene lists using DAVID bioinformatics resources. Nat Protoc. 2009;4(1):44-57. doi: 10.1038/nprot.2008.211. PubMed PMID: 19131956.

37. Huang da W, Sherman BT, Lempicki RA. Bioinformatics enrichment tools: paths toward the comprehensive functional analysis of large gene lists. Nucleic Acids Res. 2009;37(1):1-13. doi: 10.1093/nar/gkn923. PubMed PMID: 19033363; PubMed Central PMCID: PMCPMC2615629.

38. Kent WJ, Sugnet CW, Furey TS, Roskin KM, Pringle TH, Zahler AM, et al. The human genome browser at UCSC. Genome Res. 2002;12(6):996-1006. doi: 10.1101/gr.229102. Article published online before print in May 2002. PubMed PMID: 12045153; PubMed Central PMCID: PMCPMC186604.

39. Bullard JH, Purdom E, Hansen KD, Dudoit S. Evaluation of statistical methods for normalization and differential expression in mRNA-Seq experiments. BMC Bioinformatics. 2010;11:94. doi: 10.1186/1471-2105-11-94. PubMed PMID: 20167110; PubMed Central PMCID: PMCPMC2838869.

40. Wolfinger R, O’Connell M. Generalized Linear Mixed Models: A Pseudo-likelihood Approach. J Stat Comput Simul. 1993;4:233-43.

41. Breslow NE, Clayton DG. Approximate Inference in Generalized Linear Mixed Models. J Am Stat Assoc. 1993;88:9-25.

42. Kenward MG, Roger JH. Small Sample Inference for Fixed Effects from Restricted Maximum Likelihood. Biometrics. 1997;53(3):983-97.

43. Inc SI. SAS/STAT® 9.2 User’s Guide. Cary, NC: SAS Institute Inc. 2008.

44. Nettleton D, Hwang JTG, Caldo RA, Wise RP. Estimating the number of true null hypotheses from a histogram of p-values. J Agric Biol Environ Stat. 2006;11:337-56.

45. Storey JD. A direct approach to false discovery rates. J R Stat Soc Series B Stat Methodol. 2002;64(479-498).

46. Shankar K, Harrell A, Liu X, Gilchrist JM, Ronis MJ, Badger TM. Maternal obesity at conception programs obesity in the offspring. Am J Physiol Regul Integr Comp Physiol. 2008;294(2):R528-38. doi: 10.1152/ajpregu.00316.2007. PubMed PMID: 18032473.

47. Chen X, Fahy AL, Green AS, Anderson MJ, Rhoads RP, Limesand SW. β2- Adrenergic receptor desensitization in perirenal adipose tissue in fetuses and lambs with plac entalinsufficiency-induced intrauterine growth restriction. J. Physiol. (Lond.). 2010;588. doi: 10.1113/jphysiol.2010.192310.

168

48. Galan HL, Hussey MJ, Barbera A, Ferrazzi E, Chung M, Hobbins JC, et al. Relationship of fetal growth to duration of heat stress in an ovine model of placental insufficiency. Am. J. Obstet. Gynecol. 1999;180(5):1278-82.

49. Dreiling CE, Carman III FS, Brown DE. Maternal Endocrine and Fetal Metabolic Responses to Heat Stress. J Dairy Sci. 1991;74(1):312-27. doi: 10.3168/jds.S0022- 0302(91)78175-7.

50. Rehfeldt C, Kuhn G. Consequences of birth weight for postnatal growth performance and carcass quality in pigs as related to myogenesis. J Anim Sci. 2006;84 Suppl:E113-23. PubMed PMID: 16582082.

51. Thomson S, Hollis A, Hazzalin CA, Mahadevan LC. Distinct stimulus-specific histone modifications at hsp70 chromatin targeted by the transcription factor heat shock factor-1. Mol Cell. 2004;15(4):585-94. doi: 10.1016/j.molcel.2004.08.002. PubMed PMID: 15327774.

52. Zhou H, Brockington M, Jungbluth H, Monk D, Stanier P, Sewry CA, et al. Epigenetic allele silencing unveils recessive RYR1 mutations in core myopathies. Am J Hum Genet. 2006;79(5):859-68. doi: 10.1086/508500. PubMed PMID: 17033962; PubMed Central PMCID: PMCPMC1698560.

53. Li M, Wu H, Luo Z, Xia Y, Guan J, Wang T, et al. An atlas of DNA methylomes in porcine adipose and muscle tissues. Nat Commun. 2012;3:850. doi: 10.1038/ncomms1854. PubMed PMID: 22617290; PubMed Central PMCID: PMCPMC3508711.

54. Hao Y, Cui Y, Gu X. Genome-wide DNA methylation profiles changes associated with constant heat stress in pigs as measured by bisulfite sequencing. Sci Rep. 2016;6:27507. doi: 10.1038/srep27507. PubMed PMID: 27264107; PubMed Central PMCID: PMCPMC4893741.

55. Weaver IC, Cervoni N, Champagne FA, D'Alessio AC, Sharma S, Seckl JR, et al. Epigenetic programming by maternal behavior. Nat Neurosci. 2004;7(8):847-54. doi: 10.1038/nn1276. PubMed PMID: 15220929.

56. Nile CJ, Read RC, Akil M, Duff GW, Wilson AG. Methylation status of a single CpG site in the IL6 promoter is related to IL6 messenger RNA levels and rheumatoid arthritis. Arthritis Rheum. 2008;58(9):2686-93. doi: 10.1002/art.23758. PubMed PMID: 18759290.

57. Zou B, Chim CS, Zeng H, Leung SY, Yang Y, Tu SP, et al. Correlation between the single-site CpG methylation and expression silencing of the XAF1 gene in human gastric and colon cancers. Gastroenterology. 2006;131(6):1835-43. doi: 10.1053/j.gastro.2006.09.050. PubMed PMID: 17087954.

169

58. Mamrut S, Harony H, Sood R, Shahar-Gold H, Gainer H, Shi YJ, et al. DNA methylation of specific CpG sites in the promoter region regulates the transcription of the mouse oxytocin receptor. PloS ONE. 2013;8(2):e56869. doi: 10.1371/journal.pone.0056869. PubMed PMID: 23441222; PubMed Central PMCID: PMCPMC3575498.

59. Coolen MW, Stirzaker C, Song JZ, Statham AL, Kassir Z, Moreno CS, et al. Consolidation of the cancer genome into domains of repressive chromatin by long-range epigenetic silencing (LRES) reduces transcriptional plasticity. Nat Cell Biol. 2010;12(3):235- 46. doi: 10.1038/ncb2023. PubMed PMID: 20173741; PubMed Central PMCID: PMCPMC3058354.

60. Wagner JR, Busche S, Ge B, Kwan T, Pastinen T, Blanchette M. The relationship between DNA methylation, genetic and expression inter-individual variation in untransformed human fibroblasts. Genome Biol. 2014;15(2):R37. doi: 10.1186/gb-2014-15- 2-r37. PubMed PMID: 24555846; PubMed Central PMCID: PMCPMC4053980.

61. Gilsbach R, Preissl S, Gruning BA, Schnick T, Burger L, Benes V, et al. Dynamic DNA methylation orchestrates cardiomyocyte development, maturation and disease. Nat Commun. 2014;5:5288. doi: 10.1038/ncomms6288. PubMed PMID: 25335909; PubMed Central PMCID: PMCPMC4220495.

62. Yang X, Han H, De Carvalho DD, Lay FD, Jones PA, Liang G. Gene body methylation can alter gene expression and is a therapeutic target in cancer. Cancer Cell. 2014;26(4):577-90. doi: 10.1016/j.ccr.2014.07.028. PubMed PMID: 25263941; PubMed Central PMCID: PMCPMC4224113.

63. Engel N, West AG, Felsenfeld G, Bartolomei MS. Antagonism between DNA hypermethylation and enhancer-blocking activity at the H19 DMD is uncovered by CpG mutations. Nat Genet. 2004;36(8):883-8. doi: 10.1038/ng1399. PubMed PMID: 15273688.

64. Peters J. The role of genomic imprinting in biology and disease: an expanding view. Nat Rev Genet. 2014;15(8):517-30. doi: 10.1038/nrg3766. PubMed PMID: 24958438.

65. Fritah S, Col E, Boyault C, Govin J, Sadoul K, Chiocca S, et al. Heat-shock factor 1 controls genome-wide acetylation in heat-shocked cells. Mol Biol Cell. 2009;20(23):4976- 84. doi: 10.1091/mbc.E09-04-0295. PubMed PMID: 19793920; PubMed Central PMCID: PMCPMC2785740.

66. Norouzitallab P, Baruah K, Vandegehuchte M, Van Stappen G, Catania F, Vanden Bussche J, et al. Environmental heat stress induces epigenetic transgenerational inheritance of robustness in parthenogenetic Artemia model. FASEB J. 2014;28(8):3552-63. doi: 10.1096/fj.14-252049. PubMed PMID: 24755740.

67. Zhao YM, Chen X, Sun H, Yuan ZG, Ren GL, Li XX, et al. Effects of histone deacetylase inhibitors on transcriptional regulation of the hsp70 gene in Drosophila. Cell Res. 2006;16(6):566-76. doi: 10.1038/sj.cr.7310074. PubMed PMID: 16775628. 170

68. Molkentin JD, Olson EN. Combinatorial control of muscle development by basic helix-loop-helix and MADS-box transcription factors. Proc. Natl. Acad. Sci. U.S.A. 1996;93(18):9366-73. PubMed PMID: 8790335; PubMed Central PMCID: PMCPMC38433.

69. Naya FJ, Olson E. MEF2: a transcriptional target for signaling pathways controlling skeletal muscle growth and differentiation. Curr Opin Cell Biol. 1999;11(6):683-8. PubMed PMID: 10600704.

70. de Nadal E, Ammerer G, Posas F. Controlling gene expression in response to stress. Nat Rev Genet. 2011;12(12):833-45. doi: 10.1038/nrg3055. PubMed PMID: 22048664.

71. Blais A, Tsikitis M, Acosta-Alvear D, Sharan R, Kluger Y, Dynlacht BD. An initial blueprint for myogenic differentiation. Genes Dev. 2005;19(5):553-69. doi: 10.1101/gad.1281105. PubMed PMID: 15706034; PubMed Central PMCID: PMCPMC551576.

72. Vissing K, McGee SL, Roepstorff C, Schjerling P, Hargreaves M, Kiens B. Effect of sex differences on human MEF2 regulation during endurance exercise. Am. J. Physiol. Endocrinol. Metab. 2008;294(2):E408-15. doi: 10.1152/ajpendo.00403.2007. PubMed PMID: 18042665.

73. Dworatzek E, Mahmoodzadeh S, Schubert C, Westphal C, Leber J, Kusch A, et al. Sex differences in exercise-induced physiological myocardial hypertrophy are modulated by oestrogen receptor beta. Cardiovasc Res. 2014;102(3):418-28. doi: 10.1093/cvr/cvu065. PubMed PMID: 24654233.

74. Wu H, Rothermel B, Kanatous S, Rosenberg P, Naya FJ, Shelton JM, et al. Activation of MEF2 by muscle activity is mediated through a calcineurin-dependent pathway. The EMBO journal. 2001;20(22):6414-23. doi: 10.1093/emboj/20.22.6414. PubMed PMID: 11707412; PubMed Central PMCID: PMCPMC125719.

75. Shalizi A, Gaudilliere B, Yuan Z, Stegmuller J, Shirogane T, Ge Q, et al. A calcium- regulated MEF2 sumoylation switch controls postsynaptic differentiation. Science. 2006;311(5763):1012-7. doi: 10.1126/science.1122513. PubMed PMID: 16484498.

76. Chi JT, Wang Z, Nuyten DS, Rodriguez EH, Schaner ME, Salim A, et al. Gene expression programs in response to hypoxia: cell type specificity and prognostic significance in human cancers. PLoS Med. 2006;3(3):e47. doi: 10.1371/journal.pmed.0030047. PubMed PMID: 16417408; PubMed Central PMCID: PMCPMC1334226.

77. Dengler VL, Galbraith M, Espinosa JM. Transcriptional regulation by hypoxia inducible factors. Crit Rev Biochem Mol Biol. 2014;49(1):1-15. doi: 10.3109/10409238.2013.838205. PubMed PMID: 24099156; PubMed Central PMCID: PMCPMC4342852.

171

78. Cruzen SM, Pearce SC, Baumgard LH, Gabler NK, Huff-Lonergan E, Lonergan SM. Proteomic changes to the sarcoplasmic fraction of predominantly red or white muscle following acute heat stress. J Proteomics. 2015;128:141-53. doi: 10.1016/j.jprot.2015.07.032. PubMed PMID: 26254011.

79. Cruzen SM, Baumgard LH, Gabler NK, Pearce SC, Lonergan SM. Temporal proteomic response to acute heat stress in the porcine muscle sarcoplasm. J Anim Sci. 2017;95(9):3961-71. doi: 10.2527/jas2017.1375. PubMed PMID: 28992025.

80. Sanz Fernandez MV, Johnson JS, Abuajamieh M, Stoakes SK, Seibert JT, Cox L, et al. Effects of heat stress on carbohydrate and lipid metabolism in growing pigs. Physiol Rep. 2015;3(2). doi: 10.14814/phy2.12315. PubMed PMID: 25716927; PubMed Central PMCID: PMC4393217.

81. Baird NA, Turnbull DW, Johnson EA. Induction of the heat shock pathway during hypoxia requires regulation of heat shock factor by hypoxia-inducible factor-1. J Biol Chem. 2006;281(50):38675-81. doi: 10.1074/jbc.M608013200. PubMed PMID: 17040902.

82. Beaudry M, Hidalgo M, Launay T, Bello V, Darribere T. Regulation of myogenesis by environmental hypoxia. J Cell Sci. 2016;129(15):2887-96. doi: 10.1242/jcs.188904. PubMed PMID: 27505427.

83. Nielsen B, Savard G, Richter EA, Hargreaves M, Saltin B. Muscle blood flow and muscle metabolism during exercise and heat stress. J Appl Physiol. 1990;69(3):1040-6. doi: 10.1152/jappl.1990.69.3.1040. PubMed PMID: 2246151.

84. Koh MY, Spivak-Kroizman TR, Powis G. HIF-1 regulation: not so easy come, easy go. Trends Biochem. Sci. 33(11):526-34. doi: 10.1016/j.tibs.2008.08.002.

85. Liu YV, Hubbi ME, Pan F, McDonald KR, Mansharamani M, Cole RN, et al. Calcineurin promotes hypoxia-inducible factor 1alpha expression by dephosphorylating RACK1 and blocking RACK1 dimerization. J. Biol. Chem. 2007;282(51):37064-73. doi: 10.1074/jbc.M705015200. PubMed PMID: 17965024; PubMed Central PMCID: PMCPMC3754800.

86. Momma K, Homma T, Isaka R, Sudevan S, Higashitani A. Heat-Induced Calcium Leakage Causes Mitochondrial Damage in Caenorhabditis elegans Body-Wall Muscles. Genetics. 2017;206(4):1985-94. doi: 10.1534/genetics.117.202747. PubMed PMID: 28576866; PubMed Central PMCID: PMCPMC5560802.

87. Sporer KR, Zhou HR, Linz JE, Booren AM, Strasburg GM. Differential expression of calcium-regulating genes in heat-stressed turkey breast muscle is associated with meat quality. Poult. Sci. 2012;91(6):1418-24. doi: 10.3382/ps.2011-02039. PubMed PMID: 22582302.

172

88. Guo J, Lao Y, Chang DC. Calcium and Apoptosis. In: Lajtha A, Mikoshiba K, editors. Handbook of Neurochemistry and Molecular Neurobiology: Neural Signaling Mechanisms. Boston, MA: Springer US; 2009. p. 597-622.

88. Boddicker, R.L. Tissue specific transcriptome response to prenatal and postnatal heat stress in pigs. Heat Stress Symposium. Kildee Hall, April 2013. Presentation. 173

Tables and Figures

Table 5.1. Primer information for quantitative amplification of each target gene Amplicon Gene Description Primers, 5'-3' size, bp GAPDH Glyceraldehyde-3- F: TGGTGAAGGTCGGAGTGAAC 104 phosphate dehydrogenase R: GAAGGGGTCATTGATGGCGA IGSF5 Immunoglobulin F: ATCTTACGGTGGTTCAGCCT 175 superfamily, member 5 R: CTAGACTCTTCTTTCTCTCT ITGB4 Integrin, beta 4 F: CCGACACAACATCATCCCCA 161 R: TGGAGCGGATACGGTCAAAG NME3 Non-metastatic cells 3, F: TCTAGACGTTGTTCGCGCTT 112 protein expressed in R: AATCACGTTCTTGCCGACCT NUBP2 Nucleotide binding protein F: CGGCACATCATCCTTGTCCT 109 2 (MinD homolog, E. coli) R: CGAGGAGCCCCACCTTCTT PLD4 Phospholipase D family, F: GTCTCCGTGGACGTGAAAGT 154 member 4 R: CCGAGGTGCTGGTGAAGTAG TMEM63A Transmembrane protein F: ACCCTCCATCATCCTGTCCA 185 63A R: GGTTCTCCCCTGACTTGGTC ZDHHC12 Zinc finger, DHHC-type F: GGATCACGCTGGTGCTGT 178 containing 12 R: GCCTCCTCCTGGGGCTG

174

Table 5.2. Comparison of qRT-PCR and RNA-Seq results for selected genes qRT-PCR, Fold Change1 RNA-Seq, Fold Change2 Gene3 TNHS4 HSTN5 HSHS6 P TNHS HSTN HSHS q Adipose TMEM63A 1.40 1.42 1.79 0.60 1.46 1.61 1.97 0.05 Liver IGSF5 6.17 3.80 2.84 0.05 4.77 3.09 3.29 0.01 ITGB4 0.63 0.77 0.53 0.02 1.42 1.24 2.44 0.04 PLD4 0.50 0.69 0.44 0.37 1.14 1.66 3.27 0.01 Muscle NME3 1.23 1.38 1.42 0.26 2.01 0.88 3.82 0.05 NUBP2 1.10 1.34 1.16 0.35 2.36 0.92 5.08 0.03 ZDHHC12 1.31 1.18 1.20 0.61 2.60 1.25 4.25 <0.01 1qRT-PCR fold changes are relative to the treatment with thermoneutral conditions during the entire gestation (TNTN) 2RNA-Seq fold changes are relative to the TNTN group 3Transmembrane protein 63A (TMEM63A); Immunoglobulin superfamily, member 5 (IGSF5); Integrin, beta 4 (ITGB4); Phospholipase D family, member 4 (PLD4); Non- metastatic cells 3, protein expressed in (NME3); Nucleotide binding protein 2 (MinD homolog, E. coli) (NUBP2); Zinc finger, DHHC-type containing 12 (ZDHHC12) 4Thermoneutral during first half of gestation and heat stress during second half of gestation 5Heat stress during first half of gestation and thermoneutral during second half of gestation 6Heat stress during the entire gestation

175

Table 5.3. Slope estimates for the relationship of transcript expression of select genes from the RNA-Seq analysis and normalized cumulative methylation status of CpGs at targeted genomic loci within 4kb of each respective gene. Gene1 Slope2 P q PYGM 0.000050 <0.01 <0.01 HSF1 -0.000013 <0.01 0.01 CTNNA1 0.000090 0.09 0.17 HIF1A -0.000132 0.12 0.17 MANEA -0.000206 0.15 0.19 GPT 0.000002 0.29 0.33 PMAIP1 -0.000060 0.41 0.40 GPR126 0.000004 0.43 0.40 PGGT1B 0.000029 0.52 0.43 CHST1 -0.000013 0.57 0.43 ZNHIT2 -0.000023 0.59 0.43 FAR1 -0.000009 0.70 0.47 MYOD1 0.000002 0.77 0.47 CARD10 -0.000009 0.80 0.47 PHF20L1 0.000003 0.83 0.47 RAC1 0.000015 0.97 0.52 1Caspase recruitment domain family, member 10 (CARD10); Carbohydrate (keratan sulfate Gal-6) sulfotransferase 1 (CHST1); Catenin (cadherin-associated protein), alpha 1, 102kDa (CTNNA1); Fatty acyl CoA reductase 1 (FAR1); FOS-like antigen 1 (FOSL1); G protein-coupled receptor 126 (GPR126); Glutamic-pyruvate transaminase (alanine aminotransferase) (GPT); Hypoxia-inducible factor 1, alpha subunit (basic helix-loop-helix transcription factor) (HIF1A); Heat shock transcription factor 1 (HSF1); Mannosidase, endo- alpha (MANEA); Myogenic differentiation 1 (MYOD1); Protein geranylgeranyltransferase type I, beta subunit (PGGT1B); PHD finger protein 20-like 1 (PHF20L1); p53- induced death domain protein (PIDD); Phorbol-12-myristate- 13-acetate-induced protein 1 (PMAIP1); Phosphorylase, glycogen; muscle (McArdle syndrome, glycogen storage disease type V) (PYGM); Ras-related C3 botulinum toxin substrate 1 (rho family, small GTP binding protein Rac1) (RAC1); Zinc finger, HIT type 2 (ZNHIT2) 2The slope was estimated using a mixed linear model where the normalized methylation value for each genomic region was used to predict the gene expression of the respective gene after accounting for the fixed (gestational environment, postnatal environment, sex, and their interactions) and random (dam nested in gestational environment and sire) effects. 176

Figure 5.1. Gestational thermal treatment alters transcriptomic profile in adipose, liver, and longissimus dorsi muscle. Fourteen first parity crossbred gilts were exposed to one of four environmental treatments consisting of heat stress (HS) or thermoneutral (TN) conditions at different time periods during gestation. Dams assigned to TNTN (n = 4) and HSHS (n = 4) were exposed to TN or HS conditions, respectively, throughout the entire length of gestation. Dams assigned to TNHS (n = 3) and HSTN (n = 3) were heat-stressed for the first or second half of gestation, respectively. The resulting 48 offspring were also subjected to postnatal thermal treatments, TN (TNTN, n = 6; TNHS, n = 6; HSTN, n = 6; HSHS, n = 6) or HS (TNTN, n = 6; TNHS, n = 6; HSTN, n = 7; HSHS, n = 5), constantly for five weeks beginning at 12 weeks of age. At 19 weeks of age, the offspring were euthanized and tissues were collected and preserved for RNA-Seq analysis. The Venn diagram demonstrates the number of differentially expressed transcripts identified (FDR < 5%) for adipose, liver, and muscle as a result of prenatal thermal treatments in offspring subjected to chronic postnatal heat stress (HS) or thermoneutral (TN) conditions. There were eight common transcripts shared between liver and muscle. Refer to Supplemental Tables S5-7 for annotated transcripts and fold changes in TNHS, HSTN, and HSHS relative to TNTN offspring. Figure adapted from Boddicker, 2013 [89]. 177

Figure 5.2. Effects of pre- and postnatal thermal treatment and sex on cumulative methylation levels of CpGs overlapping PYGM, HSF1, and HIF1A genomic loci in longissimus dorsi muscle. Fourteen first parity crossbred gilts were exposed to one of four environmental treatments consisting of heat stress (HS) or thermoneutral (TN) conditions at different time periods during gestation. Dams assigned to TNTN (n = 4) and HSHS (n = 4) were exposed to TN or HS conditions, respectively, throughout the entire length of gestation. Dams assigned to TNHS (n = 3) and HSTN (n = 3) were heat-stressed for the first or second half of gestation, respectively. The resulting 48 offspring were also subjected to postnatal thermal treatments, TN (TNTN, n = 6; TNHS, n = 6; HSTN, n = 6; HSHS, n = 6) or HS (TNTN, n = 6; TNHS, n = 6; HSTN, n = 7; HSHS, n = 5), constantly for five weeks beginning at 12 weeks of age. At 19 weeks of age, the offspring were euthanized and muscle was collected and preserved for bisulfite sequencing analysis. The total read counts for a particular CpG were accounted for to generate normalized methylation levels for each CpG evaluated; cumulative methylation levels, accounting for the normalized methylation levels of all CpGs within a 4 kb proximity of the nearest gene together, was also evaluated. An interaction was observed between gestational and postnatal thermal treatment for cumulative methylation of 41 CpGs at the PYGM locus (P = 0.03; A). A postnatal thermal treatment by sex interaction was also observed for CpGs overlapping PYGM (P = 0.04; B) such that cumulative methylation status was decreased in postnatal HS compared to TN conditions, but the difference was augmented in females (86%) compared to males (28%). Prenatal HS during the first half of gestation reduced cumulative methylation of 50 CpGs overlapping the HSF1 gene compared to offspring exposed to TN conditions during the first half of gestation (29%; C). A postnatal thermal treatment by sex interaction was also observed for 61 CpGs at the HIF1A locus (P = 0.03; D) such that cumulative methylation was numerically decreased in females subjected to HS compared to TN females (63%; P = 0.18), but tended to be increased in heat-stressed males relative to TN males (275%; P = 0.08). 178

Figure 5.3. Effects of pre- and postnatal thermal treatment and sex on abundance of select proteins, based on significant differentially expressed transcripts from RNA-Seq, in longissimus dorsi muscle. Fourteen first parity crossbred gilts were exposed to one of four environmental treatments consisting of heat stress (HS) or thermoneutral (TN) conditions at different time periods during gestation. Dams assigned to TNTN (n = 4) and HSHS (n = 4) were exposed to TN or HS conditions, respectively, throughout the entire length of gestation. Dams assigned to TNHS (n = 3) and HSTN (n = 3) were heat-stressed for the first or second half of gestation, respectively. The resulting 48 offspring were also subjected to postnatal thermal treatments, TN (TNTN, n = 6; TNHS, n = 6; HSTN, n = 6; HSHS, n = 6) or HS (TNTN, n = 6; TNHS, n = 6; HSTN, n = 7; HSHS, n = 5), constantly for five weeks beginning at 12 weeks of age. Representative western blots of whole muscle protein lysates with antibodies directed toward HIF1A, HSF1, MEF2A, and MYOD1 (A) and quantifications (G). An interaction was observed between gestational thermal treatment and sex for HIF1A (P = 0.02; B), which was also maintained when considering offspring being exposed to heat stress (HS) or thermoneutral (TN) conditions during the first half of gestation (P < 0.01; D). Postnatal HS increased HSF1 protein abundance compared to TN postnatal conditions (19%; P = 0.04; C); postnatal treatment interacted with prenatal conditions when considering offspring being subjected to either HS or TN conditions during the first half of gestation (P = 0.05; E). An interaction was also observed between gestational thermal treatment and sex for MEF2A (P = 0.05; F). No main or interactive effects were detected for MYOD1 protein abundance (G). Protein band intensities corresponding to the primary antibody were normalized to Ponceau S staining for each respective lane following densitometric analysis. Different letters indicate P < 0.05. 179

CHAPTER 6. GENERAL DISCUSSION

Despite modern advancements in livestock facility design, heat stress (HS) continues to adversely affect efficient animal production and threatens future food security for a growing population. The effects of HS imposed on the pig involve the coordination of whole- animal physiological, metabolic, endocrine, and molecular mechanisms at the cellular level, culminating in many inimical phenotypes (e.g., reduced growth, altered body composition, reduced fertility, etc.). It is imperative we further our understanding of thermal biology in pigs, as well as other livestock species, to develop new strategies and/or technologies to assuage the negative consequences of the abiotic factor.

Until now, the variability in thermoregulatory parameters during acute HS has not been realized in pigs (chapter 2). Interestingly, while thermoregulation had weak associations with production variables, thermoneutral (TN) rectal temperature TR was predictive of TR during acute HS (chapter 2). Additionally, pigs deemed tolerant or susceptible to acute HS, based on our retrospective criteria, maintained their classifications during a future heat load

(Graves et al., 2018), which may be due to a genetic basis for thermoregulation (Kim et al.,

2018). Although it is known that heat-stressed pigs have increased adipose tissue, the quality of that tissue in the carcass may also be affected as HS increased moisture content in three adipose depots (chapter 3). Heat stress can impact a variety of reproductive aspects, including ovarian function, and ovarian heat shock protein responses during HS may be partly due to HS-induced leaky gut (chapter 4). Gestational HS has also been shown to program altered postnatal phenotypes in pigs (i.e., altered body compositions, increased circulating insulin) and we provide an initial understanding of the epigenetic regulation mediating those responses (chapter 5). Collectively, the studies presented in this dissertation 180 have expanded our knowledge, but also provide new avenues for future research regarding

HS in pigs.

Possible Contributions of Rectal Temperature Variation during Acute HS

In chapter 2, we observed not only an increase in thermoregulatory indices, but also an increase in the variation for each variable during acute HS relative to the TN period. One possible explanation for this increased variability in the thermoregulatory parameters could be attributed to the dynamics of the vasculature during the heat load. The cardiovascular system adapts to hyperthermic conditions by restricting blood flow to the splanchnic region while the periphery exhibits vasodilation and increased blood flow (Rowell, 1974; Hales et al., 1979; Kregel et al., 1988; Hall et al., 1999). This redistribution of blood flow is attributed as an innate strategy to facilitate heat dissipation. Considering these concepts, can the effectiveness of these vascular system mechanisms be influenced by some factor (extrinsic or intrinsic) and result in varying responses from pig to pig? For example, the effectiveness of heat dissipation via vasodilation at the periphery can be influenced by hydration level as it directly governs blood flow at the periphery (Claremont et al., 1976). However, although water intake was not monitored in this study, hematocrit measurements, are only slightly increased if at all, suggesting hydration status is marginally affected by hyperthermic conditions when ad libitum water is provided (Pearce et al., 2013; Boddicker et al., 2014;

Pearce et al., 2014).

Although less productive, it is known that the Creole pig breed has reduced thermoregulatory responses to HS compared to Large Whites, which is also accompanied by specific skin characteristics (Renaudeau et al., 2006). Specifically, the Creole has higher mast cell densities in the dermal layer and mast cells have been implicated in vasodilatory 181 responses in rodents (Schmelz and Petersen, 2001; Renaudeau et al., 2006). Moreover, mast cell density is increased during warmer seasons, regardless of breed (Renaudeau et al., 2006).

Future studies investigating thermotolerance in pigs could assess skin histology traits to see if the variability in thermoregulation is associated with varying skin characteristics. Perhaps a good starting point would be assessing skin characteristics in the most “tolerant” or

“susceptible” pigs, whatever the criteria may be, in those future studies.

In contrast to redistribution of blood flow to the periphery, maintaining a degree of perfusion to the splanchnic region may be beneficial. Exercise training increases blood flow to the ileum during subsequent heat challenges in sheep, coinciding with decreased TR during

HS (Sakurada and Hales, 1998). This is of particular interest, albeit counterintuitive due to

HS-induced vascular events, because of compromised intestinal integrity during HS (see chapter 1). Thus, increasing blood flow to the gastrointestinal tract may prevent a HS- induced hypoxic environment and downstream increased gut permeability.

In addition to the possible vasculature explanations, the variability in the thermoregulatory variables during acute HS (chapter 2) may also be partly due to differences in molecular mechanisms in the mitochondria, specifically coupling efficiency in the electron transport chain. Coupling efficiency refers to the proportion of oxygen consumption used to generate chemical energy in the form of ATP during oxidative phosphorylation (Brand,

2005). This efficiency can be altered depending on the degree of proton allocation through uncoupling proteins, which results in heat generation (Brand, 2005). Coupling efficiency may also contribute to the marginal association between TN feed intake and TN TR in chapter 2.

For instance, one gilt that ate approximately 2.5 kg during the 24 h TN period only exhibited 182

a TR of around 38.6 °C while a gilt that consumed 2 kg demonstrated an increased TR by 1.6

°C (Figure 6.1).

Figure 6.1. Relationship between rectal temperature (TR) and feed intake during period 1 (thermoneutral [TN] conditions) from the study in chapter 2.

Could the latter pig as described above have more protons being allocated toward heat production instead of being shunted through the ATPase in the electron transport chain?

Do the gilts with relatively minimal FI reduction during HS also possess better coupling efficiencies? Perhaps certain porcine mitochondrial DNA haplotypes, affiliated with differences in production-related phenotypes (St John and Tsai, 2018), could also be associated with variation in coupling efficiency. Interestingly, mitochondrial DNA 183 haplotypes differ geographically and mitochondrial mutations influencing coupling efficiency may have allowed human ancestors to adapt different climates (Wallace, 2013).

Does HS Result in Faster Corpus Luteum Regression?

It is known in many species that just prior to ovulation, there is an increase in HIF1A expression (Robinson et al., 2009). While this may suggest that the follicle is hypoxic during this time, this may be the major mechanism by which the follicle recruits blood vessels for subsequent corpus luteum formation and also allow endocrine dissemination of progesterone produced by luteal cells (Robinson et al., 2009; Ziecik et al., 2018). For instance, HIF1A protein and mRNA abundances in mice and cows, respectively, are increased in the granulosa layer during the periovulatory period (Tam et al., 2010; Berisha et al., 2017). In the pig, vascular area is increased to the follicle just before ovulation (Martelli et al., 2006;

Martelli et al., 2009) and is accompanied by thecal and granulosa cell-specific changes in

VEGF mRNA and protein abundances as the preovulatory follicle matures (Barboni et al.,

2000; Martelli et al., 2006).

After corpus luteum (CL) formation, HIF1A mRNA decreases (from d 4 to 15;

Boonyaprakob et al., 2005) or initially decreases (from d 8 to 12) and abruptly increases

(from d 12 to 14; Przygrodzka et al., 2016) during the luteal phase in pigs. Assuming the former is accurate, which is in agreement with humans and cows (van den Driesche et al.,

2008; Berisha et al., 2017), perhaps CL HIF1A mRNA could be an indicator of luteal phase progression. Based on our data in chapter 4, HS numerically decreased HIF1A mRNA abundance 12 days after estrus, which corresponds with peak circulating progesterone during the estrous cycle in pigs (Anderson, 2009). Interestingly, while circulating and CL progesterone were not decreased, HS was previously shown to reduce CL weight in these 184 samples (Bidne, 2017). Collectively, the HS-induced numerical decrease in HIF1A mRNA abundance may support the notion that HS could result in premature CL regression (Bidne,

2017). On the contrary, perhaps the HS-induced numerical decrease in HIF1A mRNA and

VEGFA protein abundances are just reflective of a smaller, yet still functional, CL. In other words, perhaps the HS CLs do not require as much angiogenesis (smaller tissue size, so less overall perfusion) as the TN CLs, so there would be decreased angiogenic signals.

Importantly, however, the study only focused on one time point during the luteal phase. Thus, it will be necessary for future studies to assess the development of the CL (e.g.,

CL structure, HIF1A mRNA abundance and other molecular markers, and CL and circulating progesterone levels) during HS at multiple time points during the luteal phase. Perhaps just as important would be to assess HS effects on vascular dynamics and hypoxia signaling in the follicle during the follicular phase and if those potential effects carry over to the CL.

HIF1A Regulation and Potential Role during HS

In chapter 5, HIF1A protein abundance was increased due to gestational HS and, as previously mentioned in the discussion, this may be the result of an oxygen-independent pathway, due to the well-characterized oxygen-mediated regulation of HIF1A protein abundance. During normoxia, HIF1A undergoes hydroxylation at specific proline residues and promotes the association of the protein with von Hippel–Lindau tumor suppressor

(pVHL) E3 ligase complex, resulting in downstream ubiquitin-mediated degradation of

HIF1A protein (Ohh et al., 2000; Jaakkola et al., 2001). The prolyl hydroxylase that catalyzes the posttranslational modifications is the oxygen-sensing mechanism as it requires molecular oxygen as a co-substrate (Jaakkola et al., 2001), suspected to be available via maintenance of perfusion to muscle during HS (Nielsen et al., 1990). 185

While not also altered due to postnatal HS in the study, thermal stimuli can influence

HIF1A protein abundance (Jackson et al., 2006; Horowitz and Assadi, 2010). Interestingly,

HIF1A mRNA and protein abundance is predominantly expressed in glycolytic muscle in rodents (Pisani and Dechesne, 2005). Whether HIF1A is involved in HS-induced metabolic alterations to glycolytic muscle is of further interest. Additionally, if and how direct effects of HS on mitochondria (Welch and Suhan, 1985; Davidson and Schiestl, 2001; Qian et al.,

2004; White et al., 2012; Huang et al., 2015) contribute to the enhanced glycolysis in glycolytic muscle could also be addressed in future studies.

Other Thoughts Regarding HS during Pre- or Postnatal Development

Interestingly, prenatal HS results in phenotypes that have been previously demonstrated with heat challenges during postnatal development (chapter 1 and 5). It is justifiable to postulate whether gestational HS also programs leaky gut postnatally, which could be indicated from upregulated immune pathways identified in the liver with Ingenuity

Pathway Analysis in chapter 5. Increased intestinal permeability has been programed in offspring in a murine maternal obesity model, suggesting programming of compromised intestinal integrity is plausible (Xue et al., 2014). However, future studies assessing circulating endotoxins, intestinal histology, as well as ex vivo intestinal measurements (i.e., assess via Ussing chambers) in gestationally heat-stressed offspring are needed to evaluate the hypothesized effect. Other future studies could also evaluate the contribution of compromised intestinal integrity in the maternal environment during gestational HS. Does gestational LPS exposure program similar postnatal phenotypes as those observed from HS during intrauterine development (i.e., more adipose tissue, less skeletal muscle, and increased circulating insulin)? Conversely, evaluating the effects of HS on gnotobiotic animals (i.e., 186

“germ free”) may also provide insight regarding the contribution of HS-induced compromised intestinal integrity during pre- and/or postnatal development (Martin et al.,

2016).

Heat shock proteins can have a direct influence on epigenetic modifications (Fritah et al., 2009; Khurana and Bhattacharyya, 2015) and could have key roles in programmed phenotypes resulting from intrauterine HS. Perhaps the potential for thermal programming could also occur during oocyte development as HSPs were increased in the ovary (chapter 4) and have been selectively increased in the oocyte (Pennarossa et al., 2012). Heat shock proteins may also be involved in the epigenetic mechanisms that underlie thermal conditioning and acclimation (Kisliouk et al., 2009; Tetievsky and Horowitz, 2010), responses of which are ill-defined in pigs.

In summary, HS undermines efficient animal production and results in major economic losses annually. Hyperthermia involves the interplay of physiological and molecular mechanisms that culminate in HS-induced phenotypes observed in pigs as well as other important agriculture-related species. In this dissertation, we demonstrate that the thermoregulatory response is highly variable and is only marginally related to production responses to acute HS in pigs. Heat stress increased adipose moisture content, which may be related to altered carcass fat quality during the summer months. Ovarian heat shock proteins are increased in response to in vivo HS, but this effect may partly be the result of HS-induced compromised intestinal integrity. Gestational HS influences postnatal development and we provide an initial understanding of the specific protein and epigenetic responses that may contribute to thermally programmed phenotypes. The studies herein provide future areas of research to further our understanding of thermal biology in pigs. 187

Literature Cited

Anderson, L. L. 2009. Reproductive Biology of Pigs. Animal Industry Report.

Barboni, B. et al. 2000. Vascular endothelial growth factor production in growing pig antral follicles. Biol Reprod 63: 858-864.

Berisha, B., D. Schams, D. Rodler, F. Sinowatz, and M. W. Pfaffl. 2017. Expression pattern of HIF1alpha and vasohibins during follicle maturation and corpus luteum function in the bovine ovary. Reprod. Domest. Anim. 52: 130-139.

Bidne, K. L. 2017. Investigating the ovarian response to endotoxemia, Iowa State University, Ames, IA.

Boddicker, R. L. et al. 2014. Gestational Heat Stress Alters Postnatal Offspring Body Composition Indices and Metabolic Parameters in Pigs. PloS ONE 9.

Boonyaprakob, U., J. E. Gadsby, V. Hedgpeth, P. A. Routh, and G. W. Almond. 2005. Expression and localization of hypoxia inducible factor-1alpha mRNA in the porcine ovary. Can J Vet Res 69: 215-222.

Brand, M. D. 2005. The efficiency and plasticity of mitochondrial energy transduction. Biochem Soc Trans 33: 897-904.

Claremont, A. D., D. L. Costill, W. Fink, and P. Van Handel. 1976. Heat tolerance following diuretic induced dehydration. Med Sci Sports 8: 239-243.

Davidson, J. F., and R. H. Schiestl. 2001. Mitochondrial respiratory electron carriers are involved in oxidative stress during heat stress in Saccharomyces cerevisiae. Mol. Cell. Biol. 21: 8483-8489.

Fritah, S. et al. 2009. Heat-shock factor 1 controls genome-wide acetylation in heat-shocked cells. Mol Biol Cell 20: 4976-4984.

Graves, K. L., J. T. Seibert, A. F. Keating, L. H. Baumgard, and J. W. Ross. 2018. Characterizing the acute heat stress response in gilts: II. Assessing repeatability and association with fertility. J Anim Sci 96: 2419-2426.

Hales, J. R., L. B. Rowell, and R. B. King. 1979. Regional distribution of blood flow in awake heat-stressed baboons. Am J Physiol 237: H705-712.

Hall, D. M., K. R. Baumgardner, T. D. Oberley, and C. V. Gisolfi. 1999. Splanchnic tissues undergo hypoxic stress during whole body hyperthermia. Am J Physiol 276: G1195- 1203.

188

Horowitz, M., and H. Assadi. 2010. Heat acclimation-mediated cross-tolerance in cardioprotection: do HSP70 and HIF-1alpha play a role? Ann N Y Acad Sci 1188: 199-206.

Huang, C. et al. 2015. Heat stress impairs mitochondria functions and induces oxidative injury in broiler chickens. J Anim Sci 93: 2144-2153.

Jaakkola, P. et al. 2001. Targeting of HIF-alpha to the von Hippel-Lindau ubiquitylation complex by O2-regulated prolyl hydroxylation. Science 292: 468-472.

Jackson, I. L., I. Batinic-Haberle, P. Sonveaux, M. W. Dewhirst, and Z. Vujaskovic. 2006. ROS production and angiogenic regulation by macrophages in response to heat therapy. Int J Hyperthermia 22: 263-273.

Khurana, N., and S. Bhattacharyya. 2015. Hsp90, the concertmaster: tuning transcription. Front Oncol 5: 100.

Kim, K. S. et al. 2018. Characterization of the acute heat stress response in gilts: III. Genome-wide association studies of thermotolerance traits in pigs. J Anim Sci 96: 2074-2085.

Kregel, K. C., P. T. Wall, and C. V. Gisolfi. 1988. Peripheral vascular responses to hyperthermia in the rat. J Appl Physiol (1985) 64: 2582-2588.

Martelli, A. et al. 2006. Spatio-temporal analysis of vascular endothelial growth factor expression and blood vessel remodelling in pig ovarian follicles during the periovulatory period. J Mol Endocrinol 36: 107-119.

Martelli, A. et al. 2009. Blood vessel remodeling in pig ovarian follicles during the periovulatory period: an immunohistochemistry and SEM-corrosion casting study. Reprod. Biol. Endocrinol 7: 72.

Martin, R., L. G. Bermudez-Humaran, and P. Langella. 2016. Gnotobiotic Rodents: An In Vivo Model for the Study of Microbe-Microbe Interactions. Front Microbiol 7: 409.

Nielsen, B., G. Savard, E. A. Richter, M. Hargreaves, and B. Saltin. 1990. Muscle blood flow and muscle metabolism during exercise and heat stress. J Appl Physiol 69: 1040- 1046.

Ohh, M. et al. 2000. Ubiquitination of hypoxia-inducible factor requires direct binding to the beta-domain of the von Hippel-Lindau protein. Nat Cell Biol 2: 423-427. 189

Pearce, S. C. et al. 2013. Heat stress reduces intestinal barrier integrity and favors intestinal glucose transport in growing pigs. PloS ONE 8: e70215.

Pennarossa, G. et al. 2012. Characterization of the constitutive pig ovary heat shock chaperone machinery and its response to acute thermal stress or to seasonal variations. Biol Reprod 87: 119.

Pisani, D. F., and C. A. Dechesne. 2005. Skeletal muscle HIF-1alpha expression is dependent on muscle fiber type. J Gen Physiol 126: 173-178.

Przygrodzka, E., M. M. Kaczmarek, P. Kaczynski, and A. J. Ziecik. 2016. Steroid hormones, prostanoids, and angiogenic systems during rescue of the corpus luteum in pigs. Reproduction 151: 135-147.

Qian, L., X. Song, H. Ren, J. Gong, and S. Cheng. 2004. Mitochondrial mechanism of heat stress-induced injury in rat cardiomyocyte. Cell Stress & Chaperones 9: 281-293.

Renaudeau, D., M. Leclercq-Smekens, and M. Herin. 2006. Differences in skin characteristics in European (Large White) and Caribbean (Creole) growing pigs with reference to thermoregulation. Anim. Res. 55: 209-217.

Robinson, R. S. et al. 2009. Angiogenesis and vascular function in the ovary. Reproduction 138: 869-881.

Rowell, L. B. 1974. Human cardiovascular adjustments to exercise and thermal stress. Physiol Rev 54: 75-159.

Sakurada, S., and J. R. Hales. 1998. A role for gastrointestinal endotoxins in enhancement of heat tolerance by physical fitness. J Appl Physiol (1985) 84: 207-214.

Schmelz, M., and L. J. Petersen. 2001. Neurogenic inflammation in human and rodent skin. News Physiol Sci 16: 33-37.

St John, J. C., and T. S. Tsai. 2018. The association of mitochondrial DNA haplotypes and phenotypic traits in pigs. BMC Genet 19: 41.

Tam, K. K. et al. 2010. Hormonally regulated follicle differentiation and luteinization in the mouse is associated with hypoxia inducible factor activity. Mol Cell Endocrinol 327: 47-55.

190

Tetievsky, A., and M. Horowitz. 2010. Posttranslational modification in histones underlie heat acclimation-mediated cytoprotective memory. J Appl Physiol 109: 1552-1561. van den Driesche, S., M. Myers, E. Gay, K. J. Thong, and W. C. Duncan. 2008. HCG upregulates hypoxia inducible factor-1 alpha in luteinized granulosa cells: implications for the hormonal regulation of vascular endothelial growth factor A in the human corpus luteum. Mol Hum Reprod 14: 455-464.

Wallace, D. C. 2013. A mitochondrial bioenergetic etiology of disease. J Clin Invest 123: 1405-1412.

Welch, W. J., and J. P. Suhan. 1985. Morphological study of the mammalian stress response: characterization of changes in cytoplasmic organelles, cytoskeleton, and nucleoli, and appearance of intranuclear actin filaments in rat fibroblasts after heat-shock treatment. J Cell Biol 101: 1198-1211.

White, M. G. et al. 2012. Mitochondrial dysfunction induced by heat stress in cultured rat CNS neurons. J Neurophysiol 108: 2203-2214.

Xue, Y., H. Wang, M. Du, and M. J. Zhu. 2014. Maternal obesity induces gut inflammation and impairs gut epithelial barrier function in nonobese diabetic mice. J. Nutr. Biochem. 25: 758-764.

Ziecik, A. J., E. Przygrodzka, B. M. Jalali, and M. M. Kaczmarek. 2018. Regulation of the porcine corpus luteum during pregnancy. Reproduction 156: R57-R67. 191

APPENDIX A. SUPPLEMENTARY TABLES AND FIGURES FOR CHAPTER 5

Table A.1. Genes selected based on the RNA-Seq analysis for bisulfite sequencing of targeted genomic regions Gene Description CARD10 Caspase recruitment domain family, member 10 CHST1 Carbohydrate (keratan sulfate Gal-6) sulfotransferase 1 CTNNA1 Catenin (cadherin-associated protein), alpha 1, 102kDa FAR1 Fatty acyl CoA reductase 1 FOSL1 FOS-like antigen 1 GPR126 G protein-coupled receptor 126 GPT Glutamic-pyruvate transaminase (alanine aminotransferase) Hypoxia-inducible factor 1, alpha subunit (basic helix-loop- HIF1A helix transcription factor) HSF1 Heat shock transcription factor 1 MANEA Mannosidase, endo-alpha MYOD1 Myogenic differentiation 1 PGGT1B Protein geranylgeranyltransferase type I, beta subunit PHF20L1 PHD finger protein 20-like 1 PIDD p53-induced death domain protein PMAIP1 Phorbol-12-myristate-13-acetate-induced protein 1 Phosphorylase, glycogen; muscle (McArdle syndrome, PYGM glycogen storage disease type V) Ras-related C3 botulinum toxin substrate 1 (rho family, small RAC1 GTP binding protein Rac1) ZNHIT2 Zinc finger, HIT type 2 192

Table A.2. Functional annotation clusters of biological terms representing processes affected from gestational heat stress (HS) in adipose tissue Annotation Enrichment Cluster1 Score2 Biological Terms3 1 1.74 ubiquitin-conjugating enzyme E2, catalytic domain homologues (3), ubiquitin-conjugating enzyme/RWD-like (3), active site:Glycyl thioester intermediate (3), ubiquitin-protein ligase activity (3), small conjugating protein ligase activity (3), ubiquitin mediated proteolysis (3), acid-amino acid ligase activity (3), ligase activity, forming carbon-nitrogen bonds (3), ligase (3) 1The fifteen most significant annotation clusters identified based on the gene list submitted for analysis through DAVID. 2The enrichment score is determined through DAVID and ranks the significance of each annotation cluster based on the relatedness of the terms and the genes associated with them. 3This column represents terms in the annotation clusters. The gene ontology (GO) terms were gathered based on the known annotation of the submitted genes with respect to biological process, cellular component, and molecular function; as well as biological pathway membership and protein domains. The number in parenthesis indicate the number of differentially expressed genes contribute to the clustered term. 193

Table A.3. Functional annotation clusters of biological terms representing processes affected from gestational heat stress (HS) in liver Annotation Enrichment Cluster1 Score2 Biological Terms3 1 2.50 ATP synthesis coupled electron transport (6), mitochondrial ATP synthesis coupled electron transport (6), cellular respiration (7), respiratory electron transport chain (6), respiratory chain (6), energy derivation by oxidation of organic compounds (8), NADH dehydrogenase complex (5), mitochondrial respiratory chain complex I (5), respiratory chain complex I (5), respiratory chain (6), mitochondrial electron transport, NADH to ubiquinone (5), NADH dehydrogenase (ubiquinone) activity (5), NADH dehydrogenase activity (5), NADH dehydrogenase (quinone) activity (5), oxidoreductase activity, acting on NADH or NADPH, quinone or similar compound as acceptor (5), electron transport (6), ubiquinone (4), oxidative phosphorylation (6), mitochondrial respiratory chain (5), oxidative phosphorylation (7), electron transport chain (6), oxidoreductase activity, acting on NADH or NADPH (5), Alzheimer's disease (7), Parkinson's disease (6), mitochondrial inner membrane (8), mitochondrial membrane part (5), mitochondrion inner membrane (6), generation of precursor metabolites and energy (8), organelle inner membrane (8), Huntington's disease (6), mitochondrial membrane (8), mitochondrial envelope (8) 2 1.94 phosphorus metabolic process (19), phosphate metabolic process (19), phosphorylation (15) 3 1.82 WD40 repeat, subgroup (8), WD40 repeat, conserved site (8), WD40/YVTN repeat-like (8) 4 1.79 ribosome (6), cytosolic large ribosomal subunit (4), cytosolic ribosome (5), ribosomal subunit (6), translational elongation (5), structural constituent of ribosome (6), large ribosomal subunit (4), protein biosynthesis (6), ribosomal protein (6), cytosolic part (5), translation (7) 5 1.65 SH2 domain (5), SH2 motif (5), SH2 (5) 6 1.57 microtubule-based movement (6), microtubule (7), microtubule (7), microtubule-based process (6) 7 1.57 atp-binding (21), nucleotide binding (33), ATP binding (24), adenyl ribonucleotide binding (24), purine ribonucleotide binding (28), ribonucleotide binding (28), nucleotide-binding (24), purine nucleotide binding (28), adenyl nucleotide binding (24), purine nucleoside binding (24), nucleoside binding (24) 8 1.53 blood vessel morphogenesis (7), blood vessel development (7), vasculature development (7)

194

Table A.3 continued Annotation Enrichment Cluster1 Score2 Biological Terms3 9 1.38 tyrosine-specfic protein kinase (4), tyrosine protein kinase, active site (5), tyrosine-protein kinase (5), tyrosine protein kinase (5), protein amino acid autophosphorylation (4), autophosphorylation (3), peptidyl-tyrosine phosphorylation (3), peptidyl-tyrosine modification (3), phosphotransferase (4) 1The fifteen most significant annotation clusters identified based on the gene list submitted for analysis through DAVID. 2The enrichment score is determined through DAVID and ranks the significance of each annotation cluster based on the relatedness of the terms and the genes associated with them. 3This column represents terms in the annotation clusters. The gene ontology (GO) terms were gathered based on the known annotation of the submitted genes with respect to biological process, cellular component, and molecular function; as well as biological pathway membership and protein domains. The number in parenthesis indicate the number of differentially expressed genes contribute to the clustered term.

195

Table A.4. Functional annotation clusters of biological terms representing processes affected from gestational heat stress (HS) in longissimus dorsi muscle Annotation Enrichment Cluster1 Score2 Biological Terms3 1 2.71 intracellular organelle lumen (83), organelle lumen (84), membrane-enclosed lumen (85), nuclear lumen (66) 2 2.11 HR1-like rho-binding repeat (4), rho effector or protein kinase C- related kinase homology region 1 homologues (3), protein kinase C activity (3) 3 2.08 zinc (94), transition metal ion binding (118), zinc ion binding (98) 4 2.07 UBX, domain present in ubiquitin-regulatory proteins (4) 5 1.82 nucleotide-binding (78), atp-binding (64), adenyl ribonucleotide binding (68), ATP binding (67), ribonucleotide binding (79), purine ribonucleotide binding (79), purine nucleoside binding (70), adenyl nucleotide binding (69), nucleoside binding (70), purine nucleotide binding (80), nucleotide binding (92) 6 1.79 protein transport (40), establishment of protein localization (40), protein localization (43) 7 1.73 SNARE interactions in vesicular transport (6), t SNARE (5), target SNARE coiled-coil region (5), syntaxin/epimorphin, conserved site (4), SNAP receptor activity (4) 8 1.58 phosphorylation (41), phosphorus metabolic process (46), phosphate metabolic process (46) 9 1.55 protein kinase, C-terminal (6), extension to serine/threonine-type protein kinases (6), AGC-kinase, C-terminal (6) 10 1.55 ATP binding site (32), protein kinase domain (25), protein kinase, ATP binding site (25), protein kinase, core (25), protein kinase activity (29), protein amino acid phosphorylation (32), serine/threonine protein kinase-related (19) 11 1.40 organelle inner membrane (20), mitochondrial inner membrane (18), mitochondrial membrane (20), mitochondrial envelope (20) 12 1.37 thioredoxin domain (5), thioredoxin, conserved site (4), cell redox homeostasis (6), thioredoxin-like (4) 13 1.35 SH2 domain (9), SH2 motif (9), SH2 (9) 14 1.32 transcription from RNA polymerase II promoter (15), transcription, DNA-dependent (17), RNA biosynthetic process (17) 1The fifteen most significant annotation clusters identified based on the gene list submitted for analysis through DAVID. 2The enrichment score is determined through DAVID and ranks the significance of each annotation cluster based on the relatedness of the terms and the genes associated with them. 3This column represents terms in the annotation clusters. The gene ontology (GO) terms were gathered based on the known annotation of the submitted genes with respect to biological process, cellular component, and molecular function; as well as biological pathway membership and protein domains. The number in parenthesis indicate the number of differentially expressed genes contribute to the clustered term.

196

Table A.5. Functional annotation clusters from Ingenuity Pathway Analysis of biological terms representing processes affected from gestational heat stress (HS) in longissimus dorsi muscle z-score1 Cellular Function TNHS HSTN HSHS P Muscle apoptosis 0.6 1.4 0.4 <0.01 cell death of kidney cells 0.0 2.0 -1.0 <0.01 cell death of hematopoietic 1.3 2.0 0.4 <0.01 progenitor cells apoptosis of fibroblast cell 1.0 2.3 0.6 <0.01 lines cell death of neuroblastoma 0.3 2.6 -0.4 <0.01 cell lines metabolism of protein 2.2 0.8 2.1 0.01 quantity of reactive oxygen 1.7 2.1 1.9 0.01 species cell death of cortical neurons 0.9 1.9 0.8 0.01 autophagy -1.0 -1.7 -1.2 0.01 cell death of thymocytes 2.1 2.6 1.4 0.01 apoptosis of muscle cell lines 0.5 2.2 0.5 0.01 cell death of epithelial cells 0.8 1.8 -0.2 0.02 proliferation of cells 1.2 -0.6 0.5 0.02 apoptosis of hematopoietic 1.1 2.1 0.7 0.02 progenitor cells Liver adhesion of macrophages 2.0 2.1 2.0 <0.01 migration of leukocyte cell 2.0 2.0 2.0 <0.01 lines differentiation of mononuclear 3.1 1.8 2.1 0.01 leukocytes production of reactive oxygen 1.9 2.5 2.2 0.01 species 1Refers to the level of activation of the implicated biological function relative to the controls subjected to thermoneutral (TN) conditions during the entire gestation (TNTN). Other gestational treatments: TNHS = thermoneutral during first half of gestation and heat stress during second half of gestation; HSTN = heat stress during first half of gestation and thermoneutral during second half of gestation; HSHS = heat stress during the entire gestation 197

Table A.6. Genomic loci of amplicons from bisulfite sequencing and number of CpGs assessed/amplicon Chromosome Strand Start End Forward Reverse Nearby Gene1 No. CpGs 5 + 7899113 7899288 GAGGAAGAGGTGTTGAGTATTTA TTCTAAAACTAAAATAATAAACACACCAA CARD10 16 2 - 18179501 18179764 GTTGAATTTTAAGGTTATTTAGTTGGTT TCTTCATAAAATTCCTAACCAAATCC CHST1 23 GAGTTTTTAYGATTGTGATTTTTATTTTTTGG 2 - 18179854 18180061 CHST1 18 AG ACCRCCAAAATCAAATTCAACAAAC AAATAAAAAACRTCCAAATACTAATTAAA 2 - 18180169 18180467 CHST1 18 GTGGAGAYGTAGTTTTATGAAGGTGTTTA AAACTAAC 2 + 146440487 146440687 GGYGGAGAGTAAGAGGGTTTTT CCTATCCRACTCCAAAAAAAAAATAAACC CTNNA1 22 2 + 146440654 146440922 GGGYGGTTTATTTTTTTTTTGGAG ACCAAATCCCAAACCCRAAAC CTNNA1 33 2 - 48903708 48903985 GTTTYGTTTTTATTTTGGGTTGTTTTAAGTAG CRAAATCCCCATCACAAACCTTT FAR1 20 2 - 48904085 48904378 AGTTAGGAGGAGGGTGAGAGGTTT ACCAAACRTCCCTAATTCTCACCTAAA FAR1 26 2 - 48904305 48904505 GGTYGGTTGAAGAGTTTATTTTT CTTTTCCACRTCTTATCTTAATTAACTCTCT FAR1 16 2 + 5534097 5534334 GAGGAAATTGTATTATGTATGGGTAGTT CRAACTAAACCACTTCCACTAAAA FOSL1 19 RTAAAATTCCTCTCACCTACTAACCTATCA 2 + 5534249 5534538 FOSL1 18 GGTGYGATTGTAGAGGTGGAGTTT C 2 + 5538545 5538826 FOSL1 20 GTTTYGTTAGGGTTTTATTTTTGGGTTAG TAAACCRATACCCCCTACAACCC 196 1 + 25289686 25289884 GTTAGTGGTTYGGGAATTTGGTAG AAACCCRAAAATCACCTAATAACCTAAC GPR126 13 ATTCCTCCTCCCAAAAAAATTAAAAAAAAA 1 + 25290241 25290443 GPR126 16 GGTAYGAGAGTTATAAAGGGGTGATTA TTAAAAC TTTTAATTTTTTTTTAATTTTTTTGGGAGGAG 1 + 25290407 25290707 GPR126 24 G TTACTCAACTAAAATCCCCAAAAACC 4 - 390204 390413 GTYGTGTAGGTGGATTATTATTTGG AAACRATCAAAAACAAACTAACCC GPT 32 GTTTAGGGGTTTATAGTATTAGTTTTGGTATT 4 - 390492 390773 GPT 26 TAG CCRAACACTAACAACTTAAACAC ACAAAACCCAAAACAAAACCTACAAAATA 4 - 394652 394887 GPT 15 GAGGGGTAGGGGGTATTTATGATTTTT AA 4 - 394791 395033 GTTTYGGAATTTTTAGTAATTTGGGTTT CTTATACTTCAACTCCAACAACTTAACA GPT 8 1 + 212130642 212130906 GTTYGTTATTGGTTTTTAAGGAATT TCCCCCRAAACAACTTATCAAAAC HIF1A 34 CTAACCRACRAAAAAATAAACTTACTTTTT 1 + 212130889 212131164 HIF1A 27 TAAGTTGTTTYGGGGGAGGTTTTT CTTATC 4 - 630587 630874 GTTTTATTAGTTTGTTGAGTTGGTTTTT AAACCTATACACTTCAAAATCTTCCT HSF1 19 4 - 632244 632492 GGGTAGGTGATTTATATGGTTTGATT ATAACAAACAACAACTCAATAAAATCCA HSF1 16 4 - 632588 632852 TTGGGTAGGAAGGGTTTTTGTTT AACTCRAAACCAACCTAAACTTTATAA HSF1 15 1 + 71027048 71027290 GTTTGTGTTTTGAAATTAGTGTTTTATTTATT CATATCTACAAAACCCTAAATCTACCA MANEA 19 2 - 44484006 44484306 AGYGGTTGTTTAAGGTGGAAATTTT CAACTAAAAACCAACTCCTTACCCTC MYOD1 28

198

Table A.6 continued Chromosome Strand Start End Forward Reverse Nearby Gene1 No. CpGs 2 - 44484278 44484518 GGTTTCGGGAGTTTTTGAGGA CRCAAAATTTCCACCTTAAACAACC MYOD1 25 2 - 44484488 44484777 TYGTTGGGATTGGGTAAAATTAGGAT ACATACTCRTCCTCAAAAACTCC MYOD1 27 2 - 124301896 124302191 GTATTTYGGTAGGGTTGTGTATTTT CRAAACCTCTAAACCCCTAAAACC PGGT1B 33 2 - 124334234 124334519 TTATTATYGGGAAGAATAGGGGTAGG TAACCCCCAATTACCTACTAATCTCAA PGGT1B 32 4 - 8510218 8510513 GGTTTGAAGGGTTYGGATTTGTTTT AAAACRAACCCAAAAAAAACACCAAC PHF20L1 35 4 - 8510484 8510698 GTTTTTTAGGGYGGTTTTTTAAT AAAAAAAACAAATCCRAACCCTTCAAAC PHF20L2 33 2 + 168960 169260 TGGGTTTTGTAGGGTTTGAGAGG CTCRACTACACTCAACTTCCTACC PIDD 27 2 + 169288 169554 GAGGGGTTYGGAGATTTTTGTT ACTTCCTAACCCCTAAAAAAAAACTCTAC PIDD 43 2 + 171315 171521 TTTTTGGGTTTATGGAGAGTGTTTA TACAAAAACCCTTTCAATCCCTAC PIDD 14 2 + 171496 171675 GGTAGGGATTGAAAGGGTTTTTGTAG CRAACCAAACCAACAAACACTAAAAC PIDD 12 1 - 178857039 178857332 TAGTTTGTTATTTGTAGAGTTAAGGTG RCCCCTAAAACATATACAAAAACA PMAIP1 23 1 - 178857389 178857679 TAAGGAAGTAATTTTTAGGTTGGGTTT CTAAACRCCCAAAAACCTCTAAA PMAIP1 36 2 + 6541990 6542288 TTGTTTTTGATTTTGTAAAAGGTAGGGAG CCTAAACRATACCTCCAACCAAAAA PYGM 17 197 2 + 6546787 6547087 GTGTGGTTGGGAGTTTTTTTTGG ACTAAAACRAAAAAAACTAACACACATC PYGM 24

5 + 8135280 8135482 TTTTTYGTTAAAGATGGGGATAGAG CAACRCCTTCTAAATCTAAAAATACC RAC1 10 5 + 8135447 8135647 GGTTTTAYGGGTATTTTTAGATTTAGA CCRACTAAAACTAAAAATAAATAAATTCCT RAC1 1 2 + 6207307 6207599 AAATTGTYGAGAGTTTTTAGGGTTT CRCAACTCTCCCAACACCTAATCA ZNHIT2 28 2 + 6207562 6207739 AAGATTTTTATYGTGATTAGGTGTTGGG ACCRCTCCCAAAATCCTAAAAAAC ZNHIT2 19 2 + 6207849 6208087 TTTTTGGAGGAATTGGGTGATGTT AAACRAACAACACATTAAACAACTAAAAA ZNHIT2 30 1Caspase recruitment domain family, member 10 (CARD10); Carbohydrate (keratan sulfate Gal-6) sulfotransferase 1 (CHST1); Catenin (cadherin-associated protein), alpha 1, 102kDa (CTNNA1); Fatty acyl CoA reductase 1 (FAR1); FOS-like antigen 1 (FOSL1); G protein-coupled receptor 126 (GPR126); Glutamic-pyruvate transaminase (alanine aminotransferase) (GPT); Hypoxia-inducible factor 1, alpha subunit (basic helix-loop-helix transcription factor) (HIF1A); Heat shock transcription factor 1 (HSF1); Mannosidase, endo-alpha (MANEA); Myogenic differentiation 1 (MYOD1); Protein geranylgeranyltransferase type I, beta subunit (PGGT1B); PHD finger protein 20-like 1 (PHF20L1); p53-induced death domain protein (PIDD); Phorbol-12-myristate-13-acetate-induced protein 1 (PMAIP1); Phosphorylase, glycogen; muscle (McArdle syndrome, glycogen storage disease type V) (PYGM); Ras-related C3 botulinum toxin substrate 1 (rho family, small GTP binding protein Rac1) (RAC1); Zinc finger, HIT type 2 (ZNHIT2)

199

Table A.7. Quality of the targeted bisulfite sequencing analysis Total Read Pair Mapping Unique Avg. CpG Bisulfite Conversion ID Number Efficiency CpGs Coverage Rate 2 290,889 64% 1,010 5,526X >99% 3 350,175 68% 986 7,489X >99% 17 345,431 64% 969 6,934X >99% 19 284,263 67% 943 6,161X >99% 20 293,615 67% 1,011 5,893X >99% 26 276,795 71% 1,010 6,079X >99% 30 343,786 64% 986 6,647X >99% 32 325,393 67% 993 6,770X >99% 42 120,499 62% 936 2,297X >99% 44 254,702 63% 987 4,635X >99% 45 234,984 65% 984 4,803X >99% 46 232,568 63% 969 4,613X >99% 57 248,778 64% 1,010 4,540X >99% 58 147,775 63% 986 2,801X >99% 59 208,337 66% 986 4,141X >99% 60 140,597 64% 968 2,756X >99% 63 116,113 65% 987 2,273X >99% 73 238,781 67% 986 4,997X >99% 79 180,545 63% 985 3,447X >99% 80 297,131 64% 969 5,970X >99% 81 217,855 66% 1,008 4,355X >99% 82 230,932 63% 986 4,433X >99% 83 132,149 68% 986 2,777X >99% 85 102,934 60% 986 1,767X >99% 89 217,748 64% 966 4,281X >99% 105 268,503 65% 986 5,038X >99% 109 246,773 64% 952 4,990X >99% 110 213,390 63% 936 4,339X >99% 111 280,976 63% 986 5,216X >99% 113 218,705 64% 969 4,397X >99% 114 217,143 63% 1,010 3,869X >99% 115 155,690 61% 978 2,901X >99% 118 101,445 63% 986 1,928X >99% 130 207,908 67% 1,010 4,141X >99% 132 252,734 66% 969 5,202X >99% 133 210,881 63% 986 4,066X >99% 134 248,103 56% 969 4,262X >99% 152 100,674 68% 975 2,135X >99% 153 213,273 65% 931 4,313X >99% 154 97,004 70% 1,009 2,042X >99% 157 177,581 62% 985 3,281X >99% 166 177,343 67% 986 3,627X >99% 200

Table A.7 continued Total Read Pair Mapping Unique Avg. CpG Bisulfite Conversion ID Number Efficiency CpGs Coverage Rate 167 230,188 68% 1,010 4,828X >99% 171 274,208 63% 969 5,270X >99% 174 209,730 64% 986 4,077X >99% 176 181,204 65% 993 3,566X >99% 186 233,922 65% 1,010 4,535X >99% 187 142,118 66% 986 2,898X >99%

201

Table A.8. Slope estimates for the relationship of transcript expression of select genes from the RNA-Seq analysis and normalized individual methylation status of CpGs at targeted genomic loci within 4kb of each respective gene Base Chromosome position Gene1 Strand Slope P q 4 8510552 PHF20L1 - -0.0013 0.001 0.284 1 178857564 PMAIP1 - 0.3902 0.001 0.284 2 146440738 CTNNA1 + -0.0840965 0.001 0.312 4 8510421 PHF20L1 - 0.1458 0.006 0.905 5 8135334 RAC1 + -0.0038 0.006 0.905 1 212130762 HIF1A + -0.018301 0.007 0.905 2 48904448 FAR1 - -0.0013063 0.008 0.905 5 8135408 RAC1 + 0.0045 0.011 0.905 2 169130 PIDD + -1.3197 0.012 0.905 2 146440526 CTNNA1 + -0.0120373 0.012 0.905 2 18179655 CHST1 - 0.00057265 0.013 0.905 2 146440780 CTNNA1 + -0.0216047 0.013 0.905 2 6207944 ZNHIT2 + 0.0055 0.015 0.905 2 124302127 PGGT1B - -0.0011 0.016 0.905 2 44484689 MYOD1 - -0.4242 0.017 0.905 4 390645 GPT - 0.01629379 0.017 0.905 2 18179888 CHST1 - -0.0012639 0.017 0.905 2 6207348 ZNHIT2 + 0.0124 0.018 0.905 2 48904469 FAR1 - 0.10831734 0.019 0.905 1 25290456 GPR126 + -0.1741218 0.019 0.905 2 171400 PIDD + -0.0007 0.021 0.905 1 212130824 HIF1A + -0.011656 0.022 0.905 4 630833 HSF1 - 0.00035485 0.026 0.905 2 6207374 ZNHIT2 + -0.0101 0.027 0.905 2 44484391 MYOD1 - 0.0003 0.027 0.905 2 5534403 FOSL1 + -0.3127787 0.029 0.905 4 630646 HSF1 - 0.00116124 0.032 0.905 5 7899190 CARD10 + 0.00035053 0.034 0.905 5 8135454 RAC1 + -0.0939 0.035 0.905 4 632748 HSF1 - 0.00086575 0.035 0.905 2 5538801 FOSL1 + -0.0335602 0.035 0.905 2 171491 PIDD + 0.0014 0.035 0.905 2 44484226 MYOD1 - 0.0036 0.036 0.905 2 18180306 CHST1 - -0.053442 0.036 0.905 2 6207557 ZNHIT2 + -0.1728 0.037 0.905 2 18179973 CHST1 - 0.0002837 0.037 0.905 2 18179631 CHST1 - -0.0002936 0.038 0.905 2 48904427 FAR1 - -0.022446 0.038 0.905 202

Table A.8 continued Base Chromosome position Gene1 Strand Slope P q 2 171402 PIDD + 0.0012 0.038 0.905 5 7899171 CARD10 + 0.00045401 0.039 0.905 1 212131077 HIF1A + -0.3511686 0.040 0.905 2 48903790 FAR1 - -0.0008958 0.040 0.905 5 8135386 RAC1 + -0.0028 0.040 0.905 2 6207487 ZNHIT2 + -0.2508 0.045 0.911 1 212131068 HIF1A + 0.25033152 0.046 0.911 2 146440895 CTNNA1 + 0.00083831 0.049 0.911 4 632616 HSF1 - -0.0017429 0.051 0.911 2 146440747 CTNNA1 + -0.0609245 0.052 0.911 1 71027189 MANEA + -0.0005808 0.059 0.911 2 169327 PIDD + -0.0035 0.059 0.911 1 25290680 GPR126 + 0.04676346 0.059 0.911 4 632372 HSF1 - -0.0002309 0.060 0.911 1 25290283 GPR126 + 0.03584546 0.060 0.911 2 169190 PIDD + -0.0131 0.060 0.911 5 7899238 CARD10 + -0.0001979 0.061 0.911 4 390547 GPT - 0.01623522 0.061 0.911 2 169468 PIDD + 0.0081 0.061 0.911 2 168987 PIDD + 0.0072 0.062 0.911 2 6207884 ZNHIT2 + 0.0238 0.063 0.911 2 18180322 CHST1 - -0.0625368 0.065 0.911 2 146440892 CTNNA1 + 0.06309619 0.065 0.911 2 171562 PIDD + -0.0001 0.069 0.911 2 48903933 FAR1 - 0.0018672 0.071 0.911 2 146440810 CTNNA1 + 0.00070354 0.072 0.911 2 48903828 FAR1 - -0.0011072 0.072 0.911 2 169118 PIDD + 0.6568 0.073 0.911 2 6207573 ZNHIT2 + 0.3646 0.074 0.911 1 212130851 HIF1A + 0.0310801 0.077 0.911 2 48904337 FAR1 - 0.00049499 0.077 0.911 2 44484163 MYOD1 - 0.0233 0.081 0.911 2 44484587 MYOD1 - 0.4179 0.082 0.911 2 124302085 PGGT1B - -0.0347 0.082 0.911 2 6207384 ZNHIT2 + 0.0874 0.083 0.911 1 212131022 HIF1A + -0.1498507 0.084 0.911 1 25290447 GPR126 + 0.12817438 0.086 0.911 2 124302045 PGGT1B - -0.0396 0.086 0.911 2 18179978 CHST1 - -0.0002248 0.087 0.911 2 146440845 CTNNA1 + -0.0432194 0.087 0.911 203

Table A.8 continued Base Chromosome position Gene1 Strand Slope P q 4 8510459 PHF20L1 - -0.0559 0.090 0.911 2 6207874 ZNHIT2 + -0.1988 0.091 0.911 1 25290577 GPR126 + 0.05927968 0.091 0.911 1 25290665 GPR126 + 0.03380698 0.092 0.911 1 178857595 PMAIP1 - 0.2408 0.094 0.911 1 212130845 HIF1A + 0.07133579 0.097 0.911 2 171384 PIDD + -0.0005 0.097 0.911 4 390299 GPT - 3.1132E-05 0.099 0.911 1 178857555 PMAIP1 - -0.1453 0.100 0.911 4 8510558 PHF20L1 - 0.0035 0.101 0.911 4 8510297 PHF20L1 - -0.0400 0.104 0.911 2 169153 PIDD + 0.3762 0.109 0.911 2 124301935 PGGT1B - 0.0005 0.110 0.911 2 44484649 MYOD1 - 0.0023 0.111 0.911 2 48904243 FAR1 - -0.0006854 0.111 0.911 4 8510454 PHF20L1 - -0.0458 0.113 0.911 2 44484125 MYOD1 - -0.0011 0.114 0.911 2 169523 PIDD + 0.0249 0.115 0.911 2 18180222 CHST1 - 0.06938606 0.115 0.911 5 7899158 CARD10 + 0.00033234 0.116 0.911 2 44484106 MYOD1 - -0.0062 0.117 0.911 2 48904144 FAR1 - 0.00046974 0.118 0.911 2 171529 PIDD + 0.0005 0.119 0.911 4 394824 GPT - -0.0007326 0.123 0.911 4 394830 GPT - -0.0009104 0.125 0.911 4 632681 HSF1 - 0.00106193 0.125 0.911 2 146440632 CTNNA1 + 0.0804633 0.126 0.911 1 25289855 GPR126 + -0.000284 0.128 0.911 4 390270 GPT - 3.1975E-05 0.128 0.911 2 171418 PIDD + -0.0004 0.128 0.911 5 8135395 RAC1 + -0.0117 0.129 0.911 2 5538650 FOSL1 + -0.0037162 0.129 0.911 1 212130780 HIF1A + -0.0311096 0.131 0.911 1 178857070 PMAIP1 - 0.0572 0.131 0.911 2 124302079 PGGT1B - 0.0006 0.132 0.911 2 18179614 CHST1 - -0.0003895 0.132 0.911 2 169338 PIDD + 0.2640 0.133 0.911 2 44484682 MYOD1 - -0.1007 0.133 0.911 2 6208056 ZNHIT2 + -0.0004079 0.134 0.911

204

Table A.8 continued Base Chromosome position Gene1 Strand Slope P q 2 48904296 FAR1 - -0.0008604 0.135 0.911 2 6207608 ZNHIT2 + -0.0010 0.135 0.911 4 8510417 PHF20L1 - 0.0109 0.135 0.911 2 6207916 ZNHIT2 + 0.0024 0.136 0.911 2 169368 PIDD + 0.0084 0.136 0.911 2 169365 PIDD + 0.0085 0.136 0.911 4 632793 HSF1 - -0.0006458 0.138 0.911 2 6207466 ZNHIT2 + -0.2599 0.138 0.911 2 6207489 ZNHIT2 + 0.0129 0.138 0.911 2 124334434 PGGT1B - 0.0580 0.139 0.911 2 5538720 FOSL1 + 0.01557989 0.139 0.911 2 44484716 MYOD1 - -0.0226 0.139 0.911 2 124334309 PGGT1B - -0.0080 0.141 0.911 2 124334492 PGGT1B - 0.0618 0.142 0.911 2 124302154 PGGT1B - -0.0682 0.142 0.911 4 8510608 PHF20L1 - 0.0029 0.144 0.911 2 44484679 MYOD1 - -0.3194 0.145 0.911 2 6207381 ZNHIT2 + 0.1878 0.145 0.911 2 44484456 MYOD1 - 0.0009 0.146 0.911 1 212130980 HIF1A + -0.0493202 0.147 0.911 2 44484168 MYOD1 - -0.0221 0.148 0.911 2 6207526 ZNHIT2 + 0.2001 0.149 0.911 2 48903843 FAR1 - -0.0001765 0.149 0.911 4 632827 HSF1 - -0.0871544 0.151 0.911 2 48903749 FAR1 - 0.00100197 0.152 0.911 2 146440886 CTNNA1 + 0.04369636 0.153 0.911 4 632400 HSF1 - -0.0028348 0.153 0.911 2 124301997 PGGT1B - 0.0049 0.154 0.911 1 178857157 PMAIP1 - 0.0731 0.156 0.911 1 212130871 HIF1A + 0.02287064 0.156 0.911 2 124334297 PGGT1B - -0.0710 0.157 0.911 4 390701 GPT - -0.0104651 0.157 0.911 2 5534365 FOSL1 + -0.0058932 0.157 0.911 2 124302117 PGGT1B - 0.0642 0.160 0.911 4 390684 GPT - 0.01037995 0.160 0.911 2 169123 PIDD + 0.0246 0.161 0.911 1 212130764 HIF1A + -0.0143994 0.161 0.911 2 44484323 MYOD1 - 0.1080 0.161 0.911 1 25290498 GPR126 + 0.60406622 0.161 0.911

205

Table A.8 continued Base Chromosome position Gene1 Strand Slope P q 2 171575 PIDD + 0.0003 0.162 0.911 4 8510534 PHF20L1 - 0.0020 0.165 0.911 4 8510611 PHF20L1 - 0.0027 0.166 0.911 4 390261 GPT - 5.7761E-05 0.167 0.911 2 44484191 MYOD1 - -0.0033 0.167 0.911 2 44484518 MYOD1 - 0.1750 0.168 0.911 2 124334346 PGGT1B - -0.0064 0.170 0.911 2 171443 PIDD + -0.0005 0.171 0.911 1 71027180 MANEA + 0.01483967 0.172 0.911 2 48904429 FAR1 - 0.01819858 0.174 0.911 2 146440814 CTNNA1 + 0.00319422 0.175 0.911 2 18179948 CHST1 - -0.0006302 0.175 0.911 1 178857112 PMAIP1 - 0.0016 0.176 0.911 5 7899184 CARD10 + 0.0046921 0.177 0.911 2 48904182 FAR1 - -0.0070101 0.177 0.911 1 71027178 MANEA + -0.0012356 0.178 0.911 2 146440721 CTNNA1 + -0.0499645 0.178 0.911 2 5534233 FOSL1 + 0.02070265 0.179 0.911 2 48904254 FAR1 - 0.00141313 0.180 0.911 2 124334278 PGGT1B - -0.1100 0.180 0.911 4 394820 GPT - -0.0009291 0.182 0.911 2 6207540 ZNHIT2 + 0.0100 0.184 0.911 2 5534493 FOSL1 + -0.010233 0.184 0.911 4 8510443 PHF20L1 - -0.0502 0.185 0.911 2 48903770 FAR1 - 0.02981798 0.185 0.911 2 5534332 FOSL1 + -0.3254123 0.186 0.911 4 394788 GPT - -0.0007049 0.187 0.911 2 124334352 PGGT1B - -0.0023 0.188 0.911 1 212130919 HIF1A + 0.00804243 0.188 0.911 4 8510592 PHF20L1 - 0.0026 0.189 0.911 2 44484653 MYOD1 - 0.0018 0.190 0.911 2 5534172 FOSL1 + -0.3107751 0.191 0.911 2 146440743 CTNNA1 + 0.03930912 0.191 0.911 2 48904416 FAR1 - 0.00278175 0.191 0.911 4 8510595 PHF20L1 - 0.0027 0.192 0.911 4 632464 HSF1 - -0.0005355 0.192 0.911 2 124302030 PGGT1B - -0.0251 0.192 0.911 2 169476 PIDD + 0.0903 0.192 0.911 4 632370 HSF1 - -0.000275 0.193 0.911

206

Table A.8 continued Base Chromosome position Gene1 Strand Slope P q 2 48904248 FAR1 - 0.01124941 0.194 0.911 2 124302008 PGGT1B - -0.0029 0.195 0.911 2 146440770 CTNNA1 + -0.020301 0.197 0.911 4 8510406 PHF20L1 - 0.0115 0.198 0.911 2 124334452 PGGT1B - 0.0140 0.198 0.911 1 212131028 HIF1A + -0.0158355 0.200 0.911 2 6207919 ZNHIT2 + -0.0167 0.201 0.911 2 169200 PIDD + 0.6791 0.203 0.911 4 630746 HSF1 - -0.0001067 0.203 0.911 1 71027164 MANEA + -0.0043239 0.203 0.911 2 146440698 CTNNA1 + -0.0270925 0.204 0.911 1 178857502 PMAIP1 - 0.1480 0.205 0.911 4 632291 HSF1 - 0.00015493 0.205 0.911 4 8510589 PHF20L1 - 0.0023 0.206 0.911 2 44484279 MYOD1 - 0.0014 0.207 0.911 4 8510616 PHF20L1 - 0.0024 0.207 0.911 2 169230 PIDD + 0.0131 0.208 0.911 1 212130708 HIF1A + -0.0002546 0.208 0.911 4 390365 GPT - -4.746E-05 0.209 0.911 1 212130836 HIF1A + 0.02519601 0.209 0.911 2 169224 PIDD + 0.0670 0.210 0.911 1 212130873 HIF1A + -0.0006076 0.210 0.911 1 212130758 HIF1A + -0.01916 0.210 0.911 2 44484259 MYOD1 - 0.0022 0.211 0.911 1 178857539 PMAIP1 - 0.0011 0.212 0.911 2 44484267 MYOD1 - 0.0012 0.212 0.911 2 146440866 CTNNA1 + 0.00054905 0.214 0.911 2 48903825 FAR1 - 0.00047405 0.216 0.911 4 632758 HSF1 - 0.00142254 0.217 0.911 4 632403 HSF1 - -0.0004503 0.217 0.911 1 212131121 HIF1A + -0.0834123 0.217 0.911 2 48904414 FAR1 - -0.009211 0.217 0.911 2 169092 PIDD + 0.0062 0.219 0.911 1 212131112 HIF1A + 0.0678805 0.220 0.911 2 146440881 CTNNA1 + 0.01254999 0.221 0.911 2 146440858 CTNNA1 + -0.0044563 0.222 0.911 2 124301981 PGGT1B - -0.0237 0.224 0.911 2 124302113 PGGT1B - -0.0417 0.224 0.911 1 25290268 GPR126 + -0.1897537 0.225 0.911

207

Table A.8 continued Base Chromosome position Gene1 Strand Slope P q 2 146440585 CTNNA1 + -0.0302613 0.227 0.911 2 124334414 PGGT1B - 0.0132 0.233 0.911 5 7899182 CARD10 + 0.00198151 0.234 0.911 2 18179580 CHST1 - -0.0007748 0.234 0.911 1 212131088 HIF1A + -0.1534316 0.235 0.911 2 169439 PIDD + 0.0912 0.235 0.911 2 171386 PIDD + 0.0003 0.236 0.911 2 18180365 CHST1 - -0.0418302 0.236 0.911 2 6207531 ZNHIT2 + 0.2314 0.236 0.911 2 6208045 ZNHIT2 + 0.0007 0.236 0.911 2 5534446 FOSL1 + -0.0060541 0.236 0.911 1 178857143 PMAIP1 - -0.0011 0.237 0.911 2 5538706 FOSL1 + 0.00866734 0.237 0.911 2 124302024 PGGT1B - 0.0032 0.238 0.911 2 169461 PIDD + 0.1689 0.240 0.911 1 25290540 GPR126 + -0.1865567 0.240 0.911 2 48904481 FAR1 - -0.0427453 0.241 0.911 4 394920 GPT - -0.0008315 0.242 0.911 2 6207879 ZNHIT2 + -0.0859 0.242 0.911 5 8135414 RAC1 + 0.0018 0.243 0.911 2 6208005 ZNHIT2 + 0.0016 0.245 0.911 2 44484173 MYOD1 - -0.0932 0.245 0.911 4 8510613 PHF20L1 - 0.0025 0.246 0.911 2 18179726 CHST1 - -0.0005163 0.248 0.911 2 169499 PIDD + 0.0458 0.256 0.911 2 5534126 FOSL1 + -0.0207182 0.257 0.911 2 169417 PIDD + -0.0116 0.258 0.911 2 44484585 MYOD1 - 0.0259 0.258 0.911 2 48904123 FAR1 - -0.0007071 0.258 0.911 1 178857184 PMAIP1 - 0.0097 0.258 0.911 2 18179991 CHST1 - -0.0001749 0.258 0.911 2 5534285 FOSL1 + 0.00689869 0.259 0.911 2 18179928 CHST1 - 0.00020389 0.260 0.911 1 178857631 PMAIP1 - -0.1462 0.261 0.911 2 44484379 MYOD1 - -0.0019 0.262 0.911 1 25290326 GPR126 + -0.0030274 0.262 0.911 4 8510309 PHF20L1 - -0.0268 0.262 0.911 2 18179563 CHST1 - 0.0003705 0.263 0.911 2 44484700 MYOD1 - 0.2820 0.265 0.911

208

Table A.8 continued Base Chromosome position Gene1 Strand Slope P q 1 25290653 GPR126 + 0.02243928 0.265 0.911 1 25290496 GPR126 + 0.06906009 0.266 0.911 2 169057 PIDD + -0.0127 0.266 0.911 2 6207910 ZNHIT2 + -0.0171 0.267 0.911 2 5534298 FOSL1 + 0.22267347 0.268 0.911 4 8510319 PHF20L1 - 0.0375 0.270 0.911 1 25289745 GPR126 + 4.0466E-05 0.270 0.911 1 71027093 MANEA + 0.01352622 0.271 0.911 1 178857224 PMAIP1 - -0.0963 0.274 0.911 2 124334439 PGGT1B - 0.0075 0.274 0.911 2 48903855 FAR1 - 0.00030464 0.275 0.911 2 18179728 CHST1 - -0.0005482 0.275 0.911 4 8510660 PHF20L1 - 0.0265 0.276 0.911 2 169431 PIDD + -0.0249 0.276 0.911 4 8510576 PHF20L1 - 0.0023 0.277 0.911 1 212130716 HIF1A + -0.0166288 0.279 0.911 2 18180331 CHST1 - -0.0367124 0.280 0.911 2 169139 PIDD + 0.0054 0.280 0.911 5 8135312 RAC1 + 0.0024 0.284 0.911 2 169336 PIDD + 0.0566 0.284 0.911 1 71027119 MANEA + 0.02181874 0.284 0.911 2 5534481 FOSL1 + -0.0040206 0.285 0.911 4 8510549 PHF20L1 - 0.0022 0.286 0.911 2 5534216 FOSL1 + -0.067642 0.286 0.911 4 390708 GPT - -0.0082118 0.286 0.911 4 630648 HSF1 - 0.00011913 0.288 0.911 1 178857628 PMAIP1 - -0.0353 0.289 0.911 1 25290604 GPR126 + -0.0791479 0.290 0.911 4 8510324 PHF20L1 - 0.0002 0.291 0.911 2 124302157 PGGT1B - 0.0329 0.291 0.911 1 25290280 GPR126 + -0.0806289 0.291 0.911 4 8510400 PHF20L1 - 0.0364 0.291 0.911 1 178857296 PMAIP1 - 0.0055 0.292 0.911 4 632817 HSF1 - -0.0061554 0.294 0.911 4 394773 GPT - -0.0005597 0.295 0.911 2 169377 PIDD + 0.0023 0.295 0.911 2 44484352 MYOD1 - -0.0017 0.297 0.911 2 18180354 CHST1 - -0.0463507 0.298 0.911 2 5538656 FOSL1 + -0.002522 0.300 0.911

209

Table A.8 continued Base Chromosome position Gene1 Strand Slope P q 2 6207988 ZNHIT2 + 0.0003 0.301 0.911 1 25290452 GPR126 + 0.06148502 0.302 0.911 4 630821 HSF1 - -0.0001226 0.305 0.911 1 212130760 HIF1A + -0.0116598 0.306 0.911 2 124334466 PGGT1B - 0.0391 0.307 0.911 4 8510518 PHF20L1 - 0.0046 0.308 0.911 4 8510648 PHF20L1 - 0.0381 0.308 0.911 2 146440766 CTNNA1 + 0.04213434 0.309 0.911 2 171586 PIDD + 0.0002 0.310 0.911 2 48904263 FAR1 - 0.00010787 0.311 0.911 2 18179987 CHST1 - 0.0001603 0.313 0.911 2 169345 PIDD + 0.0046 0.313 0.911 2 48904412 FAR1 - -0.0071955 0.316 0.911 2 44484399 MYOD1 - 0.0005 0.319 0.911 2 5534398 FOSL1 + 0.06220345 0.319 0.911 2 18180316 CHST1 - 0.48840573 0.320 0.911 2 124334396 PGGT1B - 0.0602 0.322 0.911 2 169171 PIDD + 0.0068 0.322 0.911 1 178857530 PMAIP1 - -0.0050 0.323 0.911 1 71027231 MANEA + 0.00067863 0.323 0.911 2 48904167 FAR1 - 0.01638301 0.323 0.911 2 169149 PIDD + 0.1665 0.324 0.911 2 6207946 ZNHIT2 + -0.0086 0.324 0.911 4 8510674 PHF20L1 - -0.0350 0.325 0.911 1 25290346 GPR126 + -0.0024368 0.325 0.911 1 178857608 PMAIP1 - -0.1422 0.325 0.911 2 18179681 CHST1 - -0.000577 0.326 0.911 2 18180415 CHST1 - 0.02964289 0.327 0.911 4 8510670 PHF20L1 - -0.0260 0.328 0.911 4 630755 HSF1 - -0.0001147 0.330 0.911 2 6207710 ZNHIT2 + 0.0003 0.332 0.911 2 44484223 MYOD1 - 0.0049 0.335 0.911 2 44484542 MYOD1 - 0.0230 0.335 0.911 4 390328 GPT - 2.3885E-05 0.335 0.911 2 169422 PIDD + 0.0006 0.336 0.911 2 44484461 MYOD1 - 0.0003 0.336 0.911 1 25289785 GPR126 + 5.8045E-05 0.336 0.911 1 25289827 GPR126 + 0.00003289 0.336 0.911 2 5534288 FOSL1 + -0.1301242 0.337 0.911

210

Table A.8 continued Base Chromosome position Gene1 Strand Slope P q 2 18179686 CHST1 - -0.0005657 0.337 0.911 4 632788 HSF1 - -0.0007741 0.340 0.911 4 8510514 PHF20L1 - -0.0247 0.343 0.911 2 18180387 CHST1 - 0.0490936 0.346 0.911 4 8510555 PHF20L1 - -0.0001 0.347 0.911 2 171396 PIDD + -0.0003 0.347 0.911 1 212131024 HIF1A + 0.03911536 0.348 0.911 2 44484568 MYOD1 - 0.0228 0.348 0.911 2 124334366 PGGT1B - -0.0504 0.351 0.911 2 48904450 FAR1 - 0.00047332 0.352 0.911 2 169220 PIDD + 0.0504 0.353 0.911 2 146440835 CTNNA1 + -0.0191204 0.353 0.911 4 8510543 PHF20L1 - -0.0156 0.353 0.911 2 169388 PIDD + 0.0017 0.354 0.911 4 390604 GPT - -0.0144128 0.354 0.911 1 25290609 GPR126 + 0.05305858 0.355 0.911 2 6208042 ZNHIT2 + -0.0147 0.355 0.911 2 48903868 FAR1 - -0.0003305 0.355 0.911 2 44484529 MYOD1 - -0.0011 0.356 0.911 1 212130728 HIF1A + -0.0251759 0.356 0.911 2 5534195 FOSL1 + -0.1032525 0.357 0.911 2 44484431 MYOD1 - 0.0004 0.357 0.911 2 6207953 ZNHIT2 + -0.0030 0.358 0.911 2 6207665 ZNHIT2 + -0.0442 0.359 0.911 4 632799 HSF1 - 0.04058394 0.360 0.911 1 212130752 HIF1A + -0.0060734 0.361 0.911 2 18180302 CHST1 - -0.0440446 0.362 0.911 1 212130790 HIF1A + 0.0069138 0.363 0.911 4 390258 GPT - 3.4005E-05 0.364 0.911 2 6207887 ZNHIT2 + 0.0010 0.364 0.911 2 48904278 FAR1 - 0.0003167 0.365 0.911 2 146440806 CTNNA1 + 0.00013363 0.373 0.930 1 212130754 HIF1A + -0.0171411 0.375 0.930 4 390346 GPT - 1.7501E-05 0.377 0.930 1 212130703 HIF1A + 0.04202122 0.378 0.930 2 48904256 FAR1 - 0.00013836 0.379 0.930 4 8510561 PHF20L1 - 0.0142 0.379 0.930 2 6207424 ZNHIT2 + 0.0028 0.380 0.930 2 48903918 FAR1 - 0.01066115 0.380 0.930

211

Table A.8 continued Base Chromosome position Gene1 Strand Slope P q 1 212131050 HIF1A + 0.03801859 0.386 0.941 1 25290676 GPR126 + 0.04286205 0.387 0.941 4 630759 HSF1 - 0.00011006 0.390 0.946 2 6207929 ZNHIT2 + -0.0024 0.393 0.947 4 8510280 PHF20L1 - 0.0018 0.393 0.947 4 390349 GPT - -2.037E-05 0.393 0.947 1 212130733 HIF1A + -0.073202 0.396 0.951 4 390241 GPT - -3.69E-05 0.400 0.957 1 212131118 HIF1A + 0.07288044 0.401 0.957 2 6207345 ZNHIT2 + -0.0035 0.402 0.957 1 25290355 GPR126 + -0.010238 0.403 0.957 2 146440594 CTNNA1 + -0.0345755 0.403 0.957 2 146440637 CTNNA1 + 0.02212861 0.405 0.959 2 146440897 CTNNA1 + -0.0024458 0.410 0.968 2 5534473 FOSL1 + -0.0090721 0.412 0.968 2 44484221 MYOD1 - -0.0038 0.413 0.968 2 169215 PIDD + 0.0029 0.413 0.968 5 8135310 RAC1 + -0.0017 0.416 0.968 1 212130854 HIF1A + 0.00052259 0.417 0.968 4 390590 GPT - -0.0214638 0.417 0.968 2 18180260 CHST1 - -0.8314224 0.418 0.968 1 71027213 MANEA + -0.0013282 0.418 0.968 4 8510244 PHF20L1 - 0.0152 0.420 0.969 4 8510546 PHF20L1 - 0.0033 0.424 0.969 1 212130770 HIF1A + 0.00087199 0.426 0.969 2 44484738 MYOD1 - 0.0044 0.428 0.969 2 146440639 CTNNA1 + 0.0194234 0.428 0.969 2 5538574 FOSL1 + -0.0023272 0.431 0.969 2 124334441 PGGT1B - -0.0428 0.431 0.969 4 8510347 PHF20L1 - 0.0014 0.431 0.969 2 48904217 FAR1 - 0.00011706 0.432 0.969 1 212130750 HIF1A + -0.0166673 0.433 0.969 2 124302067 PGGT1B - 0.0001 0.433 0.969 2 6207364 ZNHIT2 + 0.0892 0.435 0.969 2 6207639 ZNHIT2 + 0.0006 0.436 0.969 2 44484375 MYOD1 - 0.0005 0.436 0.969 4 390655 GPT - -0.0055908 0.437 0.969 1 25290514 GPR126 + 0.02610792 0.439 0.969 1 25289729 GPR126 + 7.3516E-05 0.439 0.969

212

Table A.8 continued Base Chromosome position Gene1 Strand Slope P q 2 44484136 MYOD1 - -0.0867 0.440 0.969 4 8510362 PHF20L1 - 0.0025 0.441 0.969 2 5538733 FOSL1 + -0.0042716 0.444 0.969 4 8510481 PHF20L1 - -0.0072 0.447 0.969 4 390699 GPT - -0.0076773 0.448 0.969 2 169470 PIDD + 0.0619 0.450 0.969 2 169390 PIDD + 0.1137 0.451 0.969 2 6208052 ZNHIT2 + 0.0010 0.452 0.969 4 394763 GPT - 0.00043513 0.452 0.969 2 18179607 CHST1 - 0.00073793 0.453 0.969 2 5534505 FOSL1 + 0.00235676 0.453 0.969 2 146440575 CTNNA1 + -0.0265104 0.454 0.969 4 390369 GPT - 3.8324E-05 0.456 0.969 2 6207908 ZNHIT2 + -0.0125 0.458 0.969 2 18179943 CHST1 - -0.0001161 0.458 0.969 2 48904199 FAR1 - -0.0004135 0.460 0.969 2 44484245 MYOD1 - 0.0036 0.461 0.969 2 44484710 MYOD1 - 0.0163 0.464 0.969 4 8510410 PHF20L1 - -0.0005 0.465 0.969 2 5534308 FOSL1 + 0.00760698 0.466 0.969 1 71027207 MANEA + -0.0436731 0.467 0.969 1 178857260 PMAIP1 - 0.0012 0.471 0.969 2 146440535 CTNNA1 + -0.0139012 0.473 0.969 1 71027198 MANEA + 0.03983899 0.473 0.969 2 171573 PIDD + -0.0004 0.474 0.969 2 44484512 MYOD1 - 0.3311 0.475 0.969 4 390522 GPT - 0.00492375 0.475 0.969 2 169313 PIDD + 0.0438 0.475 0.969 1 25290278 GPR126 + 0.06250481 0.476 0.969 2 48903816 FAR1 - 0.00122714 0.477 0.969 1 212131075 HIF1A + -0.1398525 0.478 0.969 2 5538585 FOSL1 + 0.00192897 0.478 0.969 2 124302014 PGGT1B - 0.0015 0.478 0.969 2 44484274 MYOD1 - 0.0018 0.479 0.969 4 8510367 PHF20L1 - 0.0020 0.479 0.969 4 630706 HSF1 - 9.6833E-05 0.483 0.969 2 5538673 FOSL1 + 0.00184432 0.484 0.969 2 18179691 CHST1 - -0.0007546 0.485 0.969 2 171551 PIDD + 0.0002 0.486 0.969

213

Table A.8 continued Base Chromosome position Gene1 Strand Slope P q 2 6207394 ZNHIT2 + 0.1130 0.487 0.969 2 6542068 PYGM + 0.0004515 0.488 0.969 1 178857117 PMAIP1 - -0.0440 0.488 0.969 2 44484343 MYOD1 - -0.0001 0.490 0.969 2 48904266 FAR1 - -0.0001438 0.491 0.969 4 390342 GPT - -1.31E-05 0.491 0.969 2 44484496 MYOD1 - 0.0003 0.492 0.969 2 6207504 ZNHIT2 + 0.0011 0.494 0.969 4 632323 HSF1 - -9.24E-05 0.496 0.969 2 5534253 FOSL1 + 0.02926728 0.497 0.969 4 8510306 PHF20L1 - 0.0014 0.498 0.969 2 44484484 MYOD1 - 0.0002 0.499 0.969 2 48903794 FAR1 - -0.0072957 0.499 0.969 2 146440653 CTNNA1 + -0.0013397 0.499 0.969 1 212130746 HIF1A + 0.00583961 0.501 0.969 4 8510428 PHF20L1 - 0.0377 0.502 0.969 4 8510438 PHF20L1 - -0.0004 0.503 0.969 2 169486 PIDD + 0.0063 0.503 0.969 2 44484609 MYOD1 - 0.0021 0.503 0.969 2 169077 PIDD + 0.0111 0.506 0.969 2 6207977 ZNHIT2 + 0.0002 0.506 0.969 2 169121 PIDD + -0.0305 0.506 0.969 2 44484703 MYOD1 - 0.0014 0.507 0.969 4 8510646 PHF20L1 - -0.0002 0.507 0.969 4 390308 GPT - 1.3204E-05 0.510 0.969 2 146440517 CTNNA1 + -0.0002869 0.514 0.969 1 212130718 HIF1A + 0.01570469 0.514 0.969 2 124301972 PGGT1B - -0.0012 0.515 0.969 2 18179975 CHST1 - 0.00013293 0.517 0.969 2 44484109 MYOD1 - 0.0081 0.518 0.969 4 632319 HSF1 - 7.7062E-05 0.518 0.969 2 169183 PIDD + -0.0044 0.519 0.969 4 394757 GPT - -0.0003565 0.520 0.969 1 212131123 HIF1A + 0.05309831 0.523 0.969 2 124302042 PGGT1B - 0.0026 0.526 0.969 2 6207958 ZNHIT2 + -0.0051 0.526 0.969 2 171465 PIDD + 0.0005 0.528 0.969 2 168983 PIDD + -0.0015 0.528 0.969 1 25289718 GPR126 + 4.1261E-05 0.529 0.969

214

Table A.8 continued Base Chromosome position Gene1 Strand Slope P q 5 7899151 CARD10 + -0.0035487 0.529 0.969 2 5538784 FOSL1 + 0.00206329 0.535 0.969 4 630845 HSF1 - 7.0344E-05 0.537 0.969 2 6207621 ZNHIT2 + -0.0271 0.537 0.969 2 6207965 ZNHIT2 + 0.0003 0.538 0.969 1 178857197 PMAIP1 - -0.0015 0.538 0.969 2 44484406 MYOD1 - 0.0000 0.539 0.969 2 124301949 PGGT1B - -0.0209 0.540 0.969 2 124334488 PGGT1B - 0.0443 0.541 0.969 2 171536 PIDD + -0.0002 0.542 0.969 2 44484237 MYOD1 - -0.0027 0.542 0.969 2 6207648 ZNHIT2 + 0.0004 0.543 0.969 2 48904113 FAR1 - 0.00635013 0.545 0.969 2 44484303 MYOD1 - 0.0003 0.546 0.969 2 44484486 MYOD1 - 0.0003 0.546 0.969 2 5534203 FOSL1 + -0.0393945 0.549 0.969 1 71027088 MANEA + 0.00938888 0.550 0.969 4 8510579 PHF20L1 - 0.0011 0.550 0.969 2 124334405 PGGT1B - -0.0630 0.552 0.969 2 18179672 CHST1 - -0.0002007 0.554 0.969 2 171462 PIDD + -0.0004 0.557 0.969 2 6542194 PYGM + 0.00048268 0.557 0.969 1 71027148 MANEA + 0.000954 0.560 0.969 2 169357 PIDD + 0.1331 0.560 0.969 4 8510372 PHF20L1 - 0.0281 0.561 0.969 2 146440625 CTNNA1 + -0.0189538 0.561 0.969 4 390704 GPT - -0.0044225 0.561 0.969 2 18180276 CHST1 - 0.0125773 0.564 0.969 5 8135382 RAC1 + 0.0016 0.564 0.969 4 390379 GPT - -2.618E-05 0.565 0.969 2 171450 PIDD + 0.0002 0.566 0.969 4 390344 GPT - -1.26E-05 0.566 0.969 1 178857466 PMAIP1 - 0.0613 0.569 0.969 2 146440823 CTNNA1 + -0.0098398 0.570 0.969 2 5538689 FOSL1 + 0.0029349 0.571 0.969 2 146440599 CTNNA1 + 0.00193562 0.576 0.969 4 8510383 PHF20L1 - 0.0010 0.577 0.969 1 25290292 GPR126 + -0.0410048 0.577 0.969 2 18179661 CHST1 - -0.0001868 0.579 0.969

215

Table A.8 continued Base Chromosome position Gene1 Strand Slope P q 2 146440556 CTNNA1 + -0.0010971 0.579 0.969 2 48904172 FAR1 - 0.00749446 0.579 0.969 2 6207992 ZNHIT2 + 0.0004 0.581 0.969 2 44484625 MYOD1 - -0.0007 0.582 0.969 2 18179880 CHST1 - 0.00048235 0.582 0.969 2 124302070 PGGT1B - 0.0002 0.582 0.969 2 6207713 ZNHIT2 + -0.0015 0.583 0.969 4 630616 HSF1 - 5.2888E-05 0.584 0.969 4 390636 GPT - 0.00436375 0.584 0.969 4 630781 HSF1 - 7.3386E-05 0.584 0.969 2 124334419 PGGT1B - -0.0538 0.584 0.969 2 48903744 FAR1 - -0.0009309 0.585 0.969 1 178857589 PMAIP1 - -0.0729 0.585 0.969 4 630749 HSF1 - -4.771E-05 0.585 0.969 2 169454 PIDD + 0.2139 0.586 0.969 2 5534488 FOSL1 + -0.0012973 0.586 0.969 1 212130687 HIF1A + -0.0037509 0.586 0.969 2 5534496 FOSL1 + 0.00189742 0.587 0.969 4 630675 HSF1 - 7.4594E-05 0.587 0.969 2 6207632 ZNHIT2 + -0.0002 0.589 0.969 2 146440596 CTNNA1 + 0.0221036 0.590 0.969 2 169488 PIDD + 0.0029 0.590 0.969 4 630696 HSF1 - -5.984E-05 0.591 0.969 2 5534438 FOSL1 + -0.001425 0.592 0.969 1 25290276 GPR126 + -0.004547 0.594 0.969 2 44484058 MYOD1 - -0.0009 0.594 0.969 1 178857621 PMAIP1 - -0.0183 0.595 0.969 4 390275 GPT - 1.3626E-05 0.596 0.969 2 44484705 MYOD1 - 0.0010 0.598 0.969 4 8510261 PHF20L1 - 0.0148 0.599 0.969 2 6207683 ZNHIT2 + 0.0007 0.599 0.969 5 7899163 CARD10 + 0.00247967 0.600 0.969 2 169228 PIDD + 0.0178 0.601 0.969 2 6207470 ZNHIT2 + -0.0016 0.602 0.969 4 632377 HSF1 - 9.9365E-05 0.603 0.969 1 25289820 GPR126 + 2.7085E-05 0.603 0.969 2 44484330 MYOD1 - -0.0004 0.603 0.969 4 630727 HSF1 - -6.08E-05 0.604 0.969 1 25290303 GPR126 + 0.00123706 0.605 0.969

216

Table A.8 continued Base Chromosome position Gene1 Strand Slope P q 4 390696 GPT - 0.0036061 0.606 0.969 1 25289726 GPR126 + 4.0355E-05 0.608 0.969 4 632342 HSF1 - -4.865E-05 0.608 0.969 2 169323 PIDD + 0.0015 0.612 0.969 2 48903805 FAR1 - -0.0001148 0.617 0.969 1 178857125 PMAIP1 - 0.0348 0.617 0.969 4 630632 HSF1 - -7.614E-05 0.617 0.969 2 169479 PIDD + 0.0014 0.620 0.969 2 5534219 FOSL1 + -0.0628957 0.620 0.969 4 390330 GPT - 9.94E-06 0.621 0.969 2 169347 PIDD + 0.0015 0.621 0.969 1 25290630 GPR126 + -0.0366408 0.623 0.969 2 6207706 ZNHIT2 + -0.0002 0.623 0.969 1 212131003 HIF1A + -0.0457934 0.623 0.969 2 146440644 CTNNA1 + 0.01925441 0.623 0.969 2 124302136 PGGT1B - -0.0022 0.624 0.969 2 5538757 FOSL1 + 0.00335506 0.624 0.969 2 5534248 FOSL1 + -0.0058438 0.624 0.969 2 6207544 ZNHIT2 + -0.0027 0.626 0.969 2 48904271 FAR1 - -0.0034914 0.626 0.969 1 71027103 MANEA + 0.00051124 0.627 0.969 2 5534138 FOSL1 + -0.0039192 0.630 0.969 2 5534374 FOSL1 + 0.00441367 0.631 0.969 2 6207922 ZNHIT2 + 0.0005 0.632 0.969 2 18179708 CHST1 - -0.0001818 0.632 0.969 4 390248 GPT - 2.4618E-05 0.634 0.969 1 178857560 PMAIP1 - 0.0113 0.634 0.969 2 169515 PIDD + 0.0026 0.634 0.969 2 171597 PIDD + 0.0001 0.635 0.969 2 169334 PIDD + 0.0367 0.636 0.969 4 8510574 PHF20L1 - -0.0001 0.638 0.969 4 8510622 PHF20L1 - 0.0037 0.638 0.969 2 169443 PIDD + 0.0755 0.641 0.969 2 6542064 PYGM + 0.00033219 0.642 0.969 1 71027220 MANEA + 0.01316087 0.642 0.969 4 390228 GPT - 2.5642E-05 0.643 0.969 4 390236 GPT - -1.95E-05 0.643 0.969 2 48904301 FAR1 - -0.0003116 0.643 0.969 1 178857605 PMAIP1 - 0.0494 0.644 0.969

217

Table A.8 continued Base Chromosome position Gene1 Strand Slope P q 4 8510602 PHF20L1 - -0.0005 0.644 0.969 2 5538641 FOSL1 + 0.00226486 0.644 0.969 2 18179922 CHST1 - -0.0001021 0.646 0.969 2 5534485 FOSL1 + -0.0599007 0.648 0.969 4 8510246 PHF20L1 - -0.0202 0.648 0.969 1 71027130 MANEA + 0.00153862 0.649 0.969 2 5538670 FOSL1 + 0.00160096 0.649 0.969 2 171538 PIDD + 0.0001 0.650 0.969 1 178857281 PMAIP1 - 0.0297 0.650 0.969 1 178857159 PMAIP1 - 0.0360 0.651 0.969 2 48904380 FAR1 - 0.00017354 0.653 0.969 4 390362 GPT - 2.2071E-05 0.654 0.969 4 390287 GPT - -1.002E-05 0.655 0.969 2 169397 PIDD + -0.0360 0.656 0.969 2 5534395 FOSL1 + 0.00450138 0.658 0.969 1 178857302 PMAIP1 - 0.0211 0.658 0.969 2 124334402 PGGT1B - 0.0131 0.659 0.969 2 169082 PIDD + 0.0008 0.659 0.969 4 390238 GPT - -2.39E-05 0.659 0.969 2 44484574 MYOD1 - 0.0137 0.659 0.969 4 632309 HSF1 - 0.00014635 0.661 0.969 2 44484123 MYOD1 - -0.0005 0.662 0.969 2 146440874 CTNNA1 + -0.0046326 0.663 0.969 4 390337 GPT - -7.07E-06 0.663 0.969 2 6208040 ZNHIT2 + -0.0001 0.663 0.969 2 169395 PIDD + 0.0012 0.664 0.969 1 71027134 MANEA + 0.01147116 0.664 0.969 2 44484482 MYOD1 - 0.0001 0.666 0.969 2 48903841 FAR1 - 0.00013192 0.668 0.969 4 390580 GPT - -0.0073765 0.669 0.969 2 18179534 CHST1 - 0.00024968 0.669 0.969 1 178857642 PMAIP1 - -0.0932 0.676 0.969 2 44484211 MYOD1 - 0.0011 0.679 0.969 1 25290351 GPR126 + 0.00125395 0.680 0.969 1 178857434 PMAIP1 - -0.0009 0.681 0.969 2 169401 PIDD + 0.0038 0.681 0.969 1 212130744 HIF1A + 0.01826772 0.683 0.969 2 5534467 FOSL1 + 0.00166282 0.684 0.969 4 390562 GPT - -0.0043002 0.685 0.969

218

Table A.8 continued Base Chromosome position Gene1 Strand Slope P q 2 6207404 ZNHIT2 + -0.0753 0.686 0.969 2 48904298 FAR1 - 0.00632964 0.687 0.969 4 8510467 PHF20L1 - -0.0015 0.689 0.969 2 124302027 PGGT1B - -0.0002 0.689 0.969 2 44484553 MYOD1 - -0.1409 0.693 0.969 2 44484745 MYOD1 - 0.0031 0.693 0.969 4 390310 GPT - -7.77E-06 0.693 0.969 4 8510300 PHF20L1 - -0.0002 0.693 0.969 4 390246 GPT - -3.15E-05 0.693 0.969 4 390316 GPT - 7.30E-06 0.694 0.969 1 178857067 PMAIP1 - -0.0001 0.695 0.969 2 48904402 FAR1 - 0.00030177 0.695 0.969 2 124334318 PGGT1B - 0.0154 0.696 0.969 2 124301923 PGGT1B - 0.0074 0.697 0.969 4 390306 GPT - 9.04E-06 0.700 0.969 2 124301991 PGGT1B - -0.0003 0.700 0.969 2 171341 PIDD + 0.0002 0.701 0.969 2 6207446 ZNHIT2 + -0.0810 0.701 0.969 2 18180294 CHST1 - -0.0180621 0.703 0.969 1 178857520 PMAIP1 - -0.0455 0.703 0.969 4 632336 HSF1 - -9.365E-05 0.704 0.969 1 71027145 MANEA + -0.0028745 0.705 0.969 2 44484235 MYOD1 - -0.0013 0.705 0.969 1 25289753 GPR126 + 1.3996E-05 0.706 0.969 2 124334263 PGGT1B - -0.0342 0.706 0.969 2 48904161 FAR1 - 0.00540611 0.710 0.969 4 390377 GPT - -1.55E-05 0.711 0.969 4 390356 GPT - -1.47E-05 0.713 0.969 4 8510639 PHF20L1 - -0.0120 0.713 0.969 4 8510392 PHF20L1 - -0.0171 0.716 0.969 2 18180010 CHST1 - 0.00018023 0.717 0.969 2 171473 PIDD + 0.0004 0.718 0.969 2 6207484 ZNHIT2 + 0.0526 0.719 0.969 1 178857427 PMAIP1 - 0.0439 0.720 0.969 2 124302089 PGGT1B - -0.0008 0.721 0.969 2 169134 PIDD + 0.1005 0.725 0.969 2 44484148 MYOD1 - 0.0008 0.725 0.969 2 6542250 PYGM + 0.00027892 0.725 0.969 2 5534163 FOSL1 + -0.0031258 0.726 0.969 4 394725 GPT - 0.00049516 0.728 0.969 219

Table A.8 continued Base Chromosome position Gene1 Strand Slope P q 1 212131041 HIF1A + -0.036587 0.730 0.969 1 212130726 HIF1A + 0.0086727 0.731 0.969 2 124301955 PGGT1B - 0.0082 0.731 0.969 4 632283 HSF1 - -7.374E-05 0.732 0.969 4 630810 HSF1 - 6.409E-05 0.734 0.969 2 6542112 PYGM + -0.0002345 0.735 0.969 2 6542246 PYGM + 0.00026708 0.736 0.969 2 44484036 MYOD1 - 0.0008 0.738 0.969 2 146440564 CTNNA1 + -0.0014423 0.738 0.969 2 48903920 FAR1 - -0.0046209 0.739 0.969 2 6207657 ZNHIT2 + -0.0003 0.740 0.969 2 169449 PIDD + -0.0029 0.740 0.969 2 5538638 FOSL1 + -0.0019426 0.741 0.969 2 169235 PIDD + -0.0047 0.741 0.969 2 6542207 PYGM + 0.00025534 0.742 0.969 2 44484193 MYOD1 - 0.0003 0.742 0.969 2 48904437 FAR1 - -0.0007654 0.743 0.969 4 632634 HSF1 - -8.926E-05 0.744 0.969 1 178857208 PMAIP1 - -0.0282 0.746 0.969 2 6207971 ZNHIT2 + 0.0002 0.746 0.969 1 25290617 GPR126 + 0.08307219 0.747 0.969 2 18179892 CHST1 - -9.3E-05 0.747 0.969 1 25290613 GPR126 + -0.014216 0.749 0.969 1 178857518 PMAIP1 - -0.0362 0.750 0.969 4 8510665 PHF20L1 - -0.0114 0.750 0.969 2 5534454 FOSL1 + 0.00300582 0.750 0.969 4 8510352 PHF20L1 - -0.0131 0.751 0.969 4 390575 GPT - -0.0052543 0.751 0.969 2 48903901 FAR1 - 0.00012573 0.752 0.969 2 5534147 FOSL1 + -0.0013564 0.755 0.969 2 169458 PIDD + 0.0517 0.756 0.969 2 124334362 PGGT1B - 0.0002 0.761 0.969 2 18179589 CHST1 - -0.0001545 0.761 0.969 2 169341 PIDD + 0.0276 0.762 0.969 2 124334383 PGGT1B - 0.0014 0.764 0.969 4 390731 GPT - -0.0029777 0.764 0.969 1 212131127 HIF1A + -0.0725511 0.766 0.969 1 212130722 HIF1A + 0.00813342 0.766 0.969 2 124334339 PGGT1B - -0.0017 0.768 0.969 1 25290405 GPR126 + 0.00139602 0.768 0.969 220

Table A.8 continued Base Chromosome position Gene1 Strand Slope P q 1 71027132 MANEA + 0.00894985 0.768 0.969 2 146440612 CTNNA1 + -0.0008129 0.768 0.969 4 390626 GPT - -0.0019354 0.769 0.969 2 44484311 MYOD1 - 0.0002 0.770 0.969 2 48903807 FAR1 - 0.00015985 0.771 0.969 1 178857490 PMAIP1 - 0.0065 0.772 0.969 4 630694 HSF1 - -0.0001444 0.772 0.969 2 124334459 PGGT1B - -0.0230 0.774 0.969 2 44484050 MYOD1 - 0.0005 0.774 0.969 2 48903904 FAR1 - 0.00038926 0.777 0.969 2 124334324 PGGT1B - 0.0074 0.777 0.969 4 394704 GPT - 0.00043347 0.779 0.969 2 48904391 FAR1 - 0.00051559 0.780 0.969 2 124334377 PGGT1B - -0.0071 0.781 0.969 2 124301960 PGGT1B - 0.0001 0.781 0.969 2 124334468 PGGT1B - 0.0074 0.782 0.969 1 178857481 PMAIP1 - -0.0288 0.783 0.969 2 6542146 PYGM + 0.00021029 0.784 0.969 2 6207950 ZNHIT2 + 0.0002 0.787 0.969 2 124302075 PGGT1B - -0.0002 0.787 0.969 2 146440679 CTNNA1 + 0.0048113 0.789 0.969 5 7899144 CARD10 + -0.0062563 0.789 0.969 2 5538799 FOSL1 + 0.00442379 0.790 0.969 2 44484203 MYOD1 - 0.0004 0.791 0.969 4 394739 GPT - 0.00016358 0.791 0.969 2 146440714 CTNNA1 + 0.00588154 0.791 0.969 2 6207995 ZNHIT2 + -0.0005 0.792 0.969 1 178857572 PMAIP1 - -0.0163 0.792 0.969 4 8510473 PHF20L1 - 0.0172 0.793 0.969 1 178857484 PMAIP1 - 0.0312 0.793 0.969 4 8510632 PHF20L1 - -0.0041 0.793 0.969 1 25289764 GPR126 + -6.58E-06 0.796 0.969 2 169363 PIDD + 0.0674 0.797 0.969 5 7899137 CARD10 + 0.00046703 0.799 0.969 2 6207410 ZNHIT2 + -0.0085 0.799 0.969 4 390585 GPT - -0.0065094 0.801 0.969 2 146440540 CTNNA1 + -0.0006209 0.802 0.969 2 18179937 CHST1 - -4.174E-05 0.803 0.969 2 169403 PIDD + 0.0233 0.803 0.969

221

Table A.8 continued Base Chromosome position Gene1 Strand Slope P q 2 6207703 ZNHIT2 + 0.0077 0.803 0.969 2 6542242 PYGM + -0.0002061 0.805 0.969 2 6207659 ZNHIT2 + -0.0066 0.805 0.969 2 169202 PIDD + 0.0632 0.807 0.969 5 7899180 CARD10 + 7.7395E-05 0.808 0.969 2 44484617 MYOD1 - -0.0003 0.809 0.969 2 6207421 ZNHIT2 + 0.0489 0.810 0.969 4 8510606 PHF20L1 - 0.0009 0.810 0.969 2 5534153 FOSL1 + 0.05105836 0.811 0.969 1 212130701 HIF1A + 0.00786567 0.812 0.969 2 171556 PIDD + 0.0001 0.812 0.969 2 18179920 CHST1 - -4.899E-05 0.813 0.969 2 124334370 PGGT1B - 0.0087 0.813 0.969 5 7899204 CARD10 + -0.0001956 0.814 0.969 1 25290546 GPR126 + -0.0659677 0.815 0.969 1 212131100 HIF1A + -0.0450113 0.816 0.969 1 25290526 GPR126 + -0.0350096 0.817 0.969 2 48904433 FAR1 - -0.0058499 0.817 0.969 4 390515 GPT - -0.0018763 0.818 0.969 2 124302165 PGGT1B - 0.0084 0.818 0.969 2 6542239 PYGM + 0.00018485 0.819 0.969 2 5534367 FOSL1 + 0.00056843 0.820 0.969 1 25289720 GPR126 + 1.3278E-05 0.820 0.969 1 178857602 PMAIP1 - -0.0014 0.822 0.969 2 6207332 ZNHIT2 + 0.0868 0.822 0.969 4 395003 GPT - -0.000532 0.824 0.969 2 48904455 FAR1 - -0.0198216 0.825 0.969 2 44484601 MYOD1 - -0.0029 0.826 0.969 4 390716 GPT - -0.0015651 0.826 0.969 1 178857271 PMAIP1 - 0.0033 0.828 0.969 1 212130819 HIF1A + 0.00022969 0.829 0.969 2 18180283 CHST1 - -0.006452 0.829 0.969 2 18180227 CHST1 - 0.01243842 0.830 0.969 4 390231 GPT - 8.00E-06 0.830 0.969 2 6207442 ZNHIT2 + -0.0645 0.831 0.969 4 394785 GPT - 6.6385E-05 0.833 0.969 2 6542040 PYGM + 0.00015505 0.833 0.969 1 212130879 HIF1A + 0.00053862 0.834 0.969 2 44484372 MYOD1 - 0.0000 0.835 0.969

222

Table A.8 continued Base Chromosome position Gene1 Strand Slope P q 2 5534273 FOSL1 + -0.0520255 0.835 0.969 2 146440621 CTNNA1 + -0.0053107 0.836 0.969 4 394775 GPT - 0.00021618 0.837 0.969 4 8510644 PHF20L1 - 0.0002 0.838 0.969 2 18179735 CHST1 - -5.544E-05 0.839 0.969 4 390244 GPT - -9.32E-06 0.840 0.969 2 146440650 CTNNA1 + -0.0047092 0.840 0.969 2 6207354 ZNHIT2 + -0.0061 0.844 0.969 1 178857582 PMAIP1 - 0.0457 0.845 0.969 1 212130857 HIF1A + -0.0077681 0.848 0.969 1 25290273 GPR126 + 0.00069513 0.848 0.969 2 169420 PIDD + 0.0018 0.849 0.969 2 6208048 ZNHIT2 + 0.0001 0.850 0.969 2 6542227 PYGM + 0.00015419 0.850 0.969 2 18180013 CHST1 - -9.053E-05 0.851 0.969 2 169000 PIDD + -0.0021 0.851 0.969 2 124334432 PGGT1B - 0.0067 0.851 0.969 2 44484475 MYOD1 - 0.0000 0.852 0.969 2 5538730 FOSL1 + 0.00126545 0.852 0.969 2 6207986 ZNHIT2 + 0.0000 0.853 0.969 2 6542104 PYGM + 0.00015329 0.853 0.969 2 18179703 CHST1 - -9.708E-05 0.853 0.969 1 25290286 GPR126 + -0.0004278 0.854 0.969 2 6207599 ZNHIT2 + -0.0001 0.854 0.969 1 178857562 PMAIP1 - 0.0032 0.854 0.969 1 178857247 PMAIP1 - -0.0021 0.858 0.971 2 6207340 ZNHIT2 + -0.0279 0.858 0.971 1 178857436 PMAIP1 - -0.0076 0.859 0.971 4 390642 GPT - 0.00149604 0.861 0.972 4 8510479 PHF20L1 - -0.0014 0.864 0.973 2 44484524 MYOD1 - -0.0005 0.866 0.973 4 390386 GPT - -7.26E-06 0.867 0.973 2 44484182 MYOD1 - 0.0017 0.867 0.973 1 178857424 PMAIP1 - 0.0174 0.867 0.973 1 178857108 PMAIP1 - 0.0004 0.871 0.974 1 212130695 HIF1A + -0.0017044 0.875 0.974 2 146440851 CTNNA1 + -0.000115 0.879 0.974 2 48904314 FAR1 - -0.0040224 0.880 0.974 2 18179626 CHST1 - 7.2212E-05 0.881 0.974

223

Table A.8 continued Base Chromosome position Gene1 Strand Slope P q 4 632417 HSF1 - 5.2838E-05 0.881 0.974 1 178857611 PMAIP1 - 0.0192 0.882 0.974 4 390615 GPT - 0.00251917 0.883 0.974 1 178857616 PMAIP1 - -0.0101 0.883 0.974 1 178857618 PMAIP1 - 0.0170 0.885 0.974 2 169483 PIDD + -0.0237 0.885 0.974 2 146440819 CTNNA1 + 9.4431E-05 0.886 0.974 5 7899247 CARD10 + -1.84E-05 0.887 0.974 1 178857199 PMAIP1 - -0.0001 0.887 0.974 5 8135443 RAC1 + 0.0462 0.887 0.974 2 18179693 CHST1 - -0.0001275 0.888 0.974 2 44484423 MYOD1 - 0.0000 0.888 0.974 1 178857088 PMAIP1 - -0.0001 0.888 0.974 2 48904230 FAR1 - -3.276E-05 0.890 0.974 2 169406 PIDD + 0.0017 0.894 0.974 1 25290532 GPR126 + 0.04193612 0.895 0.974 1 212130667 HIF1A + 0.00912647 0.895 0.974 2 48904439 FAR1 - -0.0020817 0.897 0.974 4 394718 GPT - -0.0001145 0.897 0.974 2 48904339 FAR1 - -9.326E-05 0.897 0.974 4 394959 GPT - -0.00021 0.897 0.974 2 146440750 CTNNA1 + 4.592E-05 0.898 0.974 1 212131115 HIF1A + 0.01631477 0.899 0.974 1 71027216 MANEA + 0.0001094 0.900 0.974 2 146440705 CTNNA1 + -8.048E-05 0.902 0.974 4 394965 GPT - 0.00018568 0.903 0.974 2 18180210 CHST1 - 0.00483911 0.903 0.974 2 6207642 ZNHIT2 + 0.0002 0.904 0.974 4 8510582 PHF20L1 - 0.0000 0.905 0.974 1 25289743 GPR126 + 2.39E-06 0.906 0.974 2 6207597 ZNHIT2 + 0.0002 0.907 0.974 4 394896 GPT - 8.0904E-05 0.907 0.974 2 171610 PIDD + 0.0001 0.908 0.974 2 18179565 CHST1 - -4.042E-05 0.909 0.974 2 146440841 CTNNA1 + 2.5923E-05 0.909 0.974 2 44484381 MYOD1 - 0.0025 0.910 0.975 2 146440787 CTNNA1 + -3.344E-05 0.912 0.975 2 171369 PIDD + 0.0000 0.917 0.979 1 25289794 GPR126 + -6.52E-06 0.918 0.979

224

Table A.8 continued Base Chromosome position Gene1 Strand Slope P q 5 8135427 RAC1 + -0.0001 0.918 0.979 2 18179529 CHST1 - 2.8325E-05 0.919 0.979 2 5534460 FOSL1 + 0.01693812 0.923 0.980 2 5538652 FOSL1 + -0.0002604 0.924 0.980 1 178857091 PMAIP1 - -0.0001 0.924 0.980 4 8510445 PHF20L1 - -0.0020 0.925 0.980 2 146440582 CTNNA1 + -0.0020443 0.926 0.980 2 124334313 PGGT1B - -0.0053 0.927 0.980 2 5538622 FOSL1 + 0.00097878 0.927 0.980 4 394927 GPT - 8.2037E-05 0.928 0.980 2 124301927 PGGT1B - 0.0004 0.929 0.980 2 6542167 PYGM + -6.909E-05 0.931 0.980 4 8510303 PHF20L1 - -0.0001 0.931 0.980 1 212130942 HIF1A + -0.0017606 0.933 0.980 5 7899149 CARD10 + 0.00057556 0.933 0.980 2 6542049 PYGM + -5.67E-05 0.936 0.981 1 212131055 HIF1A + 0.01723244 0.936 0.981 2 146440701 CTNNA1 + -0.0036569 0.937 0.981 2 44484566 MYOD1 - 0.0004 0.941 0.982 2 48903873 FAR1 - 3.6478E-05 0.941 0.982 2 169310 PIDD + -0.0012 0.942 0.982 2 6542089 PYGM + 4.6393E-05 0.945 0.982 4 632786 HSF1 - 0.00045541 0.946 0.982 2 6208038 ZNHIT2 + 0.0000 0.947 0.982 4 390729 GPT - 0.00041691 0.954 0.982 2 146440768 CTNNA1 + -5.489E-05 0.954 0.982 4 632649 HSF1 - -0.0006338 0.954 0.982 4 394742 GPT - 2.7037E-05 0.956 0.982 1 178857418 PMAIP1 - 0.0003 0.956 0.982 4 394745 GPT - 0.00014975 0.956 0.982 2 18180373 CHST1 - -0.0007899 0.956 0.982 4 8510483 PHF20L1 - 0.0019 0.956 0.982 1 25290308 GPR126 + 0.00100195 0.959 0.982 2 146440869 CTNNA1 + -0.0009804 0.960 0.982 1 25290399 GPR126 + 0.00551628 0.961 0.982 2 48904170 FAR1 - -1.45E-05 0.962 0.982 1 178857510 PMAIP1 - -0.0043 0.963 0.982 2 18180006 CHST1 - 1.4307E-05 0.964 0.982 2 124302142 PGGT1B - -0.0023 0.966 0.982

225

Table A.8 continued Base Chromosome position Gene1 Strand Slope P q 2 169386 PIDD + -0.0109 0.967 0.982 4 632662 HSF1 - 4.8929E-05 0.967 0.982 1 178857441 PMAIP1 - -0.0006 0.967 0.982 2 6542043 PYGM + 2.6122E-05 0.968 0.982 2 48904225 FAR1 - 0.00062622 0.971 0.982 4 390556 GPT - 0.00031854 0.972 0.982 5 7899243 CARD10 + 1.4183E-05 0.975 0.982 4 630614 HSF1 - 4.01E-06 0.975 0.982 2 44484409 MYOD1 - 0.0000 0.976 0.982 2 44484402 MYOD1 - 0.0000 0.977 0.982 2 6207356 ZNHIT2 + 0.0005 0.978 0.982 2 5538635 FOSL1 + -7.804E-05 0.979 0.982 2 44484369 MYOD1 - 0.0000 0.981 0.982 4 394936 GPT - 2.3398E-05 0.982 0.982 2 124301969 PGGT1B - 0.0000 0.983 0.982 2 169048 PIDD + 0.0000 0.983 0.982 1 178857138 PMAIP1 - 0.0000 0.985 0.982 4 632738 HSF1 - -8.69E-06 0.986 0.982 2 18180400 CHST1 - 0.00061569 0.986 0.982 2 44484632 MYOD1 - -0.0011 0.987 0.982 1 178857574 PMAIP1 - 0.0020 0.989 0.982 2 171602 PIDD + 0.0000 0.989 0.982 4 390524 GPT - 0.00010171 0.991 0.982 2 6207603 ZNHIT2 + 0.0000 0.991 0.982 2 124302053 PGGT1B - 0.0000 0.992 0.982 5 7899200 CARD10 + 5.18E-06 0.992 0.982 4 632440 HSF1 - -2.56E-06 0.992 0.982 2 18179950 CHST1 - 2.01E-06 0.992 0.982 4 632764 HSF1 - -1.45E-05 0.992 0.982 1 178857579 PMAIP1 - -0.0009 0.993 0.982 2 18179570 CHST1 - -4.39E-06 0.995 0.982 4 394932 GPT - 3.58E-06 0.997 0.982 2 124334273 PGGT1B - -0.0002 0.997 0.982 2 146440657 CTNNA1 + 3.158E-05 0.998 0.982 2 6207677 ZNHIT2 + 0.0000 0.998 0.982 4 632385 HSF1 - 2.89E-07 0.998 0.982

226

Table A.8 continued Base Chromosome position Gene1 Strand Slope P q 4 8510356 PHF20L1 - 0.0000 1.000 0.982 2 124334436 PGGT1B - 0.0000 1.000 0.982 1Caspase recruitment domain family, member 10 (CARD10); Carbohydrate (keratan sulfate Gal-6) sulfotransferase 1 (CHST1); Catenin (cadherin-associated protein), alpha 1, 102kDa (CTNNA1); Fatty acyl CoA reductase 1 (FAR1); FOS-like antigen 1 (FOSL1); G protein- coupled receptor 126 (GPR126); Glutamic-pyruvate transaminase (alanine aminotransferase) (GPT); Hypoxia-inducible factor 1, alpha subunit (basic helix-loop-helix transcription factor) (HIF1A); Heat shock transcription factor 1 (HSF1); Mannosidase, endo-alpha (MANEA); Myogenic differentiation 1 (MYOD1); Protein geranylgeranyltransferase type I, beta subunit (PGGT1B); PHD finger protein 20-like 1 (PHF20L1); p53-induced death domain protein (PIDD); Phorbol-12-myristate-13-acetate-induced protein 1 (PMAIP1); Phosphorylase, glycogen; muscle (McArdle syndrome, glycogen storage disease type V) (PYGM); Ras-related C3 botulinum toxin substrate 1 (rho family, small GTP binding protein Rac1) (RAC1); Zinc finger, HIT type 2 (ZNHIT2)

227

Table A.9. Effects of gestational treatment, sex, and their interactions on protein abundance in muscle from the western blot analysis. Only offspring exposed to thermoneutral conditions postnatally were evaluated Gest1 Sex2 P Protein3 TNTN TNHS HSTN HSHS SEM F M SEM Gest Sex Gest*Sex HIF1A 1.39 1.39 2.69 1.92 0.32 2.15 1.55 0.23 0.08 0.14 0.15 HSF1 0.64 1.06 1.23 1.40 0.24 1.08 1.08 0.16 0.08 0.95 0.74 HSPA1A 1.07 0.96 1.71 1.20 0.43 1.30 1.17 0.27 0.70 0.68 0.23 MEF2A 1.03 2.25 3.89 2.27 0.60 2.84 1.88 0.39 0.06 0.12 0.16 MYOD1 2.80 3.01 3.99 2.06 1.56 3.64 2.29 0.88 0.86 0.16 0.11 1Gestational treatments: TNTN = thermoneutral during the entire gestation; TNHS = thermoneutral during first half of gestation and heat stress during second half of gestation; HSTN = heat stress during first half of gestation and thermoneutral during second half of gestation; HSHS = heat stress during the entire gestation 2 Sex of the pig: F = female; M = male 3Hypoxia-inducible factor 1, alpha subunit (basic helix-loop-helix transcription factor) (HIF1A); Heat shock transcription factor 1 (HSF1); Heat shock 70 kDa protein 1A (HSPA1A); Myocyte enhancer factor 2A (MEF2A); Myogenic differentiation 1 (MYOD1)

228

Table A.10. Effects of thermal treatment during the first half of gestation, sex, and their interactions on protein abundance in muscle from the western blot analysis. Only offspring exposed to thermoneutral conditions postnatally were evaluated Gest1 Sex2 P Protein3 TN First Half HS First Half SEM F M SEM Gest Sex Gest*Sex HIF1A 1.40 2.19 0.25 2.09 1.50 0.25 0.03 0.10 0.03 HSF1 0.77 1.38 0.14 1.08 1.07 0.11 0.01 0.93 0.35 HSPA1A 1.26 1.36 0.29 1.31 1.31 0.28 0.78 1.00 0.50 MEF2A 1.48 2.84 0.48 2.63 1.69 0.42 0.08 0.10 0.09 MYOD1 3.55 2.73 0.86 3.46 2.82 0.79 0.52 0.54 0.53 1Gestational treatments: TN First Half = thermoneutral conditions during the first half of gestation; HS First Half = heat stress conditions during the first half of gestation 2 Sex of the pig: F = female; M = male 3Hypoxia-inducible factor 1, alpha subunit (basic helix-loop-helix transcription factor) (HIF1A); Heat shock transcription factor 1 (HSF1); Heat shock 70 kDa protein 1A (HSPA1A); Myocyte enhancer factor 2A (MEF2A); Myogenic differentiation 1 (MYOD1)

229

Table A.11. Effects of postnatal treatment, sex, and their interactions on protein abundance in muscle from the western blot analysis. Only offspring exposed to thermoneutral conditions throughout gestation were evaluated Post1 Sex2 P Protein3 TN HS SEM F M SEM Post Sex Post*Sex HIF1A 1.41 1.69 0.29 1.16 1.95 0.29 0.45 0.11 0.27 HSF1 0.63 1.16 0.14 0.89 0.90 0.14 0.01 0.89 0.41 HSPA1A 1.21 1.34 0.33 1.10 1.45 0.33 0.79 0.52 0.40 MEF2A 1.10 2.32 0.53 1.87 1.55 0.53 0.15 0.64 0.45 MYOD1 3.33 2.28 0.86 2.53 3.08 0.86 0.45 0.68 0.42 1Postnatal treatments: TN = thermoneutral conditions during 5-week postnatal thermal challenge; HS = heat stress conditions during 5-week postnatal thermal challenge 2 Sex of the pig: F = female; M = male 3Hypoxia-inducible factor 1, alpha subunit (basic helix-loop-helix transcription factor) (HIF1A); Heat shock transcription factor 1 (HSF1); Heat shock 70 kDa protein 1A (HSPA1A); Myocyte enhancer factor 2A (MEF2A); Myogenic differentiation 1 (MYOD1)

230

Figure A.1. Significant up- and down-regulated transcripts pertaining to highlighted cellular pathways in muscle (A) and liver (B) identified by IPA in HSTN relative to TNTN gestational treatment. The log2 fold change (FC) is noted with different colors corresponding to different degrees of FC; green shades indicate down-regulated transcripts while red shades indicate up-regulated transcripts. An asterisk (*) indicates that a given gene was represented in the data set more than once. The effects of each gene on the affiliated function or other genes are represented as red (activating), green (inhibiting), or yellow (inconsistent with prediction) arrows; blue lines are protein-protein interactions. Refer to Table A-1.5 for z scores associated with the above pathways for each gestational treatment relative to the TNTN group.

Figure A.2. Effects of pre- and postnatal thermal treatment and sex on protein abundance of HSPA1A in longissimus dorsi muscle. Fourteen first parity crossbred gilts were exposed to one of four environmental treatments consisting of heat stress (HS) or 231 thermoneutral (TN) conditions at different time periods during gestation. Dams assigned to TNTN and HSHS were exposed to TN or HS conditions, respectively, throughout the entire length of gestation. Dams assigned to TNHS and HSTN were heat-stressed for the first or second half of gestation, respectively. The offspring were also subjected to postnatal thermal treatments, TN or HS, constantly for five weeks beginning at 12 weeks of age. Representative western blot of whole muscle protein lysates with an antibody directed toward HSPA1A (A) indicated no differences detected for gestational treatments (B), postnatal treatments (C), and sex (D); no interactive effects were detected as well. Different letters indicate P < 0.05. 232

APPENDIX B. DAILY 12 HOUR RECTAL TEMPERATURE AND WEEKLY AVERAGE DAILY GAIN MEASUREMENTS DURING HEAT STRESS FOR 21 DAYS

APPENDIX C. HOURLY RECTAL TEMPERATURE RESPONSE DURING ACUTE HEAT STRESS

Figure C.1. Rectal temperature (TR) response during 24 h of heat stress (HS) from the first replication of study 1 (chapter 2). A) mean hourly TR of all rep 1 gilts (black; n = 48), gilts deemed tolerant (TOL; blue; n = 10), and gilts deemed susceptible (SUS; red; n = 10. B-C) Calculated slopes during hours 1-12, 13-19, and 20-24 for all gilts during rep 1. D) Calculated slopes during same time intervals for gilts deemed TOL and SUS.