An Inbred Rat Model of Exercise Capacity: the Path to Identifying Alleles Regulating Variation in Treadmill Running Performance and Associated Phenotypes

Health Science Campus

FINAL APPROVAL OF DISSERTATION Doctor of Philosophy in Biomedical Sciences

An Inbred Rat Model of Exercise Capacity: The Path to Identifying Alleles Regulating Variation in Treadmill Running Performance and Associated Phenotypes

Submitted by: Justin A. Ways

In partial fulfillment of the requirements for the degree of Doctor of Philosophy in Biomedical Sciences

Examination Committee

Major Advisor: George Cicila, Ph.D.

Academic Abraham Lee, Ph.D., PT Advisory Committee: Yasser Saad, Ph.D.

Joana Chakraborty, Ph.D.

John C. Barbato, Ph.D.

Senior Associate Dean College of Graduate Studies Michael S. Bisesi, Ph.D.

Date of Defense: June 22, 2007

An Inbred Rat Model of Exercise Capacity:

The Path to Identifying Alleles Regulating

Variation in Treadmill Running Performance

and Associated Phenotypes

Justin Andrew Ways

The University of Toledo College of Medicine

2007

DEDICATION

I am dedicating this dissertation to: all the people in my life who believe in me, especially when I don’t believe in myself; all others who struggle with who they are and what they want to do in life, it is never too late to find your path; all the teachers who had a profound impact in my life, not only educational, but spiritual and moral as well - I can only hope to give to others as much as you have given to me, my nephews, Nathan and Ethen, who I hope and pray achieve all their dreams in life.

iii

ACKNOWLEDGEMENTS

I am deeply grateful to the following:

God, for giving me the ability to achieve the pursuit of higher education and

follow my dreams; my family, for understanding the commitment it takes to pursue higher education

and for their support, even if they can’t remember what it is I do;

Drs. George Cicila and Soon Jin Lee for providing me the opportunity to pursue

my doctoral degree in their lab and for allowing me to embark on a teaching

career while performing my graduate studies;

Dr. John Barbato, for being a trusted mentor, friend, and brother;

Dr. Abraham Lee, for his guidance, both academic and spiritual;

Dr. Joana Chakraborty, for her outstanding support of students, personal

guidance, and providing me with the opportunity to utilize and practice

physiological education;

Dr. Yasser Saad, for challenging discussions that forced intellectual growth;

Dr. Brian Smith, an excellent rat, and probably human, surgeon;

Dr. Eric Morgan, for being a trusted friend and excellent cardiovascular resource;

Sarah, Kris, and Ramona, without whom none of the work in the past five years

would have been possible; my mentors and friends at Mercy College of Northwest Ohio; and last, but not least, Mom (a.k.a. Marianne), without whom I would not be

where I am or who I am today. You will always be the one I hold most dear.

TABLE OF CONTENTS

Introduction…………………………………………………………………….1

Literature Review………………………………………………………..…….6

Materials and Methods……………………………….……………..….…...53

Results…...…………………………………………………………….……..85

Discussion...…..……..……………………………………………….…….113

Conclusions...………………………………………………………………138

Summary……………………………………………………………………141

References……………………………………………………………….…142

Appendix A……………………………………………………………….…184

Appendix B……………………………………………………………….…185

Abstract..………………………………………………………………….…186

A LONG TIME AGO, WISE MEN ONCE SAID…

‘‘In a word, all parts of the body which were made for active use, if moderately used and exercised at the labor to which they are habituated, become healthy, increase in bulk, and bear their age well, but when not used, and when left without exercise, they become diseased, their growth is arrested, and they soon become old.” --Hippocrates

Socrates: "And is not bodily habit spoiled by rest and illness, but preserved for a long time by motion and exercise?" Theaetetus: "True." --Plato, approximately 400 B.C. (Plato: Theaetetus, Great Books, Vol 7, 1952, p. 518.)

INTRODUCTION

Aerobic exercise tests measure the integrative ability of multiple physiological

systems to adapt to acute aerobic exercise and are often used to determine physical fitness, assess overall health, and predict mortality (Hammond and

Froelicher, 1984; Lee et al., 1999; Myers et al., 2002). The functional capacity of

each system determines the overall efficiency of adaptation, which in turn

determines the quality of performance on an exercise test. Low performance,

then, reflects a compromise in the functional capacity of one or more

physiological systems, denotes substandard physical fitness, and may indicate

an elevated risk for disease development. While the associations between

aerobic performance, physical fitness, and overall health are well known, the

underlying factors involved are poorly understood (Blair et al., 2001). Identifying

these underlying factors is therefore an essential step toward a more throrough

understanding of the relationship between physical fitness and health.

Aerobic exercise capacity is a decidedly complex trait for which measures of

performance display continuous variation from low to high values in populations

(Britton and Koch, 2001; Koch and Britton, 2005). The additive effect of multiple

genetic and environmental factors influencing the physiological systems involved

in the adaptive response to exercise gives rise to the observed variation in

performance. Genetic factors alone may account for a significant portion of this

variation as human and rodent studies estimate the heritability of exercise

performance to range from 39% to 73% (Bouchard et al., 1998; Koch et al., 1998; 1

Bouchard et al., 1999; Lightfoot et al., 2001; Lightfoot et al., 2007). These

genetic factors are likely a complex mixture of multiple genes, or rather their allelic variants, each exerting relatively minor effects on performance (Jennings et al., 1989; Blair et al., 1995; Bouchard et al., 1999; Koch et al., 1999; Lightfoot et al., 2001; Barton and Keightley, 2002; Lerman et al., 2002). As numerous physiological systems are involved in determining the response to exercise, subsets of the aforementioned genes may be operating in only one or a few of these systems and behaving as discrete units. The sum effect of the response of each allelic variant in each organ system determines the overall response to exercise, and therefore overall performance capacity.

Complex systems such as those that determine aerobic exercise performance are difficult to study without the use of appropriate models. Such models help reduce the complexity of the system by providing the means to dissect it into simpler components. Rats have historically been used as models for physiological processes (James and Lindpaintner, 1997), but in more recent years they have been used to study the genetic component of complex traits such as hypertension, diabetes, and aerobic running capacity (Dahl et al., 1962b;

Dahl et al., 1962a; Colle et al., 1983; Barbato et al., 1998; Koch et al., 1998). A vast array of genetic tools has been developed to help resolve the molecular mechanisms underlying the complex physiology of the rat. However, congenic strains have been particularly useful because they isolate individual genetic factors and allow their effects on trait variation to be investigated independent of

other causative genetic factors (Joe, 2005; Lazar et al., 2005). Furthermore,

congenic strains facilitate the study of natural variation in complex traits rather

than using physical or chemical interventions that attempt to mimic trait variation

or induce disease (Goldblatt et al., 1934; Rossini et al., 1977; Zolotareva and

Kogan, 1978; Ahn et al., 2004; Pinet et al., 2004). Combined with the knowledge extracted from the relatively recent sequencing of the rat genome (Gibbs et al.,

2004), congenic rat strains provide a powerful tool for understanding factors that

contribute to the natural physiological variation for which so much data exists, including exercise performance.

Work in our laboratory has focused on studying the relationship between the complex physiological nature of aerobic exercise performance and the underlying allelic variants that determine variation in performance by developing rat models of aerobic running capacity (ARC). A treadmill running test similar to that of the

Bruce test used for human cardiovascular performance evaluations (Bruce et al.,

1963) has been used to selectively breed for high and low ARC in rats (Koch et al., 1998; Koch and Britton, 2001) and to survey the ARC of commercially available inbred rat strains, which identified the Copenhagen (COP) and DA strains as being the most divergent among eleven inbred strains evaluated

(Barbato et al., 1998). While the use of selectively bred rat strains is superior to available inbred strains with respect to being more highly divergent for ARC specifically, they are still genetically heterogeneous, which makes identifying the genetic factors responsible for the divergence difficult (Flint and Mott, 2001). The

COP and DA rats, on the other hand, are genetically homogeneous, naturally divergent for ARC, and readily available for the application of existing genetic tools that can be used to identify the underlying genetic factors responsible for variation in ARC.

Using a segregating F2 population bred from COP and DA rats we performed a genome scan to identify chromosomal regions, or quantitative trait loci (QTLs), linked to the strain variation in ARC. We detected two such regions on rat chromosome 16 (RNO16) and one region on rat chromosome 3 (RNO3) (Ways et al., 2002). We also observed an interaction between loci in two intervals carrying ARC QTLs, D16Rat55 and D3Rat56, where at least one DA allele was required at each locus to achieve a significant improvement in aerobic performance (Ways et al., 2002). Genes involved in the metabolism of energy substrates, particularly lipids, were identified near the ARC linkage peaks as potential candidates to explain the strain differences in ARC. Complementary observations using COP and DA rats also suggested that differential baseline cardiovascular function and utilization of metabolic substrates were significantly associated with the ARC strain differences (Barbato et al., 1998; Koch et al.,

1999; Chen et al., 2001; Walker et al., 2002; Ways et al., 2002; Lee et al., 2005).

Presently, we sought to confirm that the chromosomal regions identified in the

F2(COPxDA) genome scan do indeed influence variation in treadmill running performance. We developed congenic strains by transferring high performing DA chromosomal regions from the ARC QTL-containing intervals onto a low

performing COP genetic background. Similarly, we developed a reciprocal

congenic strain by transferring low performing COP chromosomal segments from

the ARC QTL-containing interval into a high performing DA genetic background.

Because of the high degree of complexity in the ARC trait, we sought to further

characterize these strains using anatomical, physiological, and molecular

measurements as means to confirm previously found and to identify new

potential intermediate phenotypes and genes that may help reduce the

complexity of the ARC model. These measurements included 1) weights of key

organs involved in energy homeostasis, 2) concentrations of key energy

substrates used as fuel for aerobic performance, 3) in vivo cardiac performance,

and 4) cardiac global gene expression analysis. We anticipated narrowing of the

list of underlying intermediate phenotypes and candidate genes within the QTL- containing intervals potentially responsible for heritable differences in ARC between COP and DA rats. In doing so, we hoped to develop the tools that would allow us to gain a more thorough understanding of the connections between exercise, fitness, and health.

LITERATURE REVIEW

Aerobic Exercise Capacity

Exercise Capacity and Overall Health

The broad concept of overall health attempts to explain how efficiently an

organism is capable of maintaining homeostasis, a state in which the

physiological processes occurring in the body, down to the molecular level, are

maintained within a relative state of balance. Disruption of homeostasis initiates a period of robust biochemical and physiological adjustments designed to either eliminate the source of disruption or to compensate for loss of function (Nowak,

1994). Failure to adjust to the disruption of normal physiological processes, however, produces a state of deterioration in the organism that compromises functional capacity and often results in the onset of disease. An understanding of the underlying causes, mechanisms, and pathways involved in the capacity to adjust to factors that alter homeostasis is therefore essential to understanding the relationship between health and disease.

Tests of exercise capacity are often used for clinical and research purposes because they allow investigators to measure the capacity of an individual to adapt to homeostatic disruption from physical stress (Shephard et al., 1968;

Hammond and Froelicher, 1984; Brooks, 2005). These tests are an indirect measure of how the body might respond to other stressors that are imposed by

etiological agents such as the manifestation of genetic abnormalities, congenital defects, and environmental variables (Nowak, 1994; Noonan and Dean, 2000). It is generally assumed then, that the better an individual performs during an exercise test, the greater capacity that individual has to maintain homeostasis when challenged, significantly reducing the risk for disease development.

Regular physical activity provides a constant stimulus to the body that forces chronic adaptational responses and allows the body to function at a higher capacity, ultimately leading to improvements in exercise performance, physical fitness, and overall health. In 1998, the American College of Sports Medicine

(ACSM) recommended a minimum exercise program of at least 20 minutes per day, three days per week at 55% of maximal heart rate for apparently healthy individuals to develop and maintain cardiorespiratory fitness and overall health

(Pollock et al., 1998). In fact, a regular routine of moderate aerobic exercise elicits significant improvements in the control of lipid abnormalities, coronary heart disease, osteoporosis, diabetes mellitus, hypertension, and obesity, significantly reducing the relative risk of morbidity and mortality from these and other related anomalies (Pollock et al., 1998; Lee et al., 1999; Bray, 2000; Myers et al., 2002; Brooks, 2005; Warburton et al., 2006).

It is not clear, however, whether physical fitness or regular physical activity is more important in determining overall health. While it is true that regular physical activity can lead to improvement in physical fitness, regardless of the initial state of fitness, can innate fitness substitute for the benefits of regular physical

activity? Booth (2007) and Blair (2001) conclude that overall fitness as determined by an individual’s genetic makeup and interacting environmental factors is more influential on overall health than regular physical activity and is enough to confer at least some protective effects on overall health, even in sedentary individuals (Blair et al., 2001; Booth and Lees, 2007). However, they also note that physical inactivity, regardless of initial fitness, leads to maladaptive health consequences, including an increased prevalence metabolic and cardiovascular disorders (Spargo et al., 2007). Because it is difficult to remove environmental effects such as daily physical activity, training status, diet, and lifestyle from intrinsic ability, more well-designed studies are needed to accurately address the extent to which regular physical activity improves health beyond that of intrinsic physical fitness.

Bouchard and colleagues determined through studying monozygotic and dizygotic twins that genetic factors account for 40-70% of the variation in exercise performance phenotypes (Bouchard et al., 1986). Furthermore, they revealed that variation in intrinsic aerobic capacity (Bouchard et al., 1998) and variation in the improvement in aerobic capacity (i.e., the response to a 20-week exercise-training program) (Bouchard et al., 1999) are regulated by separate genetic components, accounting for 50% and 47% of the variance in each phenotype, respectively. They then attempted to identify the chromosomal locations of the genes responsible for the observed variation in both intrinsic capacity and the adaptational response to training (Bouchard et al., 2000).

Interestingly, while each phenotype generally mapped to discrete chromosomal

locations, there was some overlap between the regions found for intrinsic

capacity and the response to training, suggesting that some of these genes may be one in the same. This provided evidence that intrinsic capacity does not exist as an isolated phenotype and physical activity is probably necessary for the maintenance of intrinsic capacity as well as the related fitness and health benefits.

Overall health and aerobic fitness are related through complex interactions between genetic and environmental components influencing multiple molecular and physiological systems and their ability to adapt to stressful situations such as aerobic exercise. Some of these systems include, but are not necessarily limited to, the cardiorespiratory system, musculoskeletal system, neuroendocrine system, and metabolism/body composition. Figure 1 represents an attempt to model the complex nature of exercise capacity within the realm of these systems.

The functional capacity of each system shown is regulated by its own distinct set of genes. However, each system shares a subset of those genes with other systems, manifesting as unique, hybrid phenotypes. For example, there are subsets of genes regulating metabolism/body composition, the cardiorespiratory system, and the neuroendocrine system that together are responsible for regulating blood pressure. Since blood pressure is a component of exercise, it is logical to assume that subsets of genes from all systems would be combined to determine overall exercise capacity as well. These genes likely operate in

molecular pathways requiring the utilization of molecular oxygen (O2) to obtain the amount of energy required to fuel sustained activity. A discussion of these pathways, the physiological systems involved, and their ability to be manipulated is therefore required.

Metabolism and Body Cardiorespiratory Composition System

Exercise Capacity

Musculoskeletal Neuroendocrine System System

Figure 1. Venn diagram illustrating how maximal exercise capacity is a function of the cumulative effect of genes regulating multiple organ systems. Each circle represents all the genes involved in maintaining a particular physiological system. Regions of overlap between circles represent those genes that are shared between systems and manifested as distinct phenotypes themselves. The genes that are shared by all systems determine maximal exercise capacity.

The Aerobic Pathway

Initially, exercise is fueled by anaerobic energy supply systems using stored adenosine triphosphate (ATP), creatine phosphate, and anaerobic glycolysis for immediate energy sources. These anaerobic pathways supply enough energy for approximately sixty seconds of activity, after which exercising glycolytic muscles begin to fatigue (Schnermann, 2002). However, as the duration of exercise progresses, greater energy demand is placed on aerobic energy transfer systems and recruitment of oxidative muscle fibers to provide a long- term energy supply (McArdle et al., 1996; Ashe and Khan, 2004; Brooks, 2005).

This more efficient energy utilization via oxidative pathways enhances aerobic exercise performance.

The aerobic pathway begins with the respiratory system. Atmospheric air is brought into the lungs and oxygen diffuses down its partial pressure gradient from the alveoli into the blood via the respiratory membrane into the pulmonary capillaries. Hemoglobin molecules within passing erythrocytes bind oxygen and transport it back to the heart for distribution to the systemic circulation. During exercise, increased ventilation and increased blood flow through the pulmonary vasculature facilitates an increase in the rate of oxygen diffusion into the bloodstream. Erythrocyte transit time remains slow enough, however, to allow hemoglobin molecules to still become fully saturated with molecular oxygen

(Harries, 1994; Ashe and Khan, 2004; Brooks, 2005). Thus, the pressure gradient of oxygen from the alveoli to the blood and the capacity to saturate

hemoglobin remain relatively constant at rest and during exercise (Harries, 1994;

Ganong, 1999; Ashe and Khan, 2004; Brooks, 2005). This creates a remarkable

ventilatory reserve that does not necessarily limit exercise performance, although

the observation that exercise induced arterial hypoxemia occurs in some

endurance-trained athletes may suggest otherwise (Dempsey and Wagner,

1999). Lindstedt (1988) noted that maximal oxygen consumption in elite athletes

could only be achieved when they had very high hematocrits (Lindstedt et al.,

1988). Thus, the amount of oxygen transported in the blood is most likely

restricted by the concentration of erythrocytes in circulation and is clearly a

limitation to peak exercise performance at the level of oxygen carrying capacity.

Cardiac output and local vascular constriction are responsible for the

transport and distribution of blood throughout both the pulmonary and systemic

circulations, indicating that oxygen transport through the circulatory system is

highly dependent on cardiovascular function (Harries, 1994; Lingappa and Farey,

2000; Brooks, 2005). At the onset of exercise, working muscles apply pressure

to major veins leading back to the heart, resulting in increased venous return. To

accommodate this increase in volume load, the heart wall stretches beyond

resting parameters, resulting in stronger contractions. This causes a significant

increase in cardiac output via increased stroke volume (Brooks, 2005). Despite

the initial increase in cardiac output, blood pressure drops because local factors

(increasing CO2 concentration, decreasing pH, cytokine release, and enhanced intercellular communication) in working muscles cause relaxation of smooth

muscle surrounding the arterioles leading into muscle capillary beds.

Baroreceptors detect this drop in blood pressure and signal the activation of sympathetic nerve fibers from the vasomotor control center in the brain to increase vascular constriction. This leads to increased vascular resistance in inactive tissues, further increases in venous return, and an elevated heart rate.

The result is increased blood flow to working muscles with an associated increase in oxygen and substrate availability, waste removal, and diminished blood flow to splanchnic and renal circuits. Cardiac output is matched to the demand of the working muscles and blood pressure is elevated to maintain adequate perfusion (Brooks, 2005).

Treadmill running tests are most commonly used to assess cardiac function, cardio-respiratory capacity, and cardiovascular reserve capacity (McArdle et al.,

1996). These tests induce a very high oxygen demand by the heart, measuring the capacity of the heart to adapt to an increased workload. Enhanced contractility, systolic function, and increases in heart rate and stroke volume (i.e., cardiac output) are generally observed during exercise in healthy individuals.

However, the elevated heart rate observed during exercise results in a decreased time interval between heartbeats, greatly reducing ventricular filling time. Cheng (1992) found in conscious dogs that this reduced filling time was compensated for by an increased rate of isovolumic relaxation during early diastole (Cheng et al., 1992). This phenomenon not only permitted the ventricle to remain in diastole longer, allowing more time to fill despite a decreased filling

interval, but it also created a much lower ventricular pressure, increasing the

pressure gradient between the atrium (which is in continuous circuit with venous

blood) and ventricle, allowing for more rapid and efficient filling (Cheng et al.,

1990). These data suggest that diastolic function during exercise is crucial to the ability of the body to maintain homeostasis despite changing environmental conditions such as exercise. Indeed, a study comparing elite athletes with exceptional cardiovascular fitness to sedentary adults showed that the athletes had greater stroke volumes (and greater cardiac outputs), stronger ventricular contractions, and improved diastolic filling capacities compared to sedentary groups during exercise (Di Bello et al., 1996).

Others have observed that resting diastolic function is an important determinant of exercise capacity and that impaired diastolic function can lead to systolic dysfunction and other cardiovascular abnormalities that result in exercise intolerance (Cuocolo et al., 1990; Vanoverschelde et al., 1993; Dahan et al.,

1995). For example, patients with congestive heart failure as a result of impaired resting diastolic function have diminished cardiac output compared to healthy individuals both at rest and during exercise (Hiatt, 1991; Little et al., 2000).

Indeed, Guazzi (2001) found impaired diastolic function to be the underlying cause of exercise intolerance in hypertensive rats where left-ventricular hypertrophy had developed (Guazzi et al., 2001). Sumimoto (1997) observed a similar phenomenon in patients with left ventricular systolic dysfunction after anterior myocardial infarction (Sumimoto et al., 1997). Exercise training,

however, in both healthy adults and patients with cardiovascular abnormalities, is

capable of producing greater end-diastolic volumes, greater end-systolic

volumes, a higher ratio of early to late flow velocity through the mitral valve, and

a greater isovolumic relaxation time at rest (Vanoverschelde et al., 1993; Di Bello

et al., 1996; Libonati, 1999; Libonati et al., 1999). The result is improved

cardiovascular fitness, ability to adapt to changes in the demands placed on the

myocardium more effectively, and oxygen transport capacity which leads to

greater overall exercise performance and suggests that the consequences of

impaired diastolic function on performance can be alleviated via exercise training.

Once oxygenated blood reaches the tissues, oxygen is released from

hemoglobin down its partial pressure gradient from the capillaries to the cells. As

the demand for oxygen to produce energy in working muscle increases during

exercise, so does the need to extract oxygen from the blood. This requires a

larger volume of blood to be delivered in a shorter period of time to match supply

to demand. Although the amount of blood the heart is capable of circulating per

unit time is of major importance to attaining maximal exercise capacity (Mitchell

et al., 1958), the ability of peripheral working tissues to extract oxygen to meet

metabolic demand is crucial as well. The greater efficiency with which the

tissues extract oxygen, the greater potential the cells have to generate energy,

producing a higher power output and/or the ability to sustain a constant power output during exercise (i.e., endurance). Increased metabolic activity in muscle releases vasodilatory chemicals that increase tissue blood flow, enhance cellular

communication, and increase the efficiency of working tissues to utilize oxygen

extracted from the circulation (Brooks, 2005). In addition, a dense capillary

network creates substantial surface area capable of accommodating large

increases in blood flow and facilitating the perfusion of tissues with oxygen-rich

blood (Hoppeler and Weibel, 2000). Indeed, increased capillary density and

muscle efficiency has been associated with superior running performance in rats

(Henderson et al., 2002; Howlett et al., 2003). The oxygen-poor environment of

exercising muscle establishes a large partial pressure gradient, promoting rapid

transit of oxygen from the high oxygen environment in the circulatory system to

the cells. Once inside the cellular component of working tissues, oxygen is

utilized by the energy producing pathways of the cell.

The final components of the aerobic pathway are the mitochondria, the sites

of cellular energy production. The amount of ATP produced by direct combustion

or anaerobic metabolism alone would be insufficient to sustain mammalian life,

let alone exercise. The evolution of pathways including the citric acid cycle and

the electron transport chain in the mitochondria allowed for the production of

large quantities of ATP capable of supporting all sizes of multi-cellular organisms by using molecular oxygen (O2) as a final electron acceptor for energy transfer

(Baldwin and Krebs, 1981), a process known as oxidative phosphorylation.

Many disorders associated with aging and energy consumption appear to result

from oxidative damage in cells, which tends to disrupt energy transfer processes in the mitochondria. Individuals who actively exercise, however, are capable of

significantly reducing cellular oxidative stress, mitochondrial damage, and the pathogenesis of related disorders (Wallace, 2005).

Across tissues types and mammalian species, there is little variation in the capacity of individual mitochondria to utilize oxygen (Hoppeler and Weibel,

2000). Instead, the capacity to produce energy in the form of ATP depends on the density of mitochondria in a given cell or tissue (Hoppeler and Weibel, 1998).

Exercise training increases mitochondrial density in actively trained muscles, leading to enhanced oxygen consumption in the whole organism (Hoppeler et al.,

1995). Since active muscles utilize about 90% of available oxygen during maximal exercise (Mitchell and Blomqvist, 1971), increases in mitochondrial density lead to improved aerobic exercise performance and overall fitness. It can also be assumed, then, that an intrinsically high mitochondrial density in muscles would provide a similar benefit.

Complementary Energy Pathways

Complementary energy flow pathways involving the breakdown of carbohydrates and lipids, and to a lesser extent proteins and nucleic acids

(Stryer, 1995), provide the substrates for energy production in oxygen flow pathways. Lipids are the primary fuel source for muscles at rest and during sustained exercise of low to moderate intensity. However, as exercise intensity increases, fuel preference switches to carbohydrates and lactate (Romijn et al.,

1993; Brooks and Mercier, 1994; Brooks, 1998) in a phenomenon known as the

“Crossover Concept.” Lipid stores represent a virtually inexhaustible energy supply during exercise relative to the amount of time spent exercising, whereas stored carbohydrates are limited due to their considerable capacity to bind water and expand cellular volume (Brouns and van der Vusse, 1998). The sooner an individual switches from lipids to carbohydrates, the sooner limited energy supplies will exhaust, resulting in fatigue. It is generally accepted that low to mild intensity is 30-45% of maximal oxygen consumption, moderate intensity up to 45-

65%, and high intensity is 65% and higher (Brooks and Mercier, 1994; McMurray and Hackney, 2005). Since endurance training increases the capacity to consume oxygen, the same rate of oxygen consumption would be considered low intensity for trained individuals who have a high capacity, but high intensity in untrained individuals with a lower capacity. Therefore, the level of intensity during progressive exercise will reach a maximum sooner in individuals with a lower capacity, forcing an earlier switch to carbohydrate fuel sources and subsequent fatigue.

Lipids are stored in peripheral adipose tissue, muscle, and the liver as triglycerides. During physically stressful events such as exercise, sympathetic nervous system (SNS) activation and the subsequent release of catecholamines promote the breakdown of triglyceride stores into free fatty acids (FFA)

(McMurray and Hackney, 2005). The FFA can be used immediately for energy production in the mitochondria of muscle cells or released into the circulation by adipose tissue as a readily available energy source for exercising muscle. At

rest and during low intensity exercise, the concentration of FFA in the blood

determines how much enters the muscles (both heart and skeletal) for use as

energy (Brouns and van der Vusse, 1998). However, Martin (1993) determined

that well-trained endurance athletes rely more on intramuscular triglyceride

(IMTG) stores than blood-borne fatty acids, bypassing most barriers associated

with fatty acid transport from the blood to muscle mitochondria (Martin et al.,

1993). This implies that exercise training promotes increased IMTG storage and

the utilization of circulating FFA as a reserve, making lipid utilization more efficient.

Carbohydrates are stored in liver and muscle as glycogen. Stimulation by

catecholamines also promotes glycogen breakdown and subsequent release of glucose from the liver into the circulatory system and immediate utilization by

working muscle (Wendling et al., 1996). As exercise progresses to higher

intensities and increasing amounts of lactic acid are produced, both lipolysis and

lipid oxidation are inhibited, increasing the relative circulating abundance and

utilization of glucose for energy through the remainder of high intensity exercise

(Brouns and van der Vusse, 1998). Exercise training increases the threshold of

exercise intensity at which lactate is released (Joyner, 1991), increases the

capacity for lactate clearance (Brooks, 2005), and reduces catecholamine levels

both at rest and during exercise (Winder et al., 1979). The sum effect is greater reliance on lipids as fuel sources during sustained exercise and sparing of glycogen reserves, both of which delay the fuel crossover, resulting in enhanced

exercise performance and a greater energy reserve potential for improved energy homeostasis both at rest and during exercise.

The relationship between energy homeostasis and exercise capacity has important implications for body composition and its association with health.

Increased amounts of subcutaneous and visceral abdominal fat are associated with decreased exercise capacity and cardiorespiratory fitness (Wong et al.,

2004). Futhermore, regular exercise promoting improvements in cardiorespiratory fitness attenuates obesity-related health risks by reducing the total amount of adipose tissue (Wei et al., 1999; Ross et al., 2000; Stevens et al.,

2002; Janssen et al., 2004). While this generally means that individuals with a high aerobic fitness have less abdominal and total body fat, it has been shown that obese men and women with high aerobic fitness were at a lower risk for all- cause mortality than lean men and women with low aerobic fitness (Blair et al.,

1996). Therefore, aerobic fitness appears to be more important to overall health than body composition, regardless of activity status, although active lifestyles are certainly beneficial to improving both.

The reduction in adipose tissue associated with exercise training stems from increased sensitivity to sympathetic nervous system (SNS) activity and catecholamines via increases in the number of receptors promoting lipolytic responses (Bouchard et al., 1993; Harant et al., 2002; McMurray and Hackney,

2005). Overall catecholamine release into the circulation is reduced both at rest and during low to moderate intensity exercise (Rahkila et al., 1980; Hurley et al.,

1986), which promotes the sparing of glycogen stores. This not only promotes

weight loss by reducing adipose tissue mass, but also increases overall energy

stores within muscular tissue, resulting in greater muscular efficiency (Martin et

al., 1993).

Since the ability to perform aerobic exercise requires an abundance of

energy, increased energy stores represent the potential for greater exercise

capacity. Collectively, the pathways of oxygen flow, nutrient metabolism, and

overall regulation of these pathways contribute considerably to the complex

phenotype of aerobic capacity. Each component requires intricate dissection to

understand individual differences so that the variation in exercise capacity and

fitness-related contributions to overall health can be defined more

comprehensively.

Genetic Models of Exercise Capacity

Human Studies

Exercise capacity is an exceptionally complex trait due to the involvement of

multiple physiological systems that are controlled by the combined influence of multiple genetic and environmental factors. Genetic factors explain a large amount of performance variation, with human twin and heritability studies estimating the genetic contribution to be as high as 50% (Bouchard et al., 1986;

Bouchard et al., 1999). The HERITAGE (HEalth, RIsk factors, exercise Training

And GEnetics) Family Study is an ongoing long-term study aimed at identifying

associations between genes, exercise, and health related phenotypes in humans

(Bouchard et al., 1995; Bouchard et al., 1998; Bouchard et al., 1999; Rice et al.,

1999; Bouchard et al., 2000; Chagnon et al., 2001; Rankinen et al., 2001a; Rico-

Sanz et al., 2004) and analyzing the effect of specific gene variations on human exercise performance (Rankinen et al., 2000a; Rankinen et al., 2000b; Garenc et al., 2001). Ninety Caucasian families and 40 African-American families were pre- tested for maximal exercise capacity, exercise trained for 20 weeks at a heart rate equivalent to 75% of each individual’s maximal oxygen consumption, then post-tested. All tests and training were performed on a bicycle ergometer and a battery of cardiovascular, respiratory, and metabolic measurements made on each individual throughout the study. Dietary and other lifestyle components were assessed by questionnaires to assess compliance and potential complicating factors. Due to the large amount of data generated, cell lines were established from the monocytes of each participant for a constant supply of DNA.

Linkage and heritability studies were performed to assess the extent to which

gene variation is responsible for both intrinsic ability and performance

improvement as a result of exercise training and to identify those genetic variants

that are causative of those phenotypic differences (Bouchard et al., 1995).

Investigators involved with the HERITAGE Family Study also publish an annually

updated review of all human genetic loci related to physical performance or

health-related phenotypes in peer-reviewed journals (Rankinen et al., 2001b;

Rankinen et al., 2002; Perusse et al., 2003; Rankinen et al., 2004; Wolfarth et al.,

2005; Rankinen et al., 2006). However, heterogeneity (genetic and

environmental) within human populations and ethical considerations makes it

extremely difficult to isolate and manipulate the effects of individual genes on

exercise performance. Therefore, animal models that minimize genetic and environmental variation and allow for the utilization of more precise genetic tools to dissect the complex trait of exercise capacity are required.

Rodent Models

Although genetic studies in the mouse and rat started at approximately the

same time, the mouse became the mammal of choice for genetic studies while

the rat was used extensively as a model organism for physiological research

because of its larger size (Jacob, 1999). The wealth of information regarding its

general biology, behavior, biochemistry, neurobiology, physiology,

pharmacology, and biomedicine, however, has rendered the rat at least as

equally important as the mouse as a genetic model (James and Lindpaintner,

1997; Jacob, 1999). Now that three mammalian genomes; human (Lander et al.,

2001; Venter et al., 2001), mouse (Waterston et al., 2002), and rat (Gibbs et al.,

2004) have been sequenced, comparative genomics can be used to associate

what has been found in rodents to the human genome to aid in identifying human

genes associated with health and disease (Lindblad-Toh, 2004). DiPetrillo and

colleagues agree that because the human and rodent genomes maintain a fairly

high degree of similarity, rodent genomes provide a starting point for understanding the nature of quantitative variation in human health and disease

(DiPetrillo et al., 2005).

The heritability of intrinsic, or untrained, exercise capacity in rodents has been estimated to range from 39-73% (Koch et al., 1998; Lightfoot et al., 2001; Lerman et al., 2002), which is similar to heritability estimates in humans (Bouchard et al.,

1986; Bouchard et al., 1998; Bouchard et al., 1999). Although the number of rodent studies aimed at unraveling the quantitative nature of exercise capacity has been growing recently, few investigators have maintained an ongoing pursuit to that end. Those rat models currently being exploited to study quantitative variation in aerobic running capacity are 1) long-term selective breeding for high and low treadmill running capacity, and 2) inbred strains naturally divergent for treadmill running capacity. While the pursuit of both models began in the same laboratory (Barbato et al., 1998; Koch et al., 1998), they are currently being investigated independently at two different institutions. The merits, limitations, and progress of both models will be discussed.

Selective Breeding Approach

Selective breeding begins with an outbred population of rats that have been carefully mated to maintain maximal genetic variation. Outbred rats may be heterozygous at any given locus and allelic variants are likely to have randomly segregated throughout the population. Ideal outbred rat stocks used for genetic

research are those that are created from the intercrossing of several inbred rat strains to create a genetically heterogeneous population for which the characteristics of the founders are known (Hansen and Spuhler, 1984). The advantage of using a widely outbred population is that there is a large amount of genetic variation that can be selected upon to concentrate the alleles determining the extreme values for the phenotype of interest (Britton and Koch, 2001). The animals with the highest values for the phenotype are bred together, as are those with the lowest values for the phenotype, in such a way that maximizes the intensity of selection and minimizes population size and the extent of inbreeding

(Britton and Koch, 2001). After several generations of artificial selection, the alleles causative of high and low phenotypic values become increasingly concentrated in the high and low populations, respectively. When the trait values begin to plateau, the selection process is likely complete and inbreeding can begin (discussed below). The ultimate goal of artificial selection is to “fix” the characteristics of the selected lines by creating inbred strains exhibiting, theoretically, the most extreme values for trait of interest so that an ideal genetic substrate for studying the trait of interest is produced (Britton and Koch, 2001).

Selective breeding for treadmill running capacity in a heterogeneous population of Sprague-Dawley rats was initially used to estimate the heritability of the trait (Britton and Koch, 2005). The measure of this trait consisted of each rat running for five consecutive days on a motorized treadmill set at a constant 15° incline and an initial starting velocity of 10 m/min. The velocity was increased 1

m/min every 2 minutes until the rats were exhausted, which was defined as the

third time the rat would endure the shock grid located at the rear of the treadmill

for two seconds rather than run. The best day out of five was used as the

estimate of running performance as it was deemed most closely associated with

the genetic component of running capacity. After three generations, the low and

high lines that were selected differed in running capacity by 70%, with 39% of the

variation in running capacity between the selected lines being genetically

determined (Koch et al., 1998). Subsequently, a large-scale selective breeding

process using the more genetically heterogeneous National Institutes of Health

(N:NIH) outbred rat stock was initiated with the intent of producing low and high selected lines that were widely divergent for treadmill running capacity. The founder population had a running capacity of 355 ± 144 m (Koch and Britton,

2001). By generation 6, the two lines differed in running capacity by an average of 171% (Koch and Britton, 2001). Currently, the selection process is at

generation 20 and the lines differ in running capacity by 435%, with the low line

running to exhaustion at 302 ± 67 m and the high line running to exhaustion at

1617 ± 187 m (Koch, 2007). These selectively bred lines are aptly named low

capacity runners (LCR) and high capacity runners (HCR), respectively.

Although the LCR and the HCR rats are in the continuous process of

selective breeding, they have been extensively characterized for a number of

anatomical and physiological traits associated with exercise performance and

health via collaborative efforts. Generation 3 rats were tested for isolated cardiac

function using a Langendorff-Neely working heart preparation. The HCR rats

had a greater cardiac output than LCR rats (49.9 ml ⋅ min-1 ⋅ g heart weight-1 vs.

34.0 ml ⋅ min-1 ⋅ g heart weight-1, respectively), yet they maintained similar heart rates (Hussain et al., 2001). This suggests that the difference in cardiac function between the two lines was due to factors producing a greater stroke volume in

HCR rats, providing greater oxygen delivery to working muscles. This was supported by findings in generation 11 showing that in isolated cardiomyocytes,

HCR rats had superior systolic and diastolic function compared to LCR rats

(Wisloff et al., 2005).

However, a study performed using generation 7 rats revealed that the maximal cardiac output during exercise (including stroke volume and heart rate) of HCR rats did not differ from LCR rats. Furthermore, HCR rats had a higher maximal oxygen consumption than LCR rats (64.4 ± 0.4 ml ⋅ min-1 ⋅ kg –1 vs. 57.6

± 1.5 ml ⋅ min-1 ⋅ kg –1, respectively) that was not due to oxygen delivery via

cardiac output, but rather a greater skeletal muscle capillary density promoting

more efficient oxygen extraction at the tissue level (Henderson et al., 2002). This

was supported by the findings that HCR rats had greater skeletal muscle

capillary density, citrate synthase (CS) activity, β-hydroxyacyl-CoA

dehydrogenase (HADHB) activity, and phosphofructokinase (PFK) activity

compared to LCR rats (Howlett et al., 2003). In subsequent generations, HCR

rats, compared to LCR rats, also had greater concentrations of proteins related to

skeletal muscle mitochondrial function and biogenesis, including peroxisome

proliferative activated receptor gamma (PPAR-γ), PPAR-γ coactivator 1 alpha

(PGC-1α), ubiquinol-cytochrome c oxidoreductase core 2 subunit (UQCRC2), cytochrome c oxidase subunit I (COXI), uncoupling protein 2 (UCP2), and ATP

+ synthase H -transporting mitochondrial F1 complex (F1-ATP synthase) (Wisloff et

al., 2005). Thus the breeding process appeared to be selecting for peripheral

factors causing divergence in maximal oxygen consumption via differential

efficiency in skeletal muscle oxygen utilization rather than cardiovascular

differences.

However, oxygen transport experiments repeated using generation 15 rats

revealed that not only was the difference in skeletal muscle oxygen transport

capacity preserved, but the HCR rats also had greater convective transfer of

oxygen due to superior cardiac output (as a function of stroke volume) and

diminished peripheral vascular resistance compared to the LCR rats (Gonzalez

et al., 2006b). Although generation 3 and 11 rat hearts displayed a difference in

intrinsic cardiac function, the difference found at the whole animal level may not

have manifested itself until generation 15 for several possible reasons. First,

although the selection process concentrates the genes for high and low exercise

capacity in the separate lines; alleles may be segregating throughout the

populations because they are still genetically heterogeneous. This would result

in slight phenotypic differences each generation, the most pronounced being the

continued divergence in exercise capacity. As the selection process continues

further, the lines will likely become more divergent for other intermediate

phenotypes as well.

Second, intrinsic cardiac function was measured at baseline levels whereas

the measurements made at maximal oxygen consumption were taken during

high intensity exercise. Although intrinsic cardiac function may be able to predict

exercise capacity to some degree, other factors in the intact exercising animal

are capable of influencing cardiac function regardless of intrinsic capacity

(Barbato et al., 2002; Barbato et al., 2005). For example, Wisloff (2005) found that HCR rats were able to mediate vascular relaxation much more efficiently than LCR rats, reducing the load the heart has to pump against. Resting blood pressure in LCR rats was therefore higher than that of the HCR rats (Wisloff et al., 2005). Furthermore, peripheral vascular resistance was significantly greater in LCR rats by generation 15, further diminishing the functional capacity of LCR rat hearts compared to HCR rat hearts (Gonzalez et al., 2006a). Autonomic regulation of cardiovascular function was also blunted in LCR rats compared to

HCR rats, which may be a contributor to the observed dysregulation of peripheral vascular resistance and increased susceptibility to ventricular tachyarrhythmias

(Lujan et al., 2006).

Lastly, it is quite clear that cardiac function and maximal oxygen consumption, while very closely associated with maximal exercise performance, do not account for all variation in the trait. Maximal exercise performance is determined by how efficiently the individual is performing and how close to their maximum they are

maintaining performance. For example, an individual running at 90% of their

maximum oxygen consumption for an extended period of time will outperform an

individual who has a higher capacity, but only performs to 75% of that capacity

(Costill et al., 1971; Costill et al., 1973). Evidence for this exists in the

observation that the fastest marathon times do not correspond to those predicted

by measurements of maximal oxygen consumption (Costill et al., 1971). While greater maximal oxygen consumption is a requirement for achieving high levels

of performance, other factors contribute to the limitations of exercise capacity.

These factors include speed, ability to continue exercising at a high percentage

of maximal oxygen consumption, lactate clearance capacity, and economy of

performance (Joyner, 1991; Brooks, 2005) as well as the physiological systems

through which they operate.

The HCR and LCR rats have also been phenotyped for markers of central

fatigue (Foley et al., 2006) and muscle metabolism (Walsh et al., 2006; Spargo et al., 2007). While peripheral factors such as muscle efficiency, cardiovascular function, and oxygen transport are the usual suspects when considering exercise

limitations, the role of the central nervous system has been overlooked until relatively recently (Kayser, 2003). Evidence at the level of basal mRNA expression implies that HCR rats have more serotonin and dopamine receptors in areas of the brain known to participate in movement, motivation, and fatigue

(Foley et al., 2006), suggesting that perception of fatigue may be important in determining endurance capacity differences. At the muscular level, HCR rats

had greater mitochondrial sensitivity to creatine and LCR rats had dysregulated muscle lipid metabolism associated with metabolic risk factors (Wisloff et al.,

2005; Walsh et al., 2006; Spargo et al., 2007). These factors may be more important in the economy of exercise and the capacity to continue exercising despite reaching maximal oxygen consumption, leading to, at least in HCR rats, a much higher capacity than what may be explained by oxygen consumption alone.

The previous information, while very useful for identifying divergent phenotypes that are intermediates of exercise capacity in the HCR and LCR rats, reveals little information regarding the nature of the genes causative of the divergent phenotypes. Certainly protein and mRNA expression differences have been found in skeletal muscle (Howlett et al., 2003; Wisloff et al., 2005; Spargo et al., 2007), liver (Wisloff et al., 2005), and the brain (Foley et al., 2006), but no attempts have been made to characterize these differentially expressed genes at the level of DNA sequence. Furthermore, the genetic heterogeneity of each line does not lend itself to co-segregation studies that would identify QTLs containing the allelic differences causative of the divergence in running capacity between

HCR and LCR rats. The inbreeding of these lines is probably a long way off and the likelihood of identifying these genes anytime soon is very small. Therefore, the alternative approach of using readily available inbred strains is being exploited to identify genetic factors causative of variation in treadmill running capacity.

Inbred Approach

Inbred rat strains are those that have been exclusively sister-brother mated for at least 20 generations so that at least 97.5% of all genetic loci are homozygous (Lindsey, 1979). Thus, every member of a particular strain is virtually an identical twin of every other member, which minimizes genetic variation within a strain and leaves any observed variation in phenotypic values almost entirely a function of environmental influence. Minimizing environmental variation, therefore, would make phenotypic variation between strains largely the result of genetic variation. The benefits of using these strains include ease of genotyping (there is only one genotype per strain), phenotyping (increasing reproducibility and requiring less animals), and allowing multiple, independent investigators to evaluate the same genetic substrate for a given phenotype

(Rapp, 1983; Rapp, 1995; Rapp, 2000; Britton and Koch, 2001). One drawback, however, is that some strains are more sensitive to environmental variation than others which may produce misleading interpretation of results (Crabbe et al.,

1999). Therefore, gene-environment interactions must always be considered when analyzing results.

As of 2002, there were 234 inbred rat strains available for research (Jacob and Kwitek, 2002), most of which were selectively bred to model human disorders and disease characteristics (Greenhouse et al., 1990). Typically, the inbred strain modeling the disease is compared to an inbred control strain and the loci responsible for the differences in disease susceptibility can be mapped in

a segregating population bred from the two strains (Rapp, 1983). While any

given pair of inbred rat strains would be unable to capture all the genetic variance

related to a particular trait or disease, the availability of these strains represents a

quick and readily accessible substrate for quantitative genetic analyses.

Eleven inbred rat strains were evaluated for maximal exercise capacity and

intrinsic cardiac function using the same running protocol as that used for the

selectively bred model (Barbato et al., 1998). DA, so named because it

expresses the d blood group allele and has an agouti coat color, and

Copenhagen (COP) were identified as the two strains with the greatest

divergence in treadmill running capacity (182%) for both males and females.

Males from each strain, on the other hand, differed in capacity by 269%. The

significance of mentioning the capacity of only the males is that in all subsequent

genetic studies, only male rats were used to avoid the influence of the female

reproductive cycle.

Intrinsic cardiac function as measured by heart-weight adjusted cardiac

output in a Langendorff-Neely isolated working heart apparatus, was highly

associated with running capacity in both genders (r = 0.87) and in males alone (r

= 0.83). However, the COP and DA strains were not the most divergent for this

cardiac phenotype. The average cardiac output for male DA rats was 52.91 ±

0.85 ml ⋅ min-1 ⋅ g heart weight-1 compared to 40.0 ± 0.49 ml ⋅ min-1 ⋅ g heart

weight-1 for male COP rats, a 32% difference compared to the cardiac outputs of male DA and Buffalo (BUF) rats, which differed by 73% and were also widely

Table I. Summary of results from Chen et al. (2001), Koch et al. (1999), and Walker et al. (2002) % that DA was Intermediate Phenotype greater than COP 1. Isolated Papillary Muscle: - Maximal Developed Tension 38% - Rate of Tension Change During Contraction (+dT/dt) 61% - Rate of Tension Change During Relaxation (-dT/dt) 59% - Time constant of Relaxation (tau) 17% 2. Isolated Ventricular Myocytes: - Fractional Shortening 50% - Amplitude of Contractile Calcium Transient 79% 3. Left Ventricular Na+/K+-ATPase Activity 17% 4. Cardiovascular Variables : - Sympathetic Support of Arterial Blood Pressure 84% - Sympathetic Tonus at Rest 24% - Parasympathetic Tonus at Rest 192% - Blood Pressure at Rest 8% - Blood Pressure during Exercise 16% - Heart Rate during Exercise 4% - Heart Weight 32% - Relative Heart Weight (Heart Weight/Body Weight) 27% 5. Ecto-5’-Nucleotidase Activity - Left Ventricle 22% - Right Ventricle 46%

divergent for running capacity. Female DA and COP rats had a 70% difference in cardiac output, which was comparable to the difference between female DA and BUF rats (76%). This suggests that COP rats carry cardioprotective alleles that may be inactivated or not expressed in female rats of that strain. Further characterization of cardiovascular function was therefore warranted for COP and

DA rats.

To further dissect the intermediate phenotype of intrinsic cardiac function, hearts from male and female COP and DA rats were characterized at the tissue, cellular, and biochemical levels of organization (Chen et al., 2001). Table I

includes a summary of the results of that study. Hearts from DA rats clearly had superior systolic and diastolic performance compared to COP rats that appeared to be the result of more efficient calcium handling (Table I, #1-3). Because baseline intrinsic cardiac function may not be able to entirely explain the difference in running capacity observed between COP and DA rats (as discussed above for HCR and LCR rats), the effects of the autonomic nervous system on cardiovascular function were investigated (Koch et al., 1999) and are summarized in Table I (#4). Although heart rates, both at rest and during exercise, varied little between strains, the ability to regulate heart rate was much more pronounced in DA rats. In addition, the elevated blood pressure at rest and during exercise suggests more efficient tissue perfusion in DA rats. Since heart rates were not much different between the two rat strains, the greater blood pressure in DA rats, and presumably cardiac output observed by Barbato (1998), was due to a greater stroke volume. The differences in intrinsic function reported by Chen (2001) and by Walker (2002) (with regard to the greater capacity for DA hearts to produce adenosine) (Table I, #5) could reflect differences in cardiac reserve capacity available for exercise between the two strains.

Although the intermediate cardiovascular phenotype differences described above were useful physiological approaches in attempting to explain the differences in running capacity between COP and DA rats, the underlying genetic determinants remained unresolved. To address the genetic component, COP and DA rats were used to breed a segregating F2 population to perform a

genome scan that would identify chromosomal regions containing the allelic variants responsible for the variation in running capacity between COP and DA rats (Ways et al., 2002). A total of 224 rats and 210 polymorphic microsatellite markers (described below) were used to identify three regions, or QTLs, on rat chromosomes 16 and 3 as being linked to the variation in running capacity between COP and DA rats. There was significant linkage on chromosome 16 near D16Rat17 and two regions of suggestive linkage, on chromosome 16 near

D16Rat55, and on chromosome 3 near D3Rat56. The two regions demonstrating suggestive linkage were shown to interact in such a way that when a rat had at least one DA allele present at both D16Rat55 and D3Rat56, running performance was significantly greater than having one or no DA allele at either locus alone. Interestingly, genomic regions linked to strain variation in heart weight and relative heart weight were also identified, but on chromosomes

8 and 7, respectively. This suggests that differences in running performance and heart weight-related differences in cardiac function were under separate genetic control and probably unrelated. Thus, genes regulating the intrinsic cardiac function differences observed by Chen (2001) are more likely to be related to running capacity than genes regulating heart weight influenced cardiac function.

Genes involved in lipid metabolism located near the aforementioned ARC-

QTL peaks were proposed as potential candidates to explain the strain differences in running capacity (Ways et al., 2002). Lipids represent the most abundant source of energy for the body and are, in fact, the primary fuel sources

for cardiac and skeletal muscle at rest and during low intensity exercise (Ranallo and Rhodes, 1998; Goodwin and Taegtmeyer, 2000). It is possible that the difference in adenosine production between COP and DA rat hearts (Walker et al., 2002) reflects an underlying metabolic mechanism that would explain differences in cardiac function. Adenosine is known to inhibit the effects of catecholamines on the heart (Schrader, 1990). The differential adenosine production observed between COP and DA rats (Table I, #5), then, would affect the way the hearts utilize lipids during exercise and influence numerous variables of cardiac performance, including contractility (Stanley et al., 2005).

Similar approaches to identifying the underlying genetic determinants of exercise capacity have been pursued in mice. In particular, a wide range of variability in exercise capacity has been observed among inbred mouse strains

(Lightfoot et al., 2001; Hoit et al., 2002; Lerman et al., 2002; Billat et al., 2005).

While Hoit, Lerman, and Billat were interested in the interrelationships of genetic background, cardiovascular function, and exercise capacity, Lightfoot pursued identifying the underlying genetic determinants of variation in exercise capacity by performing a genome scan using an F2 population derived from inbred mouse strains identified as having high and low intrinsic exercise capacity (Lightfoot et al., 2007). Lightfoot’s observations, and the observations of the other mouse studies are consistent with the observations in the aforementioned rat studies with regard to interstrain variability, heritability estimates, and even orthologous locations of QTLs for running performance. Collectively, these studies support

and provide additional tools to aid other studies in the pursuit of identifying allelic variants responsible for variation in aerobic exercise capacity.

All information reported thus far regarding the genetic models of exercise capacity begs the question, “What next?” Numerous approaches have been taken in the past, including the identification of intermediate phenotypes that segregate with the trait of interest, proposal of candidate genes based on what is biochemically and physiologically known about the trait of interest, the construction of congenic strains that narrow down the genomic region carrying the gene(s) causative of variation in the trait of interest, and various combinations of the three (Rapp, 1995). These approaches remain the primary means by which quantitative trait analyses are pursued subsequent to genome scans, but the number of tools that complement these approaches has grown. For example, gene expression profiling has been used with congenic strains and proposed for use with segregating populations to identify gene expression patterns that co-segregate with quantitative traits for the purpose of narrowing down the list of candidate genes to study (Jansen and Nap, 2001; Lee and Cicila,

2002). Bioinformatics tools continue to grow and facilitate mining the enormous amount of data generated via high throughput genomic and expression analyses.

Some tools include comparative genomics, sequence analysis and comparison, expression databases, and pathway analysis programs that integrate multiple bioinformatics tools into one comprehensive program (DiPetrillo et al., 2005).

While these tools are useful in that they help provide a more complete picture

than what an independent investigator could generate alone, they do not

preclude the need to perform extensive laboratory work to prove causation.

Genetic Analysis of Quantitative Traits

Quantitative versus Mendelian Traits

Mendelian traits are those for which a single gene is largely responsible for phenotypic variation and a cause-and-effect relationship between genotype and phenotype is apparent. Mendelian traits are manifested as discrete phenotypes through the expression of alternate forms of the same gene, also known as allelic variants. Examples of human Mendelian disease traits include cystic fibrosis and sickle cell anemia where specific allelic variants result in a dysfunctional protein that degrades the functional capacity of the organs in which they are expressed.

On the other hand, quantitative (or complex) traits are those for which a one-to- one cause-and-effect relationship between genotype and phenotype is nonexistent (Darvasi, 1998). Phenotypic variation is regulated by multiple genetic loci (QTLs) whose products interact with each other and the environment to establish continuous phenotypic variation from low to high values (Rapp, 1983;

Rapp, 2000). Such traits include aerobic exercise capacity, blood pressure, height, body composition, and diseases such as arthritis and type II diabetes mellitus for which allelic variation at a single locus may not result in a significant overall change in phenotype.

Mendelian laws of inheritance (Figure 2) can help predict the transmission of alleles through multiple generations. In a simple example, if a gene has two alleles (B1 and B2) and each parent is homozygous for either allele, i.e., two

Genotype P B1B1 x B2B2

F1 B1B2

Intercross

F2 1 B1B1 2 B1B2 1 B2B2

Phenotype 3 B1 1 B2

Figure 2. Diagram representing Mendelian Laws of Inheritance where a cross between two homozygous parents (P) results in a heterozygous F1 population and intercrossing the F1 population results in a genotypic ratio of 1:2:1 and a phenotypic ratio of 3:1 because of the dominance effect B1 has over B2.

identical alleles for the same locus, then a mating between such parents will result in an F1 (first filial) generation where every offspring is heterozygous,

having two different alleles at the same locus. Intercrossing members of these F1 progeny will result in an F2 (second filial) generation with a genotypic distribution

of 1 B1B1, 2 B1B2, and 1 B2B2. If the gene follows a dominant mode of

inheritance where the phenotypic expression of B1 is significantly greater than B2,

then the ratio of offspring displaying the B1 phenotype to offspring displaying the

B2 phenotype will be 3:1.

The distribution presented above arises from each offspring receiving one

allele from each parent (either B1 or B2). Gametes only contain one copy of a genome (i.e., haploid) so the probability of an offspring inheriting a particular genotype can be predicted if the parental genotype at a given locus is known.

Phenotypic prediction, unfortunately, is not always as simple and straightforward.

Phenomena including, but not limited to, incomplete dominance, codominance, spontaneous mutations, pleiotropy, mosaic inheritance, variable expressivity,

multiple allelic variants (more than just two), and environmental influence can

obscure the relationship between genotype and phenotype. Despite these

complications, however, a strong one-to-one relationship often still exists

between the genotypic and phenotypic variants for a given gene (Glazier et al.,

2002).

The preceding discussion was modeled for one gene with two allelic variants.

Now imagine that model to include many genes, each with several allelic variants, the possibility of incomplete dominance, codominance, spontaneous mutations, pleiotropy, mosaic inheritance, variable expressivity, homologous recombination, and environmental influence and the result is a quantitative trait.

A basic premise of quantitative traits is that trait variation depends on genes that follow the same laws of inheritance as Mendelian traits, hence quantitative traits may be thought of as an extension of Mendelian traits (Falconer and Mackay,

1996). Given the inherent complexity of quantitative traits, however,

sophisticated methods are required for the detection, localization, and

identification of the underlying genetic factors regulating those traits.

Evaluating the Phenotype of a Quantitative Trait

To identify the allelic variants responsible for a particular trait, a

representative quantifiable measure of the trait must be developed. The

measure should follow certain criteria such as being relatively simple to perform,

objectively interpretable, gradable on a continuous numerical scale, and should

exhibit a wide range in magnitude between low and high values for the

phenotype (Britton and Koch, 2001). It should also be applicable to other

models, including and especially humans. For the aerobic capacity of rats,

distance run to exhaustion on a treadmill may be used as a measure of the trait.

A ramped treadmill running protocol modeled after the Bruce test for

cardiovascular fitness in humans (Bruce et al., 1963) is employed. It would be

impractical to measure aerobic capacity by directly measuring oxygen

consumption during exercise because the technical aspects limit the number of animals that can be studied (Henderson et al., 2002) and maximal oxygen consumption does not explain all observed variation in exercise performance

(Joyner, 1991; Gonzalez et al., 2006b). Furthermore, distance run to exhaustion has been shown to be highly variable in rodent populations (Koch et al., 1998;

Ways et al., 2002). Therefore, the best distance run on a treadmill serves as an

appropriate measure of aerobic exercise capacity and suitable phenotype for the genetic dissection of aerobic exercise capacity.

Evaluating the Genotype of a Quantitative Trait

Microsatellite markers are the most common polymorphic markers used for genotypic evaluation in the rat (Lazar et al., 2005), despite the recent increase in the use of single nucleotide polymorphisms (SNPs) for genomic analysis

(Zimdahl et al., 2004; Guryev et al., 2005). Microsatellite markers are short tandem DNA sequence repeats (typically a string of di- or tri-nucleotide repeats) that are scattered throughout mammalian genomes (rats in this case) within introns and non-translated regions surrounding genes (Weber and May, 1989).

For a given microsatellite marker, the number of sequence repeats is generally stable within a single inbred strain. However, the number of repeated sequences may vary between inbred strains, translating into DNA-size polymorphisms.

Polymerase chain reaction (PCR) amplification of the microsatellite marker using primers complementary to non-repetitive regions flanking the repetitive sequence and electrophoretically separating the products on an agarose or polyacrylamide gel reveals the size polymorphism between the strains via the migration pattern of PCR-products on the gel. The strain with the marker having more repeats generates a larger fragment that will not migrate as far down the gel as a smaller fragment generated from the strain with the marker having fewer repeats (Figure

3). Since the mouse and rat are the primary mammalian models used for genetic

research, the knowledge and availability of microsatellite markers, genome

maps, and polymorphism data for different inbred strains is vast and fairly well

organized. Hence, numerous microsatellite markers have already been used to

CACACACACACACACA Marker 1, Allele B1

Marker 1, Allele B2 CACACA

Gel Migration Pattern of Marker

B1 B1 B1 B2 B2 B2

Figure 3. Amplification of a microsatellite marker by PCR and separation by gel electrophoresis. Arrows represent forward and reverse primers. For Marker 1, allele B2 has less repeats and is thus smaller than allele B1 and migrates farther on the gel.

genotype many strains, mapped along the mouse and rat genomes, and made

available to the public. Typical websites with such information include: The Rat

Genome Database (http://rgd.mcw.edu) (Twigger et al., 2005); The Wellcome

Trust Centre for Human Genetics (www.well.ox.ac.uk/rat_mapping_resources)

(Wilder et al., 2004), Otsuka GEN Research Institute (http://ratmap.hgc.jp)

(Watanabe et al., 1999), and the National Center for Biotechnology Information

(NCBI; http://www.ncbi.nlm.nih.gov/genome/guide/rat/index.html).

Genetic Mapping

Mammalian cells contain two copies of each chromosome (i.e., diploid), which are referred to as homologous chromosomes. When meiosis occurs in germ-line cells, one copy of each chromosome is distributed to each daughter cell to produce haploid gametes. During the process of meiosis, homologous

chromosomes recombine, reciprocally exchanging chromosomal segments

(Russell and Russell, 1992). This phenomenon shuffles genetic information and

produces gametes with different combinations of genetic information. For all

autosomes, each gamete then contains one allele for each locus. When

fertilization occurs, the haploid genome from the male parent combines with the

haploid genome from the female parent to form recombinant diploid genomes in

the offspring. The first step in the process of identifying the genes underlying a

quantitative trait typically involves identifying QTLs mapping to approximate

locations in the genome (Doerge, 2002; Glazier et al., 2002; Abiola et al., 2003).

Fundamentally, this is accomplished by taking advantage of homologous

recombination as well as the Mendelian nature of single gene inheritance and the

partial influence on the quantitative phenotype (Rapp, 1995), although evidence

suggests that not all QTLs result from a single gene (Frantz et al., 2001; Garrett

et al., 2001; Saad et al., 2001; Garrett and Rapp, 2002b; Garrett and Rapp,

2002a). Allelic variants from each of two inbred strains that display divergence

for a given phenotype will segregate throughout the entire F2 population, creating

a wide distribution of phenotypic variation because of homologous recombination

and the inheritance of one set of alleles from each parent (Figure 4). Using genetic markers and calculating recombination frequencies between them, the entire genome can be ordered into a linkage map. If trait values for each

Low Strain High Strain X

F1 Entire Genome

Alleles Causative of Low Trait Value Alleles F2 Causative of High Trait Value Low High

Genes Messenger RNA Proteins Physiological Traits

Figure 4. Breeding of a segregating F2 population depicted using a four-locus model. Black squares represent alleles causative of low trait values and white squares represent alleles causative of high trait values. As the genes causative of low and high values segregate throughout F2 population so do the associated mRNA transcripts, proteins, and physiological traits.

member of the F2 population are known, a statistical association can be drawn between the genotype and variation in trait values at each genetic marker, which will map QTLs to a location on the genome. A trait value associated with a specific genotype of a genetic marker is considered linked to that marker and will cosegregate with it from parent to offspring. If a trait is not associated with the genotype of a given genetic marker, then the marker and the trait will segregate randomly with respect to each other in the population (Figure 5).

From Linkage to Gene Discovery

Once QTLs have been mapped to a general chromosomal location a number of approaches can be pursued to produce finer maps of the QTL-containing region and to identify candidate genes that may be influencing the variation observed in the trait. Congenic strains and congenic substrains can be constructed to confirm the presence of QTLs, test interaction effects, and narrow the QTL-containing region to within about 1 centimorgan (cM) (Rapp and Deng,

1995; Rapp et al., 1998; Nabika et al., 2000). The utilization of congenic strains originated with the work of George Snell studying histocompatibility genes in mice (Snell, 1948). Since then, congenic stains have been used by both the mouse and rat communities in studies of hypertension (Rapp, 2000), alcoholism

(Bennett et al., 2006), lipid phenotypes (Bottger et al., 1996), and arthritis (Joe,

2007), just to name a few.

Cosegregation No Cosegregation pe Value y Phenot

B1B1 B1B2 B2B2 Figure 5. Example of an additive modelMarker of Genotype cosegregation. Trait values associated with genotype at a particular locus will segregate with particular genotypes (cosegregation). Trait values that are not associated with genotype at a particular locus will segregate randomly with respect to genotype (no cosegregation).

To generate a congenic strain (Figure 6), two inbred strains (donor and

recipient) are intercrossed to produce an F1 population. The F1 population is

then back-crossed to the recipient strain and the resulting progeny are genotyped within the region containing a mapped QTL (Rapp and Deng, 1995). Those progeny that are heterozygous for selected loci from the donor strain are back-

crossed again to the recipient strain. This is repeated for at least eight

generations until the selected locus of the donor strain is fixed in the genome of

Donor Recipient

F1 Recipient

BC1 Recipient Entire Genome X Recipient BCn Recipient Alleles X Donor

Alleles BCn- female BCn - male

Figure 6. Schematic for constructing a congenic strain.

the recipient strain and greater than 99% of the genetic background is that of the recipient strain (Silver, 1995; Rapp, 2000). More recently, the use of “speed” congenics has improved the efficiency of making congenics by reducing the number of generations required to generate the strains (Markel et al., 1997;

Wakeland et al., 1997). Instead of eight generations, congenic strains can be generated in four to five generations by selecting against alleles from the donor strain in the background of rats in each backcross generation, while still selecting for alleles from the donor strain within the QTL-containing region. Only those

animals with the least amount of allelic contamination from the donor strain are

used to breed the next backcross generation.

Candidate gene approaches were traditionally used before the widespread

availability of genetic markers to make congenic strains (Garrett et al., 1998).

Based on the known physiology and biochemistry of a trait, individual candidate

genes were analyzed for polymorphisms in strains divergent for the trait and

tested for cosegregation using Restriction Fragment Length Polymorphisms

(RFLP) (Rapp, 1995). Now that genome scans are common practice in the

search for QTLs, candidate genes are chosen based on their location within

chromosomal regions identified as being linked to trait variances (Chagnon et al.,

2001), being functionally related to the trait of interest (Falconer and Mackay,

1996), and being differentially expressed or interacting with highly differentially

expressed genes (Lee and Cicila, 2002; DiPetrillo et al., 2005). Gene expression

analyses have been used successfully with congenic strains to identify allelic

variants in CD36 that influence insulin resistance (Aitman et al., 1999) and with linkage analysis to identify allelic variants in complement factor 5 (CF5) that influence allergen-induced susceptibility to asthma (Karp et al., 2000), among others.

The insulin-resistance QTL in which Cd36 was identified has also been linked to reductions in high density lipoprotein phospholipids (Bottger et al., 1996), high plasma blood triglycerides (Kovacs and Kloting, 1998), metabolic abnormalities in

adipocytes (Aitman et al., 1997), and hypertension (Pravenec and Kurtz, 2007),

all of which are well known cardiovascular risk factors. The fact that all of these phenotypes map to the same QTL and are probably regulated by the same Cd36 gene illustrates the importance of identifying phenotypes that are correlated and/or intermediate to the phenotype of interest when searching for the genetic basis of phenotypic variation to help reduce the complexity of the model. John

Rapp outlined certain criteria that are required to establish causation between intermediate monogenic traits and the quantitative trait being studied: 1) demonstration that there is a strain difference for the monogenic trait being studied, 2) adherence of the trait to Mendelian patterns of inheritance, 3) cosegregation with incremental changes in the quantitative trait, and 4) logical biochemical/physiological links between the monogenic trait and the quantative trait (Rapp, 1983). Although these criteria were proposed many years before the widespread use of high throughput polymorphic genetic markers, genome scans, and congenic strain construction for quantitative trait analysis, they still hold true for quantitative traits being studied today.

Once candidate genes are chosen, they can be sequenced using cDNA

(single-stranded DNA copies of ribonucleic acid (RNA) transcripts) or genomic

DNA (to confirm cDNA sequencing results or for analyzing gene regulatory regions) from the inbred strains for which the phenotypic variation exists. The goal of sequencing is to identify nucleotide polymorphisms between the two inbred strains within the candidate gene that would result in an amino acid substitution or altered protein structure/function having an effect on the

quantitative trait being studied. If the candidate gene in question is also

differentially expressed, then searching for sequence polymorphisms in gene

regulatory regions may also be required (Lee and Cicila, 2002). Although

sequence differences between strains and linkage do not prove that a candidate

gene is responsible for the phenotypic variation, they do provide support that the

gene in question at least contributes to overall phenotypic variance. The goal

then is the burden of proof: is this the only gene within a QTL that explains the phenotypic variation at that locus? Smits and Cuppen (2006) discuss a number of approaches to answer this question, but along with Glazier et al. (2002) admit that only relatively few genes have been found for QTLs, although the number has been growing rapidly in recent years (Glazier et al., 2002).

Defining every gene, allelic variant, and their roles in the complex pathways that make up quantitative traits is an enormous and complicated task. As more genes are discovered and functionally characterized, more models are developed, and the genomes of more model organisms are sequenced, our understanding of complex nature of quantitative traits will be enhanced significantly. Results can then be translated to human studies and lead to a greater understanding of disease etiology and advances in genetic screening for disease susceptibility. As pathways are progressively resolved, advances in drug development, diagnostic/prognostic strategies, and preventative therapies would be the logical next step (Jacob and Kwitek, 2002; DiPetrillo et al., 2004).

MATERIALS AND METHODS

Construction and Characterization of the COP.DA Congenic Strains

Animals

Inbred DA (DA/OlaHsd) and Copenhagen 2331 (COP/Hsd) rats were

purchased from Harlan Sprague-Dawley (Indianapolis, IN) and used to establish

a colony housed under specific pathogen-free conditions within the animal care

facilities at the University of Toledo Health Science Campus. All genetic crosses,

congenic, and inbred rats used throughout the rest of the study were bred from this colony. Rats were weaned at 28 days of age and housed two to three per cage on a 12:12 hour light-dark cycle with the light cycle coinciding with daytime.

Standard rat chow (Ralston Purina, diet 5001) and water were provided ad

libitum. All breeding and experimental procedures were carried out with the

approval of the Institutional Animal Care and Use Committee of the University of

Toledo Health Science Campus in accordance with the “Guiding Principles in the

Care and Use of Animals” as approved by the Council of the American

Physiological Society.

Congenic Breeding Paradigm

Chromosomal intervals on rat chromosomes 3 and 16 containing the putative high capacity ARC QTL alleles from DA rats were separately transferred into the

COP genetic background using a marker-assisted breeding (i.e. "speed congenic") approach (Markel et al., 1997; Wakeland et al., 1997), resulting in the

COP.DA(D3Rat233–D3Mgh14) congenic and COP.DA(D16Rat12–D16Rat90) consomic strains, respectively (Figure 7). Hereafter, these congenic strains will be referred to as COP.DA(chr 3) and COP.DA(chr 16). The breeding paradigm was as follows: male F1 rats, bred by crossing male COP rats with female DA

Figure 7. Outline of COP.DA congenic strain construction. The number of rats below each inbred/F1 strain indicates the number used to breed the next generation. The number of rats below each backcross generation indicates how many rats were produced each generation. Each generation required approximately three months to complete. Since the COP.DA(chr16) and COP.DA(chr 3) congenic strains required five and four backcross generations to complete, the entire process took approximately 1.5 years. The congenic strain populations were then expanded for testing the ARC phenotype. DA.COP consomic rats (discussed below) were bred similarly.

rats, were backcrossed to female COP rats. Male progeny heterozygous for the

RNO3 and RNO16 ARC QTL-containing regions (loci genotyped are described

below for each congenic strain) and containing the fewest number of DA alleles

at the other loci in the genome were selected for backcrossing to female COP

rats. After the first backcross generation, COP.DA(chr 3) and COP.DA(chr 16)

congenic strains were developed independently. For each generation, the male

rat heterozygous for all markers within the ARC QTL-containing region and

carrying the fewest DA alleles throughout the remainder of the genome was bred

with up to 8 female COP rats. A total of four and five backcross generations

were required to breed the COP.DA(chr 3) and COP.DA(chr 16) congenic

strains, respectively. Male and female rats heterozygous for the RNO3 or

RNO16 ARC QTL-containing regions and lacking DA alleles in the background

were then mated to fix the DA alleles in the congenic regions and COP alleles

everywhere else. Continuous brother-sister mating was subsequently used to

maintain congenic rat strains.

Genotyping

DNA was extracted from tail biopsy samples using kits (DNeasy Tissue Kits and DNeasy 96 Tissue Kits; Qiagen, Chatsworth, CA). PCR amplification and gel electrophoresis were performed as previously described (Ways et al., 2002).

Primers used to amplify polymorphic microsatellite markers were purchased from

IDT Technologies (Coralville, IA). PCR products were fractionated on 4%

agarose gels (Metaphor; Cambrex, Rockland, ME) and visualized using ethidium bromide staining with ultraviolet illumination. PCR-products unable to be

32 distinguished on agarose gels were amplified using P–labeled primers and the

PCR-products resolved on 8% polyacrylamide gels and visualized by autoradiography.

Microsatellite markers for genotyping were chosen from those used in the previous ARC QTL genome scan (Ways et al., 2002). Initially, five markers, spaced an average of 16.6 cM apart between markers D3Rat56 and D3Rat21, were used to transfer the RNO3 ARC QTL-containing region. Eight markers, spaced an average of 10.2 cM apart between D16Rat12 and D16Rat90, were used to transfer DA RNO16. A total of 105 markers, spaced an average of 17.5 cM apart, were used to genotype the rats during the selection of the two congenic strains.

Both congenic strains were then genotyped at 82 additional background loci that were included in the initial F2(COPxDA) genome scan (Ways et al., 2002), resulting in a total of 187 loci tested between the two congenic strains as follows:

COP.DA(chr 3) was genotyped at 180 loci, spaced an average of 11.2 cM apart, to test for the presence of DA alleles at locations other than the proximal portion of RNO3; COP.DA(chr 16) was genotyped at 177 loci, spaced an average of 11.4 cM apart, to test for the presence of DA alleles on chromosomes other than

RNO16. All background loci were homozygous for COP alleles in the congenic strains.

Marker density was increased within the QTL-containing regions to confirm

the lack of contamination from recipient loci. Locations of microsatellite markers

on RNO3 and RNO16 were obtained from the rat genome sequence database

(Build 3.4) at the National Center for Biotechnology Information website

(http://www.ncbi.nlm.nih.gov/genome/guide/rat/index.html). As these markers

were not used in the F2(COPxDA) genome scan (Ways et al., 2002), distances

between loci cannot be expressed in cM, and are therefore expressed in

megabasepairs (Mb). Seventeen additional markers were genotyped in the

proximal portion of RNO3, resulting in an average marker spacing of 5.4 Mb in

this interval. All of these loci were homozygous for DA alleles in the COP.DA(chr

3) congenic strain. Ten additional markers were genotyped on RNO16, resulting

in an average marker spacing of 5.0 Mb on this chromosome. All of these loci

were homozygous for DA alleles in the COP.DA(chr 16) consomic strain. Since

one Mb corresponds to approximately 0.62 cM (Glazier et al., 2002), the average

marker spacing in the congenic strains can be estimated at approximately 3.3 cM

for COP.DA(chr 3) and 3.1 cM for COP.DA(chr 16).

ARC Phenotype

A ramped test for maximal treadmill running capacity was used to assess

ARC as previously described (Barbato et al., 1998; Koch et al., 1998; Ways et al., 2002). At 10-weeks of age, male COP (n = 36), COP.DA(chr 3) (n = 39),

COP.DA(chr 16) (n = 24), and DA (n = 28) rats underwent an education week to

become acclimated to running on the treadmill (Model Exer-4, Columbus

Instruments, Columbus, OH). At 11-weeks of age, rats were tested for maximal

treadmill running capacity on five consecutive days. The test consisted of a 10 m/min starting speed that increased 1m/min every 2 minutes at a constant 15°

incline. Rats were removed from the treadmill at the point of exhaustion, operationally defined as the third time a rat would sustain three seconds on a shock grid (1.2 mAmp at 3 Hz) located at the back of the treadmill lane rather than run. At the end of each run, the rat was removed from the treadmill and its body weight was measured.

For each of the five consecutive days of running, the total distance run to exhaustion (m) was used as the estimate of performance. The single best distance run out of the five days of testing was considered the distance most closely associated with the heritable component of aerobic running performance, as previously described (Barbato et al., 1998; Koch et al., 1998; Koch and

Britton, 2001). A three-week recovery period followed the performance testing to avoid potential training effects on other phenotypic measures. Rats were then divided into fasting and ad libitum-fed subsets and used for further experiments to measure body and organ weights (ad libitum-fed animals only), as well as blood chemistries (fasted and ad libitum-fed animals), at 15-weeks of age.

Ad Libitum-Fed Organ Weight and Plasma Measurements

An ad libitum-fed subset of the previously run, 15-week-old rats [COP (n =

15), COP.DA(chr 3) (n = 18), COP.DA(chr 16) (n = 8), and DA (n = 12)] was used to obtain organ and plasma samples. Rats were weighed and then anesthetized with sodium pentobarbital (50 mg/kg body weight). A midline incision was made in each rat from the abdomen through the lower portion of the thoracic cavity. An

18 gauge, 1.5 inch needle (Becton Dickenson, Franklin Lakes, NJ) was inserted into the right ventricle through the apex of the heart and venous blood was drawn into K2EDTA coated Vacutainer tubes (Becton Dickenson, Franklin Lakes, NJ)

and set on ice. After blood collection, hearts were removed, blotted dry, and

weighed. Left ventricles were isolated by cutting the atria and major blood

vessels from both ventricles, along with the top portion of the left ventricle and

septum to avoid contamination. The right ventricle was then bisected and cut

away proximal to the left ventricle and septum, after which the left ventricle was

blotted dry and weighed. The liver was removed as a single unit, blotted dry, and

weighed. The pancreas was dissected free from its attachment to the spleen and

then medially toward the duodenum, blotted dry, and weighed. Each kidney was

separately removed, decapsulated, blotted dry, and weighed.

Abdominal fat content was dissected as follows: first, flaps of skin consisting

of subcutaneous tissue were freed from abdominal wall tissue from the sternum

to the pubic symphysis and displaced laterally. Subcutaneous abdominal fat was

dissected free from the skin between the xiphoid process and the rostral border

of the pelvic girdle. The abdominal cavity was opened and the gonadal and

mesenteric fat deposits were removed and taken together as visceral abdominal

fat. Retroperitoneal fat was then dissected free from the underlying musculature

and kidneys. The weight of each dissected fat pad was recorded.

Fasting Plasma Measurements

Another subset of the previously run, 15 week-old rats [COP (n = 20),

COP.DA(chr 3) (n = 20), and COP.DA(chr 16) (n = 15)] were fasted by food and

water deprivation for 16 hours (8 p.m. to 12 p.m.). Rats were weighed and

anesthetized with sodium pentobarbital (50 mg/kg body weight), the thoracic

cavity opened, and whole blood was collected via right ventricular puncture as

described above. Blood glucose was measured using an Accu-Check

Advantage blood glucose monitor and the appropriate test-strips (Roche

Diagnostics, Indianapolis, IN).

Following collection, blood samples were centrifuged at 2800xg for 15

minutes at 4°C. Plasma was then aliquotted and stored at -80°C until assayed.

Plasma triglyceride and non-esterified fatty acid values from each rat were assayed using the same aliquot on the same day. Plasma insulin was assayed separately using different aliquots of only the highest quality samples to avoid potential interference with the ELISA protocol.

Triglycerides: Plasma triglycerides were assayed using a kit [Triglyceride

(GPO) Reagent Set, Pointe Scientific; Lincoln Park, MI] with modifications for use

with 96-well microtiter plates. Briefly, 3 μl of plasma or diluted standards was

added in triplicate to wells, followed by the addition of 300 μl of reagent to each

well, the plate was covered and incubated at 37°C for 5 minutes. Lipase present

in the reaction mixture hydrolyzes triglycerides in each well to glycerol and free

fatty acids. Glycerol kinase, also present in the reaction mixture, then phosphorylates the glycerol, producing glycerol-3-phosphate. Finally,

glycerophosphate oxidase reacts with glycerol-3-phosphate to produce hydrogen

peroxide, which then reacts with 4-aminoantipyrine and 3-hydroxy-2,4,6-

tribomobenzoic acid (TBHB), yielding a red color. The color intensity in each well

is proportional to the concentration of triglycerides in the samples. Absorbance of

the colorimetric reaction was measured at 540 nm on a Versamax tunable

microplate reader using SoftMax Pro 4.7.1 analysis software (Molecular Devices,

Sunnyvale, CA).

Fatty Acids: Plasma non-esterified fatty acids were assayed using a kit

(NEFA C, Wako Chemicals USA; Richmond, VA), also with modifications for use

with 96-well microtiter plates. Plasma or diluted standards (5 μl) were added in

triplicate to wells and assayed according to manufacturer’s specifications at

1/10th scale. Acyl-CoA synthetase in the reaction mixture attaches acyl groups to

the fatty acids in the plasma samples. The resulting acyl-CoA molecules are then oxidized by acyl-CoA oxidase in the reaction mixture, and hydrogen peroxide is produced in the process. The hydrogen peroxide then reacts with 3- methy-N-ethyl-N(β-hydroxyethyl)-aniline (MEFA) and 4-aminoantipyrine to form a

purple color. Similar to the triglycerides, color intensity is proportional to the concentration of free fatty acids in each sample. Absorbance of the colorimetric reaction was measured at 550 nm on a Versamax tunable microplate reader using SoftMax Pro 4.7.1 analysis software (Molecular Devices).

Insulin: Plasma insulin was assayed using a kit (Rat Insulin ELISA Kit, Crystal

Chem Inc.; Downers Grove, IL) according to the manufacturer’s specifications, including quality of the plasma samples.

Statistical Analysis

Statistical analyses were performed using SPSS 13.0 for Windows. A

Shapiro-Wilk test was used to determine whether data were normally distributed followed by a Levene test for homogeneity of variance. Normally distributed data were analyzed by a one-way ANOVA to determine the overall level of significance. Data showing overall significance were further analyzed using a

Dunnett’s-t post-hoc test to determine inter-strain significance, comparing all groups to COP when homogeneity of variance was observed. When homogeneity of variance was not observed (Levene statistic, P≤0.05), the

Dunnett’s C post-hoc test was used to determine whether inter-strain differences were significant.

Data not normally distributed were transformed and the statistical procedures described above followed. Data still having a non-normal distribution had extreme outliers removed using the boxplot method on untransformed data

followed by analysis using the statistical procedures described above for

unaltered data. Data continuing to have a non-normal distribution were

subjected to non-parametric Kruskal-Wallis tests to determine the overall level of

significance, followed by a Mann-Whitney U pair-wise comparison test if

significant differences were observed. P<0.05 was selected as the criterion for

statistical significance.

Construction and Characterization of the DA.COP(chr 16) Consomic Strain

The construction and characterization of the DA.COP(chr 16) consomic strain

was essentially identical to that of the COP.DA(chr 16) consomic strain with

relatively few exceptions. To avoid a repetitive discourse of all the information

presented above, only the exceptions will be highlighted in this section.

Congenic Breeding Paradigm

The chromosomal interval on rat chromosome 16 containing the putative high capacity ARC QTL alleles from COP rats was transferred into the DA genetic background, resulting in the DA.COP(D16Rat12–D16Rat90) consomic strain.

Similar to the COP.DA consomic strain, the identity of this DA.COP consomic strain will be hereby abbreviated to DA.COP(chr 16). The breeding paradigm was as follows: male F1 rats, bred by crossing male DA rats with female COP

rats, were backcrossed to female DA rats. Male progeny heterozygous for the

RNO16 ARC QTL-containing region and containing the fewest number of COP

alleles in the remainder of the genome were selected for backcrossing to

subsequent female DA rats. A total of four backcross generations were required

to breed the DA.COP(chr 16) consomic strain. Continuous brother-sister mating

was again used to maintain this particular consomic strain.

Genotyping

Genotyping was performed exactly as described for the COP.DA(chr 16) consomic strain.

ARC Phenotype

At 10-weeks of age, male COP (n = 39), DA.COP(chr 16) (n = 28), and DA (n

= 36) underwent the same two-week exercise protocol and recovery period as

described for the COP.DA congenic strains. They were then divided into fasting

and ad libitum-fed subsets and used for further experiments to measure body

and organ weights (ad libitum-fed animals only), as well as blood chemistries

(fasted only), when they were 15-weeks of age.

Ad Libitum-Fed Organ Weights

An ad libitum-fed subset of the previously run, 15-week-old rats [COP (n =

18), DA.COP(chr 16) (n = 17), and DA (n = 20)] was used to obtain organ

weights. Because there was some overlap in the testing dates for the COP.DA

and DA.COP congenic strains, the performance of the parental strains in this overlap period were used in both the COP.DA and DA.COP strain analyses.

This was done to normalize running performance to testing dates, as testing for all congenic strains was spread across an 8-month time period, during which significant seasonal variation in performance was observed in the DA rats.

However, body weights and organ weight measurements remained separated from the COP.DA congenic strain measurements.

Fasting Plasma Measurements

Another set of 15 week-old rats [DA.COP(chr 16) (n=11) and DA (n=12)] were fasted by food and water deprivation for 16 hours (8 p.m. to 12 p.m.). All procedures were performed exactly as described for the fasting COP.DA congenic strain measurements. No fasting insulin measurements were obtained.

Statistical Analysis

Statistical analyses were performed using SPSS 13.0 for Windows exactly as described for the COP.DA congenic strains above except that all data were compared back to DA rats instead of COP rats.

In Vivo Cardiac Function

Measurement of Cardiac Function

All cardiovascular measurements were conducted in a blinded manner;

computer files were saved using a three-digit code, analyzed, and then organized

according to strain. Rats [10 each from COP, DA, COP.DA(chr 16), and

DA.COP(chr 16)] were anesthetized with sodium pentobarbital (50 mg/kg, i.p.)

and the tail vein was cannulated with PE-10 tubing for dobutamine infusion

(Naumova et al., 2003; Jiang et al., 2005). A microtip pressure-volume catheter

(SPR-838; Millar Instruments, Houston, TX) was inserted through the right

common carotid artery and advanced into the left ventricle (LV). After a 10-

minute stabilization period, baseline LV functional data were continuously

recorded using the MPVS-400 pressure-volume conductance system (Millar

Instruments, Houston, TX) at a sampling rate of 1000/sec for 10 minutes.

Dobutamine (a β-1 adrenergic agonist) infusion (10µg/kg/min) was initiated immediately following baseline data collection, during which time LV functional data were continuously recorded. Ten minutes of dobutamine infusion was

followed by a 15-minute washout period. At the conclusion of each experiment,

10 µl of a 15% saline solution was injected into the inferior vena cava for

determination of parallel conductance volume (Vp), i.e., the volume of the

surrounding myocardium. Relative volume units, as a function of conductance,

were converted to true volume units via calibration procedures using heparinized

blood. True Vp was subsequently calculated by PVAN3.5 analysis software

(Millar Instruments, Houston, TX) to correct for the cardiac mass volume as

previously described in detail (Feldman et al., 2000).

Analysis of Cardiac Function Variables

Under the conditions of baseline and dobutamine infusion, PVAN3.5 software

calculated the following parameters: heart rate (HR), stroke volume (SV),

cardiac output (CO), end systolic pressure (ESP) and volume (ESV), maximum

left ventricular pressure (LVPmax), maximum rate of left ventricular pressure

increase during systole (+dP/dt), end diastolic pressure (EDP) and volume

(EDV), maximum rate of left ventricular pressure decrease during diastole

(-dP/dt), left ventricular relaxation time constant (tau), ejection fraction (EF),

stroke work (SW), arterial elastance (Ea), maximal power generated during the

cardiac cycle (PWRmax), preload-adjusted maximal power (PAMP), and volume at maximum rate of pressure increase during systole (V@dP/dtmax). Stroke

volume, stroke work, and cardiac output were normalized to body weight (SVI-

bw, SWI-bw, and CI-bw, respectively) and heart weight (SVI-hw, SWI-hw, and

CI-hw, respectively), and the contractility index [(+dP/dt)/P@dP/dtmax] was calculated. Body weight and heart weight-adjusted values were calculated because previous results generated for cardiac output in the isolated working heart were reported as a function of heart weight (Barbato et al., 1998), because heart weight and body weight indices can be used as indicators of cardiac and

whole animal perfusion, respectively, and to normalize for cardiac size

differences.

Following collection of left ventricular cardiac function data, the conductance

catheter was removed from rats and hearts were resected, blotted dry, and

weighed. Left ventricles were prepared as previously described for the congenic

strains, wrapped in aluminum foil, frozen in liquid nitrogen, and stored at -80°C.

Statistical Analysis

Statistical analyses were performed using SPSS 13.0 for Windows. A

Shapiro-Wilk test was used to determine whether data were normally distributed

followed by a Levene test for homogeneity of variance. Normally distributed data

were analyzed using a priori contrast coefficients for planned comparisons between groups. The primary comparison made was between COP and DA rats at baseline to confirm previous results (Barbato et al., 1998; Koch et al., 1999;

Chen et al., 2001). The COP rats were then compared to the COP.DA(chr 16) strain and DA rats were compared to DA.COP(chr 16) strain to determine if the transfer of chromosome 16 would affect cardiac function as well as ARC. Inter- consomic, DA versus COP.DA(chr 16), and COP versus DA.COP(chr 16) comparisons were not made. Data not normally distributed were analyzed with

Mann-Whitney U pair-wise comparison tests. Data from the dobutamine infusion were analyzed as the percent change over baseline for all parameters and

subjected to the same statistical analysis as baseline data. P≤0.05 was selected as the criterion for statistical significance.

Global Gene Expression Analysis in COP and DA Left Ventricles

Phenotyping and RNA Preparation

Aerobic running capacity (ARC) was estimated in male COP (n=4),

F1(COPxDA) (n=4), and DA (n=4) rats by tests of treadmill running capacity as described above for the congenic strains. Rats were killed by pentobarbital overdose at 15-weeks of age and their body and heart weights measured. Left ventricles were collected as described above and were processed immediately for total RNA isolation as described (Lee et al., 2003). Briefly, left ventricles were homogenized in a guanidine thiocyanate/phenol solution (Ultraspec, Biotecx

Laboratories, Houston, TX). Total RNA was extracted with chloroform and isolated using RNA Tack Resin (Biotecx Laboratories). RNA was further purified by absorption to the RNA binding matrix of a spin column and ethanol precipitation (RNeasy, Qiagen, Valencia, CA). Left ventricular RNA integrity was assessed by electrophoretic size-fractionation on 1% agarose gels under denaturing conditions.

Gene Expression Profiling

Oligonucleotide microarray analysis was performed using the Affymetrix Rat

Genome U34 array set (GeneChips U34A, U34B and U34C totaling 26,379 probe sets; approximately 7,000 known genes and 1,000 expressed sequence tag (EST) clusters for U34A and 8000 ESTs each for B and C chips) according to the manufacturer’s protocol. Double-stranded cDNA was synthesized from 15 μg total RNA from each rat (four rats per strain) with reverse transcriptase

(SuperScript II, InVitrogen, Carlsbad, CA) and T7-(dT)24 as the oligonucleotide primer. Biotinylated cRNA was synthesized using the Enzo Bioarray High Yield

RNA Transcript Labeling Kit (Enzo Diagnostics, Farmingdale, NY) and cRNAs were purified with columns (RNeasy mini kit, Qiagen). Each cRNA sample was fragmented according to the protocol in the Affymetrix GeneChip Expression

Analysis manual and quality was assessed by hybridization of 5 μg to a test chip

(Test3 Chip, Affymetrix). Fragmented cRNA (5 μg) was then hybridized to each of a set of three rat GeneChips (U34A, B, and C; Affymetrix). Hybridization, washing, staining with streptavidin-phycoerythrin, and scanning were performed at the University of Toledo Health Science Campus Genomics Core Facility according to the manufacturer's (Affymetrix) instructions. Transcription and hybridization were validated using bacterial sequences as an external control, as well as several “housekeeping” genes as internal controls.

Identification of Differentially Expressed Probe Sets

Absolute and comparative analyses of data from COP and DA left ventricles were conducted using the default settings of the Affymetrix software (Micro Array

Suite, MAS-5.0) (Liu et al., 2002). Microarray images were scaled to an average hybridization intensity of 150 to normalize signals between individual chips.

These data were deposited into the Gene Expression Omnibus (GEO) database

(http://www.ncbi.nlm.nih.gov/projects/geo/) as series GSE1795. Data from four different DA rats were compared to data from four different COP rats (i.e., a 4x4 matrix comparison) using Data Mining Tool (DMT 3.0, Affymetrix) software.

Student’s t-test, with P<0.05 as the criterion for significance, was used to initially identify probe sets showing significant strain-differences in expression where absolute expression values (signals) from four biological replicates/strain were compared between two strains. Probe sets with signal log ratios >0.38 or <-0.38

(i.e., a 1.3-fold change) were then selected. Not included were probe sets whose hybridization signals were below background levels, i.e., called “absent” for three of four rats in both strains. Reproducibility was assured by choosing probe sets that were differentially expressed in >50% of the 4x4 matrix comparisons, selecting for consistency of differential calls and whose signals among the four rats/strain had a standard deviation <25% of the mean.

Network Analysis of Differentially Expressed Genes and Interactions

Ingenuity Pathways Analysis (IPA; Ingenuity Systems, Mountain View, CA)

was used to identify biological networks through which the differentially

expressed sequences might operate (www.Ingenuity.com). This application was

used to search a proprietary database for interactions between our set of

differentially expressed left ventricular genes and all other genes stored in the

database to generate sets of networks, each containing a maximum of 35

genes/proteins. The IPA software computed a score for each network according

to the fit of the set of supplied differentially expressed genes, termed “focus

genes.” A score greater than 2 indicated, with at least 99% confidence, that a

particular network was not generated by chance alone by calculating the

probability of finding one or more focus genes in a set of genes selected

randomly from the global network found in the IPA database. Networks with

scores ≥2 were ranked from highest to lowest and probable functions were assigned according to the assortment of genes in the network.

Ingenuity Pathways Analysis was also used in conjunction with the Rat

Genome Database (http://rgd.mcw.edu) and NCBI (www.ncbi.nlm.nih.gov) to

determine the chromosomal location of genes within molecular networks that

contained the differentially expressed genes. Networks were organized

according to probable functions for the genes involved. The database was also

searched for genes within the 1-LOD confidence intervals of the two interacting

loci (Ways et al., 2002) to identify genes that interacted with differentially

expressed genes found within those intervals.

In Silico Gene Mapping

Chromosomal locations of Affymetrix probe sets were determined by comparing their sequences to the rat genome sequence database (builds 2.1 and

3.4) at the NCBI web site (http://www.ncbi.nlm.nih.gov) using the BLAST program. The BLAST program was also used to identify orthologous locations in the mouse genome (mouse genome sequence database build 30, NCBI) for probe sets that did not map to the rat genome. Comparative mapping information for orthologous regions of the mouse and rat genomes was also obtained from the Ensembl (http://www.ensembl.org), rat genome database

(http://rgd.mcw.edu), and mouse genome informatics

(http://www.informatics.jax.org) websites. Chromosomal locations of known genes were obtained from the rat genome database (http://rgd.mcw.edu).

Protein Expression of PIK3R1

The most differentially expressed sequence between COP and DA left

ventricles (Lee et al., 2005) represented the Pik3r1 gene. Pik3r1 codes for a regulatory subunit (known as p85α) that dimerizes with a catalytic subunit (p110) to form a heterodimer known as phosphatidylinositol 3-kinase (PI 3-kinase) which

is involved in many intracellular signaling pathways (Shibasaki et al., 1991; Inukai et al., 1997). The Pik3r1 gene can generate three isoforms (p85α, p55α, and p50α) that confer variability in the activity of the enzyme (Inukai et al., 1997).

The methods described below outline the attempt to confirm the mRNA expression differences of this gene at the protein level and determine if mRNA expression differences might be the result of differential expression of the various isoforms.

Sample Preparation

Fifteen-week old male COP (n = 6) and DA (n = 6) rats were killed via pentobarbital overdose and their hearts were removed. Left ventricles were isolated and stored as described above until analyzed. Thawed tissue was homogenized in 3x volume of cold extraction buffer [150 mM NaCl, 50 mM

HEPES (pH 7.6), 0.02% Na-azide, 1mM Na3VO4, 1% Triton X-100, and 1 protease inhibitor cocktail tablet (Complete mini tablets, Roche Diagnostics,

Indianapolis, IN)]. Insoluble material was removed by centrifugation at 16,000xg and 4°C for 30 minutes and the supernatants were aliquotted into 1.5 ml tubes.

Total protein concentration was determined using a BIO-RAD protein assay kit II with bovine serum albumin as a standard (BIO-RAD, Hercules, CA). Protein samples (100 μg) were boiled for 5 minutes at 100°C in an equal volume of 2X

Laemmli sample buffer with 5% 2-mercaptoethanol for denaturation.

Western Blot Analysis

Denatured proteins were loaded into a 10% SDS-PAGE mini-gel (BIO-RAD,

Hercules, CA). Samples were run at a constant 15 mA current until the bromophenol blue front exited the gel (~2.5-3 hours), at which point the current was stopped and the gel removed from the electrophoresis chamber. Separated proteins were transferred to a 0.45 μm PVDF membrane (Millipore, Billerica, MA) via a semi-dry electrophoretic tranfer cell (BIO-RAD, Hercules, CA) using a triple buffer system. The buffers consisted of one cathode buffer (25 mM Tris, 40 mM aminohexanoic acid, pH 9.4) in which the gel and filter paper were soaked for 10 minutes and two anode buffers [anode buffer I (300 mM Tris, pH 10.4); anode buffer II (25 mM Tris, pH 10.4)] in which the filter paper and PVDF membrane were soaked, respectively, for 10 minutes. The transfer was run at a constant

20V for 1 hour, after which the gel was stained with Coomassie blue to monitor efficiency of transfer and the membrane was placed in a 5% non-fat dry milk

(NFDM) plus 1X tris-buffered saline and 1% Tween-20 (TBST) and gently rocked at 4°C to block all non-specific sites for antibody binding. The membrane was washed in cold TBST and placed in the 5% NFDM/TBST solution with the appropriate primary antibody and gently rocked overnight (16-20 hours) at 4°C.

The membrane was washed again and gently rocked in 5% NFDM/TBST and the appropriate secondary antibody conjugated to horseradish peroxidase for one hour at room temperature. After a final wash the membrane was placed in ECL detection reagent (Pierce, Rockford, IL), removed, excess moisture blotted off,

and developed on autoradiography film. After detection, the membrane was stripped of all antibodies and re-probed to detect the normalizing protein

(GAPDH).

Antibodies

Antibodies were purchased from Santa Cruz Biotechnology (Santa Cruz, CA).

Anti-PI 3-kinase p85α (sc-1637) was the primary mouse monoclonal antibody used at a 1:200 dilution to detect the different isoforms of the p85α regulatory subunit of PI 3-kinase (p85α, p55α, and p50α) (Inukai et al., 1997). Anti-GAPDH

(sc-32233) was the primary mouse monoclonal antibody used at a dilution of

1:5000 to detect the “housekeeping” gene product GAPDH for sample normalization. Goat anti-mouse IgG conjugated to horseradish peroxidase for

ECL detection (sc-2005) was the secondary antibody used at a 1:5,000 -

1:10,000 dilution to detect the mouse primary antibodies.

Quantitation and Statistics

Films containing the developed protein bands were scanned into Scion Image

Analysis 4.0.2 (Scion Corp., Frederick, MD) for densitometry analysis. The row of bands containing the protein of interest (p85α, p55α, p50α, or GAPDH) was selected, band density was recorded for each animal, and then COP and DA bands were separately averaged. Overall protein expression for the PI 3-kinase regulatory subunit was taken as the sum of the density readings for each isoform.

All bands were normalized to GAPDH to obtain a relative density. COP values

were compared to DA values using Student’s t-test and P≤0.05 was used as the

criterion for significance.

Gene Sequencing

Several genes identified by the global expression and pathway analyses that

mapped to QTL-containing regions were sequenced to determine if there were

nucleotide sequence polymorphisms in the coding sequence that would lead to amino acid substitutions and altered protein function. Such polymorphisms could

identify these genes as functional candidates for the ARC QTLs identified for

COP and DA rats (Ways et al., 2002).

In Silico Primer Design

The cDNA sequences (accession numbers provided) for seven genes [acyl-

CoA synthetase long-chain family member 1 (Acsl1, NM012810), beta-3

adrenergic receptor (Adrb3, NM013108), insulin receptor substrate 2 (Irs2;

NM003749, AF090738, AF050159, AF087674), lipoprotein lipase (Lpl,

NM012598), and phosphatidylinositol 3-kinase regulatory subunit 2 (Pik3r2,

NM022185) were obtained from NCBI (http://www.ncbi.nlm.nih.gov). Primer

pairs were designed from these sequences using the Oligo-Lite primer design

program for MacOS (Molecular Biology Insights, Cascade, CO) to amplify cDNA

from the left ventricles of COP and DA rats. Since a single primer pair was generally insufficient to amplify an entire cDNA sequence, multiple primer pairs were designed in such a way that there was overlap of each amplified segment to prevent the introduction of sequence gaps. The same primers were used to sequence the cDNA segments (discussed below). If the primers were insufficient to complete the sequencing of cDNA segments, additional primers were designed to create overlap with previously determined sequences and used to complete the sequencing. The overlapping segments were used for quality control as well as for aligning completed sequences.

cDNA Preparation and Sequencing

Total RNA from COP and DA left ventricles was used to prepare cDNA for sequencing, as described above for the gene expression analysis. The PCR amplification of cDNA sequences from the gene-specific primer pairs was performed in a series of 50 μl reaction volumes using the Expand Long Template

PCR System (Roche Diagnostics, Indianapolis, IN) as described by the manufacturer. Products were fractionated on a 1% agarose gel and visualized using ethidium bromide staining and ultraviolet illumination to confirm product sizes. Bands were purified using the QIAquick Gel Extraction Kit (Qiagen,

Chatsworth, CA). Following purification, PCR products amplified from the same primer pair were pooled and 5 μl was used for fractionation on a 1% agarose gel with a ΦX174 DNA-HaeIII digest ladder (New England Biolabs, Ipswich, MA) for

which the relative abundance of each band was known and used to estimate the concentration of the pooled PCR product via densitometric analysis (Scion Image

Analysis 4.0.2, Scion Corp., Frederick, MD). Purified PCR products and the appropriate primers were prepared and shipped to MWG Biotech (High Point,

NC) for sequencing according to the specifications set forth by the company.

Sequence Analysis

Sequences were received electronically then analyzed, assembled, and compared using Sequencher DNA Sequence Analysis Software for MacOS

(Gene Codes Corp., Ann Arbor, MI). All received sequences were compared against provided chromatographic profiles for error checking. Sequence fragments were first assembled using the published sequences obtained from

NCBI as a guide. The assembled contigs from COP and DA rat sequences were then aligned with each other to search for polymorphisms (non-synonymous nucleotide substitutions, insertions, or deletions) that may change the amino acid composition of the protein and potentially lead to functional differences in the translated protein.

Substrate Utilization

Exercise Test and Study Design

Twelve male COP rats and 12 male DA rats were divided into control and run groups (n = 6 rats per group per strain). Ten-week old rats in the run group were introduced to the treadmill (Model Exer-4, Columbus Instruments, Columbus,

Ohio) as described above. The following week, each rat was subjected to a single evaluation of running capacity using the previously described protocol

(Barbato et al., 1998; Ways et al., 2002). Following the running capacity test, the current to the grid was stopped and the rat removed from the treadmill, weighed, and immediately anesthetized with sodium pentobarbital (50 mg/kg, i.p.). Control rats were subjected to all the same conditions as the exercised rats except that treadmills remained stationary.

Tissue Collection and Processing

A midline incision was made in each rat from the abdomen through the lower portion of the thoracic cavity. Plasma was obtained as described above. Hearts were removed and left ventricles were isolated as described above and immediately frozen between clamps cooled in liquid nitrogen, wrapped in foil, and stored at -80°C. The rats were decapitated immediately following heart removal for another study involving brain tissue (data not discussed). The medial gastrocnemius muscle from the right leg of each rat was dissected free from

surrounding leg muscles, blotted dry, and immediately frozen, wrapped in foil,

and stored at -80°C. A piece of the medial lobe of the liver was cut free from the

rest of the liver, immediately frozen, wrapped in foil and stored at -80°C. All

tissues were ground into powder under liquid nitrogen using a metal mortar and

pestle and stored in labeled 1.5 ml microcentrifuge tubes at -80°C until assayed.

Plasma Substrate Measurements

Aliquots of plasma samples were thawed and used to determine FFA,

glucose, or lactate concentrations. Samples were analyzed on separate days

using separate aliquots. The FFA concentrations were measured as described

above. Glucose and lactate concentrations were ascertained via enzyme-based

fluorometric assays based on measuring the production of NADH.

Plasma was diluted 1:9 with distilled water before proceeding with the glucose

assay. Diluted plasma or glucose standard (5 μl) was added to 1 ml of glucose

reagent (50 mM Tris ph 8.1, 0.04% bovine serum albumin, 50 μM NADP, 0.5 mM dithiothreitol, 0.3 mM ATP, 1 mM magnesium chloride) in 10x75 mm borosilicate test tubes. Relative fluorescence resulting from endogenous NADH in each sample was measured in a Turner Quantech Base Fluorometer to account for background. Glucose-6-phosphate dehydrogenase (Sigma, St. Louis, MO) and hexokinase (Sigma) were added to each tube at concentrations of 0.02 units/ml and 0.17 units/ml, respectively. All tubes were vortexed and incubated at room temperature for 30 minutes before measuring relative fluorescence as a result of

enzymatic NADH production. Glucose concentrations were calculated by

subtracting the relative background fluorescence from the second reading and

comparing the results to the relative fluorescence of NADH generated from

known glucose concentrations in a standard curve.

Plasma was diluted 1:7 with distilled water before proceeding with the lactate

assay. Diluted plasma (3 μl) or lactate standard (5 μl) was added to 1 ml of

lactate reagent (100 mM 2-amino-2-methyl-1-propanolol pH 10, 100 mM

hydrazine hydrate, 0.2 mM NAD) in 10x75 mm borosilicate test tubes. Relative

fluorescence as a result of endogenous NADH present was measured in a

Turner Quantech Base Fluorometer to account for background. Lactate dehydrogenase (Sigma) was added to each tube at a concentration of 8 units/ml.

All tubes were vortexed and incubated at room temperature for 4 hours before taking a second set of fluorometric readings. Lactate concentrations were calculated by subtracting the relative background fluorescence from the second reading and comparing the results to the relative fluorescence of known lactate concentrations in a standard curve.

Tissue Glycogen

Powdered tissue (10-40 mg) was disrupted via sonication (Fisher Sonic

Dismembrator, Model 100) in 1 ml of 0.3 M perchloric acid followed by homogenization in a glass conical mortar with a motor-driven pestle.

Homogenates were transferred to 1.5 ml microcentrifuge tubes and centrifuged

for 4 minutes at 16,000xg and 4°C. Supernatants were transferred to 1.5 ml

tubes and either assayed immediately or stored at -80°C. Liver samples were

diluted 1:5 while muscle and heart samples were used undiluted.

In 10x75 mm borosilicate test tubes, 10 μl of sample was added to 500 μl of a sodium acetate buffer (50 mM sodium acetate pH 5.5, 0.02% bovine serum albumin) both with and without 0.14 units/μl of amyloglucosidase (Roche

Diagnostics, Indianapolis, IN). Amyloglucosidase hydrolyzes glycogen into single units of glucose; therefore the samples lacking the enzyme served as controls for endogenous glucose that if not subtracted out would result in deceptively high

levels of glucose units produced as a result of glycogen hydrolysis. Tubes were

vortexed and placed in a shaking incubator at 25°C for 3 hours. After the

addition of 500 μl of glucose reagent (100 mM Tris ph 8.1, 0.04% bovine serum albumin, 40 μM NADP, 0.5 mM dithiothreitol, 1 mM ATP, 2 mM magnesium chloride, 0.58 units/ml glucose-6-phosphate dehydrogenase, 1.15 units/ml hexokinase), samples were vortexed and incubated at room temperature for 30 minutes. Relative fluorescence resulting from enzymatic NADH production was measured in all tubes. Net fluorescence was calculated as the difference between fluorescence values of tubes containing amyloglucosidase and tubes lacking the enzyme. Glucose concentrations were calculated as described above for plasma glucose then normalized to the wet weight of the muscle sample used.

Tissue Triglycerides

Powdered tissue (10-40 mg) was digested overnight (16-20 hours) in 400 μl of tetraethylammonium hydroxide (TEAH) at 50°C while shaking. The following day 175 μl of 3M perchloric acid was added to each tube. The addition of perchloric acid produced a thick, white precipitate so each tube was vortexed thoroughly then centrifuged at 3000xg for 10 minutes at room temperature.

Supernatant (~300 μl) was transferred to new 1.5 ml microcentrifuge tubes and

200 μl of 2 M potassium bicarbonate was added to neutralize the solution. The concentration of triglycerides in each tissue sample was then determined as described above for plasma triglycerides then normalized to the wet weight of the muscle sample used.

Statistical Analysis

A two-way analysis of variance was employed to determine whether the

difference in substrate concentrations between exercised rats and control rats

was dependent on strain. The significance criterion for both main and interaction

effects was P≤0.05. Since there were only two groups for running performance

and vertical work performed, a Student’s t-test was used to assess statistical

significance, again set at P≤0.05.

RESULTS

Congenic Strains

Congenic strains were constructed via marker-assisted breeding where the regions on RNO3 and RNO16 containing ARC QTLs from DA rats were transferred onto an otherwise uniform background of COP alleles, resulting in the

COP.DA(chr 3) and COP.DA(chr 16) congenic strains, respectively. This

approach was also used to transfer RNO16 from COP rats onto an otherwise

uniform background of DA alleles resulting in the reciprocal congenic strain

DA.COP(chr 16). The RNO3 congenic region was a 118-135.2 Mb interval with

the region known to contain only DA alleles extending from D3Rat233 to

D3Mgh14. The RNO16 congenic region was 85.3-90.2 Mb, with the region known to contain only donor-strain alleles extending from D16Rat12 to

D16Rat90. As the RNO16 congenic regions encompass 95-100% of RNO16,

COP.DA(chr 16) and DA.COP(chr 16) can be considered consomic strains

(Nadeau et al., 2000). Physical maps of the transferred chromosomal regions for the RNO16 consomic strains and COP.DA(chr 3) congenic strain are shown in

Figure 8.

Figure 8. Physical maps of the congenic regions carried by COP.DA(chr 3) and the two RNO16 consomic strains. Solid bars represent transferred chromosomal regions from the donor strain into the congenic and consomic strains, open bars represent chromosomal regions where crossovers have occurred marking the boundaries of the congenic region, and solid lines represent the length of the entire chromosome. Markers located at or near QTL peaks, as described in Ways et al. (2002), are underlined, bolded, and indicated by arrows. Map distances are expressed in megabasepairs (Mb).

COP.DA Performance Data

Running results and all measurements for the DA.COP(chr 16) consomic strain will be presented separately from the present discussion of COP.DA congenic strains as they were run at different times with different sets of parental control strains. No significant correlation was observed between strain and the day when the best distances run to exhaustion occurred over the five-day period of the trial (r=0.084, P=0.35). No significant strain differences were observed for the mean day of occurrence for the best distance run to exhaustion (P=0.94; day

2.56 for COP.DA(chr 3), day 2.63 for COP.DA(chr 16), day 2.81 for COP, and day 3.30 for DA rats). For the parental rat strains run with the COP.DA congenic strains, a 412.7 m greater best distance run to exhaustion for DA rats (984.6 ±

38.5 m) compared to COP rats (571.9 ± 27.5 m, Figure 9) was observed, consistent with previous results (Barbato et al., 1998; Koch et al., 1999; Ways et al., 2002). COP.DA(chr 16) rats had a significantly greater performance (696.7 ±

38.2 m, 21.8%) compared to COP rats (P=0.03; Figure 9). COP.DA(chr 3) rats, however, had a performance (643.6 ± 40.9 m) greater than that of COP rats

(12.5%), but the results were not statistically significant (P=0.21).

The DA and COP.DA(chr 16) rats both had a significantly longer average distance run to exhaustion over the 5 days of testing (755.6 ± 35.0 m, P<0.001 and 517.5 ± 27.6 m, P=0.012, respectively) compared to COP rats (406.2 ± 17.2 m). Again, while COP.DA(chr 3) rats ran for a greater average distance over

Figure 9. Mean best distance run to exhaustion for COP.DA congenic strains and parental COP and DA rat strains. Performances were as follows: DA = 985 ± 39 m (n = 28), COP.DA(chr 16) = 696 ± 38 m (n = 24), COP.DA(chr 3) = 643 ± 41 m (n = 39), COP = 571 ± 28 m (n = 36). Values are given as the mean ± the standard error of the mean (SEM). DA rats showed significantly greater performance compared to each of the other strains (P<0.0001). * significantly greater than the performance of COP (P< 0.05).

the five days of testing (463.7 ± 30.6 m) compared to COP rats, the difference

was not statistically significant (P=0.14).

COP.DA Organ Weight Measurements

Significant organ weight differences were observed between COP and DA

rats for all organs measured, except for the pancreas and kidneys. No significant

differences in heart or left ventricular weights were observed between either of the congenic strains compared with COP rats (Table II). COP.DA(chr 16) and

COP rats did not show significant differences in body weight (Table II). However,

COP.DA(chr 3) rats had significantly lower body weights compared to COP rats

(252.6 ± 4.6 g versus 271.0 ± 5.5 g, P=0.014; Table II), but comparable body weights to DA rats (257.0 ± 3.5 g; Table II). Heart and left ventricular weights 88

adjusted for body weight were significantly greater for COP.DA(chr 3) compared to COP rats (2.99 ± 0.06 mg/g versus 2.76 ± 0.03 mg/g and 1.96 ± 0.05 mg/g versus 1.75 ± 0.02 mg/g, respectively; Table II). COP.DA(chr 3) rats had significantly lower pancreas weights compared to COP rats (0.52 ± 0.02 g versus

0.62 ± 0.03 g, (P<0.05), although body weight adjusted pancreas weights were not significantly different (Table II). In addition, COP.DA(chr 3) rats had significantly lower total kidney weights compared to COP rats (1.61 ± 0.05 g versus 1.84 ± 0.05 g, P=0.002; Table II). No significant liver weight differences were observed between either COP.DA(chr 16) or COP.DA(chr 3) congenic rats compared to COP rats (Table II).

The DA rats had significantly more total abdominal fat compared to COP rats

(7.93 ± 0.34 g versus 4.95 ± 0.22 g, P<0.001; Figure 10). Separating total abdominal fat into retroperitoneal, subcutaneous, and visceral components revealed that DA rats had significantly more fat for each component compared to

COP rats (Figure10). COP.DA(chr 16) consomic rats also had significantly more total abdominal fat (6.47 ± 0.24 g versus 4.95 ± 0.20 g, P=0.001; Figure 10) and subcutaneous abdominal fat (2.25 ± 0.09 g versus1.27 ± 0.11 g, P<0.05; Figure

10), compared to COP rats. Most of the difference in total abdominal fat between

COP.DA(chr 16) and COP rats resulted from the large difference in subcutaneous abdominal fat (64%). Overall, a significant correlation existed between subcutaneous fat and best distance run (r=0.32, P=0.02) where the

89 Table II – Anatomical Comparisons Between COP.DA Congenic and Inbred Strains

COP COP.DA(chr 16) COP.DA(chr 3) DA Variable (n = 15) (n = 8) (n = 18) (n = 12)

Body Weight (BW, g) 271.0 ± 5.5 275.6 ± 4.6 252.6 ± 4.6* 257.0 ± 3.5*

Heart Weight (g) 0.76 ± 0.01 0.73 ± 0.02 0.75 ± 0.02 0.85 ± 0.02*

BW-Adjusted Heart Weight 2.76 ± 0.03 2.65 ± 0.04 2.99 ± 0.06* 3.29 ± 0.05* (mg/g)#

Left Ventricle Weight (g) 0.49 ± 0.01 0.48 ± 0.01 0.49 ± 0.02 0.55 ± 0.01*

BW-Adjusted Left Ventricle 1.75 ± 0.02 1.75 ± 0.02 1.96 ± 0.05* 2.13 ± 0.04* Weight (mg/g) #

Pancreas Weight (g) 0.62 ± 0.03 0.49 ± 0.04 0.52 ± 0.02* 0.53 ± 0.04

BW-Adjusted Pancreas Weight 2.29 ± 0.12 1.79 ± 0.12 2.05 ± 0.09 2.09 ± 0.19 (mg/g)

Liver Weight (g) 9.55 ± 0.16 9.40 ± 0.16 9.24 ± 0.31 8.02 ± 0.19*

BW-Adjusted Liver Weight † 35.3 ± 0.45 34.2 ± 0.67 36.5 ± 0.88 31.2 ± 0.51* (mg/g)

Total Kidney Weight (g) 1.84 ± 0.05 1.79 ± 0.05 1.61 ± 0.05* 1.99 ± 0.05

BW-Adjusted Total Kidney 6.80 ± 0.14 6.49 ± 0.18 6.37 ± 0.11* 7.73 ± 0.13 Weight (mg/g)

BW-adjusted values are organ weights that were normalized to body weight. # variables with extreme outliers removed from COP (n = 14). † adjusted liver weight transformed using 1/(liver weight/body weight). * significantly different from the COP value (P< 0.05).

more subcutaneous fat a particular rat had, the better the running performance was.

Figure 10. Abdominal fat weight for COP.DA congenic strains and the COP and DA strains. Abdominal fat, as its retroperitoneal, subcutaneous, and visceral compartments, was dissected from COP (n=15), DA (n=12), COP.DA(chr 16) (n=8), and COP.DA(chr 3) (n=18) rats and weighed. Values are given as means ± SEM. Total abdominal fat is the sum of the retroperitoneal, subcutaneous, and visceral fat pad components. * significantly greater than COP fat weights (P<0.05). ** significantly greater than COP fat weights (P≤0.001).

COP.DA Plasma Measurements

Fed and fasted plasma triglycerides, FFA, and glucose (fasted only) concentrations for both congenic strains were not significantly different from COP rats, with the exception of lower fasted plasma triglycerides concentrations observed in COP.DA(chr 16) rats (P<0.05, Table III). A significant correlation was also observed between fasting plasma triglycerides and best distance run

(r=0.27, P=0.04) among the three strains tested. Both COP.DA(chr 3) and

COP.DA(chr 16) rats did not have significantly different fasted plasma insulin concentrations compared to COP rats (Table III).

Table III - Circulating Metabolic Substrate Concentrations for COP and COP.DA Congenic Rat Strains

Plasma Value COP COP.DA(chr 16) COP.DA(chr 3)

Fed (n = 14) (n = 8) (n = 16)

Triglycerides (mg/dl) 51.71 ± 8.3 39.55 ± 5.02 47.59 ± 6.17

Free Fatty Acids (mEq/l) 0.12 ± 0.01 0.13 ± 0.02 0.10 ± 0.01

Fasted (n = 20) (n = 15) (n = 20)

Triglycerides (mg/dl) 26.03 ± 1.9 20.63 ± 2.8* 30.5 ± 2.5

Free Fatty Acids (mEq/l) 0.23 ± 0.02 0.29 ± 0.03 0.22 ± 0.02

Glucose (mg/dl) 119.2 ± 4.3 118.1 ± 7.6 132.1 ± 5.2

Insulin (ng/ml)‡ 1.12 ± 0.22 1.25 ± 0.29 1.02 ± 0.31 Table III. Plasma glucose and insulin (‡, 10 rats/strain) were only measured in fasted animals. Plasma triglyceride (fed and fasted) and insulin (fasted) concentration values were log transformed to achieve normal distributions. Values are given as the mean ± SEM. * significantly different from COP values (p< 0.05).

DA.COP Performance Data

Running performances between COP and DA rats that were tested alongside the DA.COP(chr 16) consomic strain were comparable to the aforementioned results observed for the COP.DA congenic strains and previous results for the

DA and COP inbred strains (Barbato et al., 1998; Koch et al., 1999; Ways et al.,

2002). No significant correlation was observed between strain and the day when the best distance run to exhaustion occurred over the five-day period of the trial

(r=0.016, P=0.87). No significant strain differences were observed for the mean day of occurrence of the best distance run to exhaustion (P=0.99); day 3.11 for

DA.COP(chr 16), day 3.10 for COP, and day 3.06 for DA rats. The DA rats ran

900.6 ± 37.5 m and COP rats ran 544.7 ± 22.7 m to exhaustion (Figure 11).

Figure 11. Mean best distance run to exhaustion for DA.COP congenic strains and parental COP and DA rats. Performances are as follows: DA = 900.6 ± 38 m (n = 36), DA.COP(chr 16) = 701.5 ± 39 m (n = 28), COP = 544.7 ± 23 m (n = 39). Values are given as the mean ± SEM. * significantly different from DA performance values (P< 0.05).

Transferring COP alleles contained within the interval on RNO16 bound by markers D16Rat12 to D16Rat90 resulted in a 199.1 m (22.1%) reduction in performance of the DA.COP(chr 16) rats (701.5 ± 38.9 m) compared to DA

(P<0.05; Figure 11). Results for the average distance run between COP and DA rats (394.1 ± 14.0 m and 698.8 ± 30.3 m, respectively) were comparable to those described for the COP.DA congenics. DA.COP(chr 16) consomic rats had an average distance run of 518.8 ± 30.8 m which was significantly lower than DA rats (P<0.001), as would be expected.

DA.COP Organ Weight and Plasma Measurements

All organ weights analyzed with the DA.COP(chr 16) consomic rats, were significantly different between COP and DA rats. The body weights of

DA.COP(chr 16) rats did not differ from the body weights of DA rats (Table IV).

Heart weights, but not left ventricular weights, were significantly less in the

DA.COP(chr 16) consomic strain compared to DA rats (Table IV), although the

DA.COP(chr 16) heart weights were much more similar to those of the DA rats

(0.80 ± 0.01 g versus 0.84 ± 0.01 g, respectively; P=0.019) than COP rats (0.67 ±

0.01 g). Heart and left ventricle weights adjusted for body weight did not change this trend (Table IV). The DA.COP(chr 16) rats had greater pancreas and liver weights than DA rats (0.57 ± 0.05 g versus 0.39 ± 0.03 g and 9.49 ± 0.23 g versus 8.65 ±0.25 g, respectively). Adjusting pancreas and liver weights for body

Table IV – Anatomical Comparisons between DA.COP Consomic and Inbred Strains

COP DA.COP(chr 16) DA Variable (n = 18) (n = 17) (n = 20)

Body Weight (g) 270.7 ± 4.5* 262.2 ± 3.2 260.2 ± 2.7

Heart Weight (g) 0.67 ± 0.01* 0.80 ± 0.01* 0.84 ± 0.01 BW-Adjusted Heart 2.49 ± 0.03* 3.06 ± 0.04* 3.24 ± 0.03 Weight (mg/g)

Left Ventricle Weight (g) 0.43 ± 0.01* 0.53 ± 0.01 0.54 ± 0.01

BW-Adjusted Left 1.60 ± 0.03* 2.02 ± 0.03 2.08 ± 0.03 Ventricle Weight (mg/g)

Pancreas Weight (g)† 0.66 ± 0.04* 0.57 ± 0.05* 0.39 ± 0.03

BW-Adjusted Pancreas 2.43 ± 0.15* 2.17 ± 0.18* 1.52 ± 0.11 Weight (mg/g)†

Liver Weight (g) 10.00 ± 0.18* 9.49 ± 0.23* 8.65 ± 0.25

BW-Adjusted Liver # 37.0 ± 0.58* 36.1 ± 0.53* 32.5 ± 0.32 Weight (mg/g)

Total Kidney Weight (g)† 1.74 ± 0.03* 1.99 ± 0.03 2.05 ±0.03

BW-Adjusted Total 6.44 ± 0.07* 7.61 ± 0.09 7.83 ± 0.08 Kidney Weight (mg/g)#

BW-adjusted values are organ weights that were normalized to body weight. # variables with extreme outliers removed from DA BW-adjusted liver weight (n = 17) and DA BW- adjusted total kidney weight (n = 19). † pancreas weight and adjusted pancreas weight transformed using 1/square root of variable; kidney weight transformed using (kidney weight squared). * significantly different from the COP value (P< 0.05).

weights did not change this trend (Table IV). Kidney weights were similar between DA.COP(chr 16) and DA rats.

As with the COP and DA rats evaluated with the COP.DA congenic strains,

the DA rats evaluated with the DA.COP(chr 16) consomic rats also had

significantly more total abdominal fat compared to COP rats (6.14 ± 0.20 g

versus 4.25 ± 0.33 g; P<0.05). When separated into individual components, DA

had significantly more visceral, subcutaneous, and retroperitoneal fat as well

(Figure 12). While the DA.COP(chr 16) strain generally had more fat than the DA

Figure 12. Abdominal fat weight data for the DA.COP(chr 16) consomic, COP, and DA strains. Abdominal fat broken down into its retroperitoneal, subcutaneous, and visceral compartments was dissected from COP (n=18), DA (n=20), and DA.COP(chr 16) (n=17) and weighed. Values are given as means ± SEM. Total abdominal fat is the sum of the retroperitoneal, subcutaneous, and visceral fat pad components. * significantly less than DA fat weights (P<0.05).

rats (total and individual components), the results were not statistically significant

(Figure 12). Fasting metabolic substrate concentrations were not significantly

different between DA and DA.COP(chr 16) consomic rats (Table V).

Table V – Metabolic Substrate Concentrations in Fasted DA and DA.COP(chr 16) Consomic Rats

DA DA.COP(chr 16) Fasted Plasma Value (n = 12) (n = 11)

Triglycerides (mg/dl) 24.42 ± 3.1 31.35 ± 4.0

Free Fatty Acids (mEq/l) 0.18 ± 0.01 0.21 ± 0.02

Glucose (mg/dl) 183.1 ± 3.1 183.7 ± 5.4 Values are given as the fasted mean ± SEM. There were no significant differences observed.

In Vivo Cardiac Performance

Baseline Function

Baseline cardiac performance using an in vivo pressure-volume conductance system was initially compared between COP and DA rats to confirm the differences in cardiac function found previously (Barbato et al., 1998; Koch et al.,

1999; Chen et al., 2001). No significant differences in heart rate were observed between the COP and DA rats (Table VI). Although SV and CO were greater in

DA compared to COP (165 ± 17 μl versus 158 ± 15 μl, 4%; 69.0 ml/min versus

59.4 ± 6.7 ml/min, 16%, respectively), the differences were not statistically significant. Similar observations were made for body weight adjusted SV and CO values (Table VI). Heart weight adjusted SV and CO values, on the other hand, tended to be lower in DA compared to COP rats (Table VI). The COP rat hearts had significantly lower ESP than DA rat hearts (108 ± 13 mmHg versus 136 ± 7 97

mmHg; P=0.03). Measures of contractile performance (+dP/dt, Contractility

Index, V@dP/dtmax, and PWRmax), with the exception of EF and PAMP, all

tended to be greater in DA rats compared to COP (Table VI). The DA rat hearts

also tended to have superior diastolic performance compared to COP rat hearts

(EDP, EDV, -dP/dt, tau; Table VI), although only EDP and tau reached statistical

significance (10.46 ± 0.67 mmHg versus 8.10 ± 0.72 mmHg, P=0.02; 10.82 ±

1.09 msec versus 7.88 ± 0.33 msec, P=0.03).

The DA.COP (chr 16) rats displayed an increased value for tau (Table VI), indicating a reduction in diastolic function as a result of transferring COP RNO16 alleles into a DA genetic background. The DA.COP(chr 16) rat hearts had significantly lower CI values than DA rat hearts (139 ± 4 mmHg2/sec versus 163

± 4 mmHg2/sec; P=0.001). The comparison made between COP.DA(chr 16) and

COP rats revealed that COP.DA(chr 16) tended to have lower CI values,

including an unexpectedly low heart weight normalized cardiac output (Table VI).

In fact, most of the variables for COP.DA(chr 16) in Table VI followed a trend that indicated inferior cardiac function compared to DA rats.

Dobutamine Infusion

Several studies involving dogs, mice, and rats were reviewed to determine

which hemodynamic parameters were significantly and consistently affected by

dobutamine and whether these parameters increased or decreased as a result of

Table VI – Cardiac Function Variables for COP, DA, and Consomic Rats

Hemodynamic COP COP.DA(chr 16) DA DA.COP(chr 16) Variables (n = 8) (n = 8) (n = 10) (n = 10) Measured HR (bpm) 379 ± 29 347 ± 24 416 ± 14 406 ± 11

SV (µL) 158 ± 15 143 ± 11 165 ± 17 154 ± 15

SVI-bw (µL/kg) 616 ± 55 529 ± 35 659 ± 77 541 ± 37

SVI-hw (µl/g)#4 235 ± 21 197 ± 13 192 ± 21 188 ± 20

CO (mL/min) #4 59.4 ± 6.7 49.7 ± 5.6 69.0 ± 7.6 57.3 ± 4.4

CI-bw (µl/min/g) #2,4 233 ± 29 167 ± 10 276 ± 34 217 ± 16 CI-hw (µl/min/mg) 88.3 ± 9.5 62.9 ± 3.7** 80.4 ± 9.6 68.5 ± 4.6 #2,4 ESP (mmHg) #4 108 ± 13 113 ± 9 136 ± 7* 140 ± 2

ESV (µL) 50.2 ± 17.1 63.1 ± 20.2 51.8 ± 11.5 52.1 ± 14.6

SW (mmHg·mL) 16.6 ± 2.0 13.4 ± 1.5 19.1 ± 2.0 17.8 ± 2.0 SWI-bw 62.5 ± 8.1 49.6 ± 4.9 76.5 ± 9.5 67.3 ± 7.3 (mmHg·µl/g) SWI-hw 23.7 ± 2.8 18.5 ± 1.8 22.2 ± 2.6 19.5 ± 1.6 (mmHg·µl/mg) #4 EF (%) 79.1 ± 6.3 75.2 ± 5.9 76.2 ± 4.7 78.1 ± 4.2

+dP/dt (mmHg/sec) 13550 ± 2087 9658 ± 1040 14774 ± 1192 14124 ± 921 Contractility Index 146 ± 13 122 ± 7 163 ± 4 139 ± 4*** (mmHg2/sec) #3,4 V@+dP/dt (µL) 190 ± 22 182 ± 4 204 ± 15 194 ± 26

PWRmax (mWatts) 97 ± 15 77 ± 10 106 ± 14 115 ± 17 PAMP 28.0 ± 4.3 24.6 ± 3.1 25.9 ± 2.3 35.5 ± 4.4 (mWatts/µL2) EDP (mmHg) #4 10.46 ± 0.67 10.60 ± 0.92 8.10 ± 0.72* 8.14 ± 0.41

EDV (µL) #2 192 ± 17 167 ± 12 203 ± 15 192 ± 24

-dP/dt (mmHg/sec) -10817 ± 1633 -8458 ± 1088 -12669 ± 1104 -12383 ± 692 tau (msec) #3 10.82 ± 1.09 12.59 ± 1.03 7.88 ± 0.33* 8.93 ± 0.33*** Ea (mmHg/µL) 0.75 ± 0.13 0.84 ± 0.11 0.93 ± 0.14 0.94 ± 0.07

HR, heart rate; SV, stroke volume; SVI, stroke volume index [sv/bw (body weight), sv/hw (heart weight)]; CO, cardiac output; CI, cardiac index (co/bw, co/hw); ESP, end-systolic pressure; ESV, end-systolic volume; SW, stroke work; SWI, stroke work index (sw/bw, sw/hw); EF, ejection fraction; +dP/dt, maximum rate of left ventricular pressure increase during systole; V@dP/dt max, volume at +dP/dt; PWRmax, maximal power output; PAMP, preload- adjusted maximal power; EDP, end-diastolic pressure; EDV, end-diastolic volume; -dP/dT, maximum rate of left ventricular pressure decrease during diastole; tau, left ventricular relaxation time constant; Ea, arterial elastance. Values are given as mean ± SEM. #2 variables with extreme outliers removed from COP.DA(chr 16) (n = 7); #3 variables with extreme outliers removed from DA (n = 9); #4 variables with extreme outliers removed from DA.COP(chr 16) (n = 9). * significantly different between COP and DA (P< 0.05); ** significantly different between COP and COP.DA(chr 16) (P< 0.05); ***significantly different between DA and DA.COP(chr 16) (P< 0.05). Those variables shaded in light gray are indices of contractility; variables shaded in dark gray are indices of relaxation.

dobutamine infusion (Little, 1985; Cheng et al., 1990; Cheng et al., 1992; Leather

et al., 2002; Nemoto et al., 2002; Schenk et al., 2004; Plante et al., 2005). Table

VII displays those parameters, the expected relative increase or decrease in

each variable as a result of dobutamine infusion, and the results obtained from

the COP, DA, and consomic rats expressed as percent change (%Δ) from

baseline values. F statistics are shown at the right of the table to draw attention

to the high degree of variability in the dobutamine response. Generally, the

parameters behaved as predicted by the aforementioned studies with the

exception of ESP, which decreased rather than increased in response to

dobutamine. For all parameters except heart rate, DA rats had a greater

response to dobutamine than COP rats, although none of the comparisons

reached statistical significance (Table VII). Interestingly, consomic rats tended to

respond more strongly to dobutamine than the parental strains (Table VII).

Statistical significance was not reached for these parameters, except for EF,

which increased 19.2 ± 5.1% in COP.DA(chr 16) rats compared to the 3.1 ± 1.6% 100

Table VII. Hemodynamic Response to Dobutamine for COP, DA, and Consomic Strains

Variables Expected to COP COP.DA(chr 16) DA DA.COP(chr 16) F Change with (n = 8) (n = 8) (n = 10) (n = 10) Statistic Dobutamine

Infusion ↑ HR (%Δ) 11.3 ± 2.7 15.3 ± 2.2 6.8 ± 1.2 7.9 ± 0.7 4.62

↑ ESP (%Δ) #1 -7.5 ± 8.4 -13.9 ± 5.6 -8.2 ± 2.1 -10.7 ± 3.0 0.35

↓ ESV (%Δ) #1 -22.3 ± 11.9 -49.4 ± 10.0 -33.6 ± 9.2 -43.0 ± 10.6 1.13

#1 ↑ EF (%Δ) 3.1 ± 1.6 19.2 ± 5.1** 17.7 ± 4.5 15.8 ± 5.2 2.10

↑ +dP/dt (%Δ) 29.0 ± 10.2 45.6 ± 11.3 42.0 ± 7.7 52.5 ± 6.9 1.20

↑ PWRmax (%Δ) #2, #3 27.5 ± 8.3 15.4 ± 12.2 37.8 ± 8.3 49.5 ± 8.5 2.44

↑ PAMP (%Δ) 26.3 ± 9.5 19.8 ± 19.2 32.6 ± 10.4 16.6 ± 10.2 0.35

↓ -dP/dt (%Δ) -7.9 ± 7.2 1.1 ± 10.5 -8.2 ± 4.0 -14.2 ± 5.4 0.85

↓ Tau (%Δ) -10.8 ± 5.8 -18.1 ± 2.2 -7.6 ± 2.7 -6.7 ± 2.4 2.26

HR, heart rate; ESP, end-systolic pressure; ESV, end-systolic volume; EF, ejection fraction; +dP/dt, maximum rate of left ventricular pressure increase during systole; V@+dP/dt, volume at +dP/dt; PWRmax, maximal power output; PAMP, preload-adjusted maximal power; -dP/dT, maximum rate of left ventricular pressure decrease during diastole; tau, left ventricular relaxation time constant. Arrows represent whether dobutamine was expected to increase or decrease parameters. Negative values indicate a decreased response. Values are given as mean ± SEM of the percent change from baseline to the dobutamine response. #1 variables with extreme outliers removed from COP (n = 7); #2 variables with extreme outliers removed from COP.DA(chr 16) (n = 7); #3 variables with extreme outliers removed from DA (n = 9). ** significantly different between COP and COP.DA(chr 16) (P< 0.05).

increase observed in COP rats (P=0.02). Appendix B displays the dobutamine

response in all hemodynamic parameters measured.

101

Molecular Characterization and Candidate Genes

Global Gene Expression Analysis in DA and COP Left Ventricles

Overall results for the gene expression and pathways analysis can be found

in Lee et al. (Lee et al., 2005). The results presented here summarize and

expand upon those reported. Male COP, DA, and F1(COPxDA) rats were

phenotyped for maximal treadmill running capacity prior to the microarray

experiments. The COP and DA rats ran 392 ± 55 m and 758 ± 35 m to

exhaustion, respectively (P<0.001). The running performance of the F1(COPxDA) rats was intermediate to that of the COP and DA strains (539 ± 26 m; P<0.05).

All performance results were consistent with and validated those described in the previous sections and published results (Barbato et al., 1998; Koch et al., 1999;

Ways et al., 2002).

The percentage of total probe sets on the Affymetrix U34 chip series corresponding to RNA transcripts expressed in left ventricles averaged 31.1% for

DA rats, 32.7% for COP rats and 36.2% for the F1(COPxDA) rats. These

percentages were not significantly different from one another. Average covariance between replicates was less than 20% for the probe sets identified as differentially expressed. The low variability observed in chip data comparisons validated the array analysis. Stringent selection criteria were applied to ensure that differential gene expression would be reproducible in all replicates for all groups.

102

Using the criteria described in Materials and Methods, 199 probe sets were identified as being differentially expressed. Twenty-eight of the 199 had multiple copies (25 in duplicate and 3 in triplicate, 59 probe sets in all) that generally represented different regions of the same expressed sequences. The aims were

1) to map these differentially expressed probe set sequences to chromosomal intervals containing the running capacity QTLs, and 2) to identify gene products located within QTL-containing intervals that interact with genes corresponding to the differentially expressed probe sets. All differentially expressed probe sets and interacting genes will hereby be referred to as genes/ESTs. Those genes/ESTs that were most closely associated with diastolic cardiac performance were considered the most superior candidates for the ARC QTLs.

Chromosomal locations of differentially expressed genes/ESTs were determined by comparing the DNA sequences used to design the probe sets on the Affymetrix gene chips with the rat genome database (Build 2.1 and 3.4) using the rat genome sequence database and Basic Local Alignment Search Tool

(BLAST) on the NCBI web server. Probe sets not mapping to the rat genome were localized using the mouse genome database (Build 30, NCBI) BLAST program to identify mouse genome sequences that were similar to probe set sequences. Comparative mapping strategies were then used to determine the orthologous rat genome locations. Due to the imprecision of QTL localization

(Darvasi et al., 1993; Hyne et al., 1995), genes/ESTs mapping to RNO16 where most of the chromosome was above the threshold for “suggestive” linkage, or the

103

proximal portion of RNO3 from D3Rat233-D3Rat98 that displayed an interaction effect with the region on RNO16 from D16Rat42-D16Rat90 (Figure 7), were considered as potential candidates for ARC QTLs. A total of six differentially expressed genes/ESTs had sequences mapping to ARC QTL-containing intervals (Table VIII).

Table VIII. Differentially Expressed Genes/ESTs Localized to ARC QTL- Containing Regions

Probe Set Mb Expression Chromosome Current Annotation ID Position

peptidylprolyl isomerase U68544_at COPDA 16 49.6 Pdlim3 TM2 domain containing AA818129_at COP>DA 16 71.4 2, Tm2d2 anaphase promoting AI072238_at COP>DA 3 3.4 complex subunit 2, Anapc2 LOC687266 similar to folylpolyglutamate synthase, mitochondrial AI072658_at COP>DA 3 10.7 precursor (folylpoly- gamma-glutamate synthetase) (FPGS)

Six of the original 199 probe sets were identified as being differentially expressed in the left ventricles of COP and DA rats mapped to QTL-containing intervals on rat chromosomes 16 and 3. Probe Set IDs are abbreviated versions of those found in the Affymentrix database (http://www.affymetrix.com/analysis/index.affx). Chromosomal positions and annotations were obtained from either the Rat Genome Database or NCBI gene query tools. Any differentially expressed genes/ESTs that were identified as being on RNO16 were considered candidate genes whereas only those genes/ESTs that were within a 1-LOD confidence interval of D3Rat56 (Ways et al., 2002) were considered as candidate genes.

104

Ingenuity Pathways Analysis identified 13 molecular networks containing 50

differentially expressed genes/ESTs, 40 of which were in the top three networks

(Lee et al., 2005). Table IX describes the likely functions/pathways associated

with top three networks identified, as well as the genes associated with the

function/pathway. Chromosomal locations were determined for the genes/ESTs

belonging to these networks that were not among the differentially expressed

genes/ESTs. Two genes, insulin receptor substrate 2 (Irs2), and acyl-CoA synthetase long-chain family member 1 (Acsl1) mapped to ARC QTL-containing regions on RNO16 (83 Mb and 49 Mb, respectively) and can be considered potential ARC QTL candidate genes. Additionally, cell division control protein 16

(Cdc16) was identified as a gene on RNO16 (81 Mb) that interacted with the differentially expressed Anapc2 on RNO3 (3.4 Mb). None of the differentially expressed genes that mapped to QTL-containing regions were identified in a network.

Confirming Expression of Pik3r1 in DA and COP Left Ventricles

The most highly differentially expressed gene identified by the microarray analysis was Pik3r1, represented by probe set U50412. Messenger RNA expression of this gene was 10-fold higher in DA left ventricles compared to

COP. This observation was confirmed by quantitative RT-PCR (Lee et al., 2005).

However, probe set AA819268, which also represented Pik3r1, was moderately

differentially expressed but in the opposite direction of U50412 (unconfirmed by

105

Table IX. Molecular Networks in Which Differentially Expressed Left Ventricular Genes/ESTs Might Be Operating and the Top Ranking Probable Functions of Each Network

Probable Network Network Genes in Network Function Acsl1, Casp8, Elovl6, Fabp5 (↓), Fads2, Lipid Metabolism Hnf4a, Irs2, Pik3ca, Pparg, Sc5dl, Scap, Scd, Srebf1, Slc3a2 (↓) 1 Endocrine System Cblb (↑), E2f1, Hnf4a, Irs2, Pik3r1* (↓), Disorders Pparg, Rela, Srebf1 Cblb (↑), E2f1, Hnf4a, Irs2, Pik3r1* (↓), Metabolic Disease Pparg, Srebf1 Akap12 (↑), Fyn, RT1-Ba (↑), Dbi (↓), Cell Cycle Id1 (↑), Id2 (↑), Inhba, Mycn, Npm1, Myc, Myod1, Nfkb1a (↑), Ptn (↑), Tp53 Hematological Fyn, Gstp1, Id1 (↑), Id2 (↑), Il2rb (↑), System Inhba, Myc, Nfkb1a (↑), Npm1, Tp53, Development and Cd74, Myod1, RT1-Da (↑) 2 Function Immune and Fyn, Gstp1, Id1 (↑), Id2 (↑), Il2rb (↑), Lymphatic System Inhba Myc, Nfkb1a (↑), Npm1, Tp53, Development and Cd74, Myod1 Function Cd9 (↑), Cebpb (↑), Dcn, Fn1* (↑), Fos, Cell Death Fosl1, Gata4, Ifng, Igfbp5 (↑), Inha (↓), Nid (↑), Nppa, Odc1, Srf, Stat3* (↓), Tcf1 Cebpb (↑), Eno1* (↑), Fn1 (↑), Fos, Fos1, Cellular Gata4, Hop* (↓), Ifng, Nkx2-5, Nmi, Srf, Development 3 Stat3* (↓), Tcf1 Cd9 (↑), Cdc2, Cebpb (↑), Crp, Dcn, Eno1* (↑), Fn1*(↑), Fos, Gata4, Ifng, Gene Expression Igfbp5 (↑), Inha (↓), Lgals3bp, Mmp11, Nkx2-5, Nppa, Odc1, Sat (↑), Serpine2, Stat3* (↓), Tcf1, Tgfbi

Thirteen molecular networks were identified containing 135 genes, 50 of which were differentially expressed. The three primary networks containing the largest number of differentially expressed genes were assigned potential functions by the Ingenuity program. Genes associated with the top three potential functions for each network are shown. Arrows indicate those genes showing decreased (↓) and increased (↑) expression in the left ventricles of COP compared to DA rats. Those genes mapping to QTL containing intervals are underlined and bolded.

106

quantitative RT-PCR). Both probe sets mapped to the predicted 3’-untranslated region of Pik3r1, but were opposite in orientation, suggesting they represent two different transcripts where U50412 is in the same orientation as the preceding

Pik3r1 gene. Therefore, only U50412 was considered further as the representative of Pik3r1.

Because Pik3r1 mRNA expression was so highly elevated in DA rat hearts compared to COP rat hearts, it was present in the top network (Table IX), and it interacts with Irs2 (which is also in the top network and on RNO16), we examined

Pik3r1 protein expression to determine if expression differences at the protein level are consistent with the expression differences observed at the mRNA level.

Pik3r1 encodes a protein known as phosphatidylinositol 3-kinase p85α, which may be expressed as three isoforms: p85α, p55α, and p50α (Inukai et al.,

1997). The antibody used was directed against an N-terminal SH2 region shared by all three isoforms, therefore isoform expression differences as well as overall expression differences could be resolved. Despite the large increase in cardiac

Pik3r1 expression observed at the mRNA level in DA rats compared to COP rats, only minor expression differences were observed for each isoform (1.5-fold greater for p85α and 1.9-fold greater for p50α in DA rats compared to COP rats) as well as for the sum of both isoforms (1.8-fold greater), none of which were statistically significant (Figure 13). The p55 isoform was not identified in left ventricular tissue, which is consistent with reports by others (Kessler et al.,

2001).

107

COP DA p85 α p50 α GAPDH

Figure 13. Left ventricles from 6 COP and 6 DA rats were used to perform a Western blot analysis of phosphatidyl inositol 3-kinase regulatory subunit p85α and its alternatively spliced isoform p50α. Densitometry readings from individual animals of each strain were normalized to the “housekeeping” gene glyceraldehyde 3-phosphate dehydrogenase (GAPDH) then averaged and used to determine fold-difference in expression.

Sequencing of Putative Candidate Genes

The decision of which putative candidate genes were to be sequenced was

based on 1) being proposed in the previous genome scan as being part of lipid

metabolism pathways (Ways et al., 2002) (Adrb3 and Lpl), 2) being present in

molecular networks containing differentially expressed genes/ESTs (Acsl1 and

Irs2), 3) being involved in cardiac metabolism and gene expression (Acsl1, Irs2,

Pik3r2), and 4) mapping to QTL-containing regions on RNO16 and RNO3. The protein coding regions of each gene were sequenced using cDNA derived from left ventricular mRNA. No sequence polymorphisms that would lead to non- synonymous changes altering the amino acid sequence were detected in the cDNA obtained from COP and DA rats. Only synonymous polymorphisms were found in two of the five genes sequenced: Six within the coding sequence of Lpl and two within the coding sequence of Irs2.

108

Metabolic Substrate Utilization

Exercise Performance

The DA rats had a significantly greater performance for a single bout of

exhaustive treadmill running compared to COP rats (954.4 ± 88.2 m versus 509.5

± 27.3 m; P=0.003). The vertical work performed during the treadmill run was

also significantly greater for DA rats than for COP rats (56.9 ± 5.5 kg⋅m versus

27.8 ± 3.5 kg⋅m; P=0.001). The results for vertical work are similar to those

found previously (Barbato et al., 1998) as were the results for distance run to

exhaustion (Barbato et al., 1998; Koch et al., 1999; Ways et al., 2002; Lee et al.,

2005), despite running performance only being measured on one day rather than

taking the best measurement out of five days.

Substrate Utilization

The results presented in Figure 14 display a measure of how COP and DA

rats differ in their utilization of metabolic substrates. Significant interaction

effects were observed for liver triglyceride, left ventricle triglyceride, and plasma

lactate concentrations, indicating that the change in these particular substrates

during exercise differed between COP and DA rats. In DA rats, exercise reduced liver triglycerides by 16%, elevated left ventricular triglycerides by 4%, and elevated plasma lactate by 113%. Conversely, in COP rats, exercise elevated liver triglycerides by 11%, reduced left ventricular triglycerides by 43%, and

109

reduced plasma lactate by 25%. Significant main effects were observed for plasma FFA where exercise increased concentrations in both COP and DA rats,

110

Figure 14. All measurements were made on plasma and tissues from animals that were provided food and water ad libitum. Glucose, free fatty acids (FFA), and lactate were measured in plasma samples whereas all other substrates were measured in the tissues indicated. Open squares () indicate COP rats and filled diamonds (♦) indicate DA rats. The left column describes measures of glucose metabolism while the right column describes measures of lipid metabolism. * significant interaction effect where the variable was significantly different for both strain and condition (P<0.05). ** significant main effect for condition where the variable was significantly different between the animals that were run and the controls (P≤0.05).

and for liver and muscle glycogen where exercise decreased concentrations in

both COP and DA rats (although to greater degrees in DA rats). Plasma

glucose, left ventricular glycogen, and muscle triglyceride concentrations were

not significantly altered as a result of exercise. The amount of circulating lactate

produced as a result of exhaustive exercise was greater in plasma from DA rats

than COP rats, but the difference was not statistically significant (6.4 ± 0.7 mM

versus 3.3 ± 1.3 mM; P=0.07). However, a positive, but insignificant (P=0.21),

correlation between vertical work performed and plasma lactate concentration

was also observed where DA rats produced more lactate than COP rats (Figure

15). The significantly greater workload achieved by DA compared to COP

accounted for 15% of the difference in lactate production between the two rat

strains.

111

Figure 15. Scatterplot showing plasma lactate concentration as a function of vertical work performed during exhaustive exercise. Open squares () represent COP rat values and filled diamonds (♦) represent DA rat values.

112

DISCUSSION

Congenic Strains and Running Performance

The 22% greater mean best distance run to exhaustion observed for the

COP.DA(chr 16) consomic strain compared with the parental COP strain indicates that transferring RNO16 alleles from DA rats into COP rats significantly improves ARC, confirming the presence of an ARC QTL on RNO16 (Figure 9).

Failing to observe a significantly greater performance for the COP.DA(chr 3) strain compared to COP rats (Figure 9) is consistent with the interaction identified between loci on RNO16 (D16Rat55) and RNO3 (D3Rat56) where a minimum of one DA allele was needed at each locus to observe a significantly greater best distance run to exhaustion (Ways et al., 2002). The COP.DA(chr 16) and

COP.DA(chr 3) congenic strains are homozygous for DA alleles at D16Rat55 and

D3Rat56, respectively, but homozygous for COP alleles at the other loci

(D3Rat56 and D16Rat55, respectively). Thus, positive epistatic interactions between these two ARC-QTLs would not be expected to occur in either strain.

The greater endurance running performance observed in COP.DA(chr 16) consomic rats compared to COP rats, therefore, is likely a product of the significant ARC-QTL identified near D16Rat17 (Ways et al., 2002).

Running performance of the DA.COP(chr 16) consomic strain was reduced by

22% compared to DA rats (Figure 11), which accounts for 56% of the difference between the COP and DA rats that were run concomitantly. Interestingly, the

113

performance improvements of COP.DA(chr 16) and COP.DA(chr 3) only

accounted for 30% and 17%, respectively, of the difference between the

concomitantly run COP and DA rats. Transferring RNO16 alleles from the COP

rat into the DA rat eliminates not only the influence of individual RNO16 QTLs

from DA alleles, but the interaction between D16Rat55 and D3Rat56 as well,

resulting in a more pronounced effect on running performance than that observed

for both COP.DA congenic strains. Thus, the DA.COP(chr 16) consomic strain

probably accounts for all identified QTLs and interactions except for any potential

individual effects of the RNO3 QTL.

The best distance run taken from five days of testing for each rat was used as

the measure of maximal running capacity because it was considered most

closely associated with the heritable component of aerobic running performance

(Barbato et al., 1998; Koch et al., 1998; Koch and Britton, 2001). The reason for

this is that on any given day, the environment (i.e. treadmill operators, room temperature/humidity, background noise and odors, cage cleaning, etc…) have a wide range of influence on individual treadmill running capacity, including the potential of reducing running performance values to zero. However, given

“optimal” environmental conditions, the genetic contribution to running capacity may be expressed to its fullest extent, resulting in the best possible genetically influenced performance. It could also be argued that this approach actually measures the effects of positive environmental influences on running performance rather than strictly a genetic phenomenon (Crabbe et al., 1999).

114

However, the parallel trend lines in Figure 16 indicate that while environmental conditions for each period of testing were highly variable, the response of each strain to the environment on their best day of performance was similar. This

Figure 16. Mean best distance for COP and DA rats for every treadmill running test that included both strains since the seminal work reported by Barbato et al., indicated at 1997 by the circled points (Barbato et al., 1998). Data points are displayed as means ± SEM. Open circles represent DA values; open squares represent COP values. Asterisks (*) indicate DA values significantly greater than COP values (P≤0.001). Dashed lines through each strain indicate the mean for all years. Solid lines indicate trend lines for each strain.

suggests that the consistently significant difference in best day of performance observed between the two strains was determined by genetic variation rather than the effect of the environment on genetic variation.

The consistently significant differences observed in average distance run to exhaustion for the inbred and congenic strains as well as the observation that

115

single bouts of running rather than the best day out of five yielded performance

differences commensurate with results obtained using best distance run further

corroborate the similar response to the environment between strains. Moreover,

the mean day of occurrence of the best distance run to exhaustion for all strains

clustered around day 3, indicating that the best day of performance for all strains

was not a function of improvement from training responses (Baldwin et al., 1977;

Dudley et al., 1982; Koch et al., 1998). These consistent responses to

environmental conditions are important observations if the argument is to be

made that strain differences in treadmill running performance are genetically determined rather than environmental.

Human and mouse studies have also linked maximal exercise capacity to chromosomal regions orthologous to those containing the RNO16 ARC QTLs

(Ways et al., 2002). Results from Lightfoot and co-workers identified a QTL for

exercise capacity in female mice at D8Mit359 which is orthologous to an interval

near the RNO16 ARC QTL defined by D16Rat17 (Figure 8) (Lightfoot et al.,

2007). A genome scan for human endurance performance using a cycle

ergometer identified a maximal power output QTL at D13S796 which is

orthologous to an interval near the RNO16 ARC QTL at D16Rat55 (Rico-Sanz et

al., 2004). While neither the human nor the mouse QTL mapped within the 1-

LOD support interval of our RNO16 ARC QTLs (Ways et al., 2002), the

imprecision of QTL localization (Darvasi et al., 1993) does not preclude the same

gene(s) from underlying any of the orthologous QTLs identified.

116

Intermediate Phenotypes

Previous work with COP and DA rats identified a number of anatomical and

physiological differences that may contribute to the strain differences in intrinsic

aerobic performance observed between inbred COP and DA rats (Table I).

However, aside from heart and body weights (Ways et al., 2002), these strain

differences have not been linked to specific chromosomal regions, making it

difficult to identify contributing genetic factors. Therefore, we took an alternative

approach to link anatomical and physiological strain differences to the ARC QTL-

containing regions by taking advantage of the newly constructed congenic strains

and the fact that they differ from recipient parental strains by only the QTL-

containing intervals transferred from donor parental strains (Appendix A). By

comparing anatomical and physiological measurements related to energy

homeostasis, cardiac function, and exercise performance made in the congenic

strains to the recipient parental strains, we can assume that observed deviations from the parental strains originate directly from specific RNO16 allelic substitutions, interactions involving the substituted alleles with the remainder of the recipient genome, or a combination of the two.

Body and Organ Weights

A number of organ weights were identified as being significantly different

between the congenic and parental strains (Tables II and IV). Organ weight

117

differences may imply the presence of functional differences in the organs, and therefore at least warrant their inclusion in future studies. For example, enlargement of left ventricles resulting from pressure overload leads to both impaired diastolic and systolic function of the heart during exercise (Cuocolo et al., 1990). Exercise training has been associated with enlargement of the pancreas from hypertrophy of exocrine pancreatic cells (Minato, 1997; Minato et al., 2000) and β-cells (Shima et al., 1997) leading to respective increases in cellular concentrations of digestive enzymes and metabolic hormones and improvements in energy homeostasis and metabolic abnormalities. Exercise training has also been shown to increase liver weight by increasing the amount of stored glycogen (Murakami et al., 1997), presumably the result of increased enzyme activities in the glycogen synthesis pathway. Although the rats in the present study were not exercise trained, allelic variation in genes that confer variation in key metabolic pathways could have similar effects.

The COP.DA(chr 3) congenic rats had significantly reduced body, pancreas, and kidney weights as well significantly greater body weight-adjusted heart and left ventricle weights compared to the parental COP rats (Table II), despite the lack of a significantly greater ARC (Figure 9). These strain-differences, then, may have resulted from interactions of transferred DA RNO3 alleles with background COP alleles, and are therefore unrelated to strain differences in

ARC. A discussion of how background alleles may affect the function of genes within transferred QTL-containing regions will be addressed below.

118

The DA.COP(chr 16) consomic rats had significantly greater pancreas

weights and liver weights compared to DA rats, despite reduced values for

running performance (Table IV and Figure 11). Reciprocal reductions in

pancreas and liver weights were observed in COP.DA(chr 16) consomic rats

compared to COP rats, although these differences did not reach statistical

significance (Table II). These organ weight values, however, trend in the

opposite direction from what would be expected based on the known effects of a

sedentary lifestyle and exercise training on performance (Troxell et al., 2003;

Koch et al., 2005; Booth and Lees, 2007) and the organ weights mentioned

(Minato, 1997; Murakami et al., 1997; Shima et al., 1997; Minato et al., 2000).

Also, liver glycogen concentrations (Figure 14) and plasma insulin concentrations

(Table III) were not different between any of the strains studied. Since all rats

were entirely sedentary except for the five consecutive days of treadmill running,

these results demonstrate that the differences in pancreas and liver weights

between strains are probably unrelated to functional differences that would be

expected to occur between rats with different running capacities. Therefore, the

genes regulating these organ weight differences are probably unrelated to the

genes regulating differences in running capacity, although further testing and

QTL delimitation will be required to be certain.

119

Cardiovascular Phenotypes

Barbato et al. reported a significantly greater cardiac output in DA rats

compared to COP rats using an isolated working heart preparation (Barbato et

al., 1998). The DA rats showed improved function over COP rats for other

cardiovascular variables as well (Table I). While heart weight (but not left

ventricle weight) and body weight QTLs were observed in the original

F2(COPxDA) genome scan (Ways et al., 2002), they did not co-segregate with

ARC. The finding of significantly altered ARC in consomic rats compared to the recipient parental strains (Figures 9 and 11), despite no relevant difference in heart or left ventricular weights (Tables II and IV), confirms that the RNO16 ARC

QTLs are independent of cardiac mass, though probably not of other functional cardiac phenotypes (Koch et al., 1999; Chen et al., 2001), which warranted further characterization of cardiovascular function (Tables VI and VII).

While most of the previous experiments aimed at characterizing variation in cardiac performance between COP and DA rats were carried out in vitro (Barbato et al., 1998; Chen et al., 2001), the present experiments were carried out in vivo.

Although in vitro experiments are generally very useful for isolating and studying specific variables such as contractility or intrinsic diastolic function, they lack the complexity of the whole organism and are often unable to account for peripheral physiological factors that would affect cardiac function, such as the autonomic nervous system, loading conditions, and substrate availability (Goodwin and

Taegtmeyer, 2000). While some of these conditions may be simulated in an in

120

vitro model to test specific hypotheses, in vitro experiments do not account for

natural strain differences in peripheral factors, and are therefore limited in their

ability to relate the underlying genetic determinants of running capacity to

intrinsic cardiac function. However, others have shown that strain differences in

cardiac function variables are maintained between in vivo and in vitro

experiments although the absolute values for those variables may not be

equivalent (Gyurko et al., 2000; Barbato et al., 2005; Jegger et al., 2006).

Therefore, we tested the hypothesis that differences in cardiac output,

contractility, and diastolic function as well as the similarity in heart rate observed

between COP and DA rats in vitro (Table I) would also be observed in vivo.

As the cardiac function of COP.DA(chr 16) consomic rats tended to be opposite to that which would be expected if enhanced cardiac function was

responsible for the greater running performance of COP.DA(chr 16) over COP

(Table VI), comparisons made using this strain will not be discussed. Otherwise,

very few hemodynamic variables were significantly different between any of the inbred and consomic strains compared in the present study. Cardiac output and cardiac output normalized to body weight (CI-bw) were greater in DA compared to COP rats. This may be explained by the tendency for DA rats to have larger stroke volumes and faster heart rates (Table VI), although none of the results were statistically significant. Heart weight-adjusted stroke volume (SV-hw) and cardiac output (CI-hw), however, tended to be greater in COP rats compared to

DA rats, suggesting that the elevated cardiac output and CI-bw in DA rats were

121

probably related to their greater heart size (Table II). However, DA.COP(chr 16)

consomic rats tended to have reduced cardiac outputs and stroke volumes (as

well as body-weight adjusted values for each) compared to DA rats, despite

similar heart sizes and heart rates (Tables IV and VI), demonstrating that RNO16

may actually harbor alleles having effects on cardiac performance independent of

heart size. CI-hw values, on the other hand, were lower than those of the DA

rats (Table VI), suggesting that peripheral variables specific to the DA genome at loci other than RNO16 are preventing the full expression of the superior cardiac

output of DA rats.

Arterial elastance (Ea), which is used as an index of afterload and is

calculated by dividing end-systolic pressure by stroke volume (ESP/SV) (Kelly et

al., 1992), was also elevated in DA rats compared to COP rats. Since DA rats

had a significantly greater ESP but similar SV compared to COP rats (Table VI),

DA rat hearts seemingly have a greater afterload to pump against than the COP

rat hearts. Similar afterload conditions appear to be present in the DA.COP(chr

16) consomic strain (Table VI), which explains why the CI-hw values were

reduced in these rats but not the COP rats despite both having reduced

unadjusted stroke volume and cardiac output values compared to DA. While it

was not determined whether the afterload differences were due to arterial stiffness or enhanced sympathetic tone, the results are consistent with the greater arterial blood pressures previously observed in DA rats resulting from enhanced sympathetic arterial tone (Table I, #4). Although none of the

122

comparisons made presently were statistically significant, the data suggest that the cardiac performance differences between DA and COP do exist in vivo, that they are associated with alleles on RNO16, and that they are blunted by a greater afterload in DA rats.

Variation in cardiac output can be attributed to conditions that alter contractility and diastolic function, both of which were observed to be significantly different between DA and DA.COP(chr 16) whereas only diastolic function differences were observed between COP and DA (Tables I and VI). None of the indices of contractility were significantly different between COP and DA rats

(Table VI), though DA.COP(chr 16) rats had a significantly lower contractility index compared to DA rats. Although other measures of contractility in this study were not significantly different between DA and DA.COP(chr 16) rats, a report by

Schenk et al. discussed that indices of contractility do not always correlate to a high degree (Schenk et al., 2004). Therefore, the relationship between cardiac contractility, RNO16, and running performance appears to be validated by these results.

Diastolic function also seems to be implicated in the effects of RNO16 on cardiac function because tau, the most commonly used index of active relaxation

(Schertel, 1998), was significantly different from DA rats in both COP and

DA.COP(chr 16) rats (Table VI). These results are opposite those found by

Chen et al., but may reflect the difference between in vivo and in vitro experiments. The DA rats were shown to have greater peak intracellular systolic

123

calcium and greater calcium storage capacity, but a slower rate of calcium normalization as indicated by a larger tau value (Chen et al., 2001). Adrenergic stimulation from the sympathetic nervous system enhances the rate of cardiac relaxation by increasing calcium uptake by the sarcoplasmic reticulum (Cheng et al., 1992; Libonati, 1999; Little et al., 2000), and DA rats were previously observed to have greater sympathetic tone compared to COP rats (Koch et al.,

1999). It appears that the differences in sympathetic tone still exist in the present study, as suggested by the greater afterload in DA rats, and that cardiac relaxation is enhanced as a result of that sympathetic tone. Although

DA.COP(chr 16) rats had similar afterload conditions to DA rats, indicating a similar sympathetic tone, the majority of the effects on cardiac relaxation seem to have been blunted and are possibly the result of the same COP genes on

RNO16 that caused the reduction in CI-hw.

Autonomic regulation of cardiac output and peripheral blood flow is a key determinant of the mammalian capacity to consume oxygen (Saltin, 1985).

Dobutamine infusion allowed us to test the effects of an adrenergic receptor agonist on cardiovascular function, since sympathetic regulation of cardiovascular function was shown to be greater in DA rats compared to COP rats (Koch et al., 1999). Dobutamine acts on beta-1 adrenergic receptors present in the heart to increase heart rate, contractility, and relaxation (Cheng et al., 1992; Schnermann, 2002). Unfortunately, the response to dobutamine was not significantly different between any of the strains for all variables listed in

124

Table VII, except for ejection fraction in COP.DA(chr 16). F statistics were

included in the table to highlight the large variability within strains compared to

between strains, which may explain the lack of significant findings. However, all

variables for all rats behaved as would be expected with dobutamine infusion

based on the work of others (Little, 1985; Cheng et al., 1990; Cheng et al., 1992;

Leather et al., 2002; Nemoto et al., 2002; Schenk et al., 2004; Plante et al.,

2005). Segel, et al. noted that sodium pentobarbital depresses overall cardiac

function by 10-30% and dobutamine responses can be blunted by high

catecholamine levels (Segel and Rendig, 1986), both of which may contribute to

the observed variability in dobutamine response, especially if the rats respond

differently to sodium pentobarbital (which was not tested).

Substrate Metabolism

Although subcutaneous fat deposits are a major source of energy stores, high

aerobic fitness is often associated with lower abdominal fat content, including

subcutaneous fat (Ross et al., 2000; King et al., 2001; Janssen et al., 2004; Ross

et al., 2004; Wong et al., 2004; Wisloff et al., 2005). The significantly greater subcutaneous abdominal fat (Figure 10) and lower fasting plasma triglyceride concentrations (Table III) observed in the COP.DA(chr 16) consomic rats suggests an increased ability to store dietary fat compared to COP rats. Since

COP.DA(chr 16) rats had a significantly greater running performance compared to COP rats, the increased subcutaneous abdominal fat store may not be

125

detrimental, but serve as an abundant energy reservoir, providing a metabolic

advantage during endurance exercise, particularly during the low-intensity portion

of the treadmill running test (Brooks and Mercier, 1994; Brouns and van der

Vusse, 1998). Most FFA released into the circulation originate from upper-body

subcutaneous fat, with smaller contributions from leg and intra-abdominal fat, even in individuals with abdominal obesity (Jensen, 2006). Indeed, DA rats appear to have a greater capacity for fatty acid release from adipose tissue when stimulated by exercise than COP rats (Figure 14, E), although this difference was not significant and has not been linked to loci on RNO16.

Transferring COP RNO16 alleles into a DA genetic background did not result in DA.COP(chr 16) rats having significantly less subcutaneous (and total) abdominal fat compared to DA (Figure 12) as would have been expected based on the results from COP.DA(chr 16) rats. The reason reciprocal effects were not observed is likely the result of COP RNO16 alleles regulating fat storage independent of ARC interacting with background DA alleles in such a way that preserves the superior capacity to store fat. Body composition and metabolism are complex traits in and of themselves, and only a subset of the genes that regulate the ability to store and utilize fat are expected to be related to exercise capacity (Figure 1). Furthermore, genes that regulate complex traits have been shown to interact with loci that do not have direct effects on phenotypic expression, but are capable of modifying the loci that do have direct effects on the phenotype in cancer models (Dietrich et al., 1993; Paul et al., 1993; Dragani,

126

2003). The effects of these modifier loci on tumor number were not observed until the locus being modified was transferred into a different genetic background.

It is possible that these modifier loci exist in the ARC model as well, for the running capacity phenotype as well as intermediate and related phenotypes, especially in the COP.DA(chr 3) congenic strain mentioned above.

One hypothesis that attempts to link metabolic differences to limitations to exercise performance is that of substrate crossover during exercise (Brooks and

Mercier, 1994). According to this hypothesis, energy derived from carbohydrates progressively predominates over energy derived from lipids as exercise intensity increases in order to maintain a high power output. The more rapidly this switch is made during the course of exercise, the sooner exhaustion will occur.

Because the treadmill running test used with the COP and DA rat model becomes progressively more intense the longer the rats stay on the treadmill, it seemed logical to determine how substrate concentrations in the blood and key tissues change as a result of exhaustive exercise in COP and DA rats. Since it would be extremely difficult, if not impossible, to measure certain tissue substrate concentrations in the same rats before and after exercise, these measurements were made in rats run to exhaustion and compared to sedentary control rats exposed to an immobilized treadmill.

The variation in running performance for COP and DA rats tested for exercise- induced changes in substrate metabolism were similar to those found previously

(Barbato et al., 1998; Koch et al., 1999; Ways et al., 2002; Lee et al., 2005),

127

despite running performance only being measured on one day rather than taking the best measurement out of five days. This is an important observation because we can assume that all differences in substrate concentrations between control and sedentary animals were the result of the level of exhaustion that would be observed if the best day out of five were used.

There was a significant depletion of liver and muscle glycogen stores for both

COP and DA rats, although to a greater extent in DA rats (Figure 14, B and C).

This makes sense since DA rats ran 87% farther on the treadmill, which corresponded to a 47% greater intensity and 84% greater workload compared to

COP rats. While the decline in concentration of muscle triglycerides was similar between COP and DA rats, liver triglyceride concentrations in DA rats were greater to begin with and declined significantly, while COP rats appeared to have an increase in concentration of liver triglycerides (Figure 14, F). These data support the conclusions based on subcutaneous fat, fasting plasma triglycerides, and the capacity to release fatty acids for use as an energy source where DA rats have a greater capacity to obtain energy from peripheral fat stores than COP rats. In the left ventricle, however, COP rats appear to rely less on intracellular carbohydrates to fuel cardiac metabolism during exercise and more on stored triglycerides (Figure 14, D and H). Coupled with the strain differences in plasma lactate produced during exercise (Figures 14-I and 15), DA rat hearts may be biased toward the utilization of a larger proportion of carbohydrates than COP rat hearts to generate a more powerful cardiac output to overcome the greater

128

peripheral resistance. Indeed, DA rats also appear to have a greater maximal power increase in response to dobutamine (Appendix B).

Cardiac lipid utilization is more dependant on circulating FFA rather than stored reserves (Stanley et al., 2005). Despite this, COP rats showed a more significant depletion of cardiac triglycerides compared to DA rats after a treadmill run to exhaustion, suggesting the presence of dysfunctional metabolic pathways in the COP heart. These interesting metabolic differences need to be studied in greater detail with larger animal sets as well as in the consomic strains to determine 1) if these differences are the result of QTLs on RNO16 and related to

ARC and 2) to dissect how these differences are contributing to variation in cardiac performance and aerobic running capacity.

Quantitative trait loci for several other traits potentially related to exercise capacity have been mapped to mouse and human genomic intervals orthologous to RNO16 and the proximal portion of RNO3. These QTLs are mapped within a

1-LOD support interval for the RNO3 and RNO16 ARC QTLs and include mouse

QTLs for lean body mass and type II diabetes mellitus (determined by glucose and insulin concentrations as well as subcutaneous and mesenteric fat weights)

(Masinde et al., 2002) and a mouse QTL for fasting plasma triglycerides (Suto and Sekikawa, 2003), respectively. Furthermore, the beta-3 adrenergic receptor

(Adrb3) and lipoprotein lipase (Lpl) genes, both located near RNO16 ARC QTLs, have been associated with differences in fat mass in humans (Sakane et al.,

1997; Garenc et al., 2001). These comparative QTL data support the observed

129

differences in fasting triglyceride concentrations in COP.DA(chr 16) consomic

rats compared with COP rats and in fat mass in both COP.DA(chr 3) and

COP.DA(chr 16) rats compared with COP rats.

Left Ventricle Expression Analysis and Candidate Genes

Only a small number (199) of the nearly 27,000 probe sets were differentially expressed in the left ventricles of 15 week-old male DA and COP rats. Criteria for determining differential expression were based on a relatively low threshold

(1.3-fold difference) because QTLs often exert small effects. Despite this liberal threshold, reliability of modest differences in expression was ensured by rigorous elimination of non-reproducible results. Reproducibility of a subset of the 199 differentially expressed probe-sets was confirmed by real-time PCR (Lee et al.,

2005).

Six differentially expressed probe sets mapped to QTL-containing intervals on

RNO16 and RNO3 (Table VIII). While these probe sets corresponding to ESTs and known genes have potential as candidates for ARC strain differences, based on their differential expression in left ventricles and location near known ARC

QTLs, additional evidence is required. As such, the relevance of these candidates will be discussed in the context of the consomic strains and their differences from parental recipient strains in terms of cardiac function, especially relaxation and metabolism, and regulatory pathways through which they operate.

130

Genes interacting with the 199 differentially expressed probe sets were sought to identify potential candidate genes for ARC QTLs that may be regulating the expression differences. While allelic differences that alter protein function may not be detected by any expression analysis, such allelic differences could affect the expression of interacting genes that are detected by expression analysis. For example, Pik3r1 was among the most differentially expressed genes identified (Lee et al., 2005) and is located on RNO2. The Pik3r2 gene, located on RNO16, is capable of restoring the activity of PI 3-kinase independent pathways when expressed in cultured brown adipose cells where Pik3r1 has been ablated (Ueki et al., 2003). Molecular networks generated from the expression data of the known genes/ESTs contained within the 199 differentially expressed probe sets were identified using Ingenuity Pathways Analysis. Genes belonging to these networks were identified as being present in cardiac tissue, mapping to QTL-containing intervals on RNO16, and interacting with the one or more of the differentially expressed genes.

Peptidyl-prolyl isomerase F (Ppif), also known as cyclophilin D, is part of a complex of proteins known as the mitochondrial permeability transition pore which contains an adenine nucleotide translocase that is activated by oxidative stress as well as calcium and is involved in cell death (Baines et al., 2005). The greater expression in DA hearts may be represented by a greater number of transition pores that would release ATP into the cell, although the expression of other known members of the transition pore is unknown (Halestrap et al., 1998).

131

Multimerin 2 (Mmrn2) is a gene that is expressed in embryonic vascular endothelial cells and endocardial cells (Leimeister et al., 2002) and may be involved in matrix assembly and adhesion in adult tissues, although its primary function remains unknown (Hayward et al., 1998). The PDZ and LIM domain 3 gene (Pdlim3) is a member of the PDZ-LIM family of proteins that are known to localize to Z lines of skeletal and smooth muscle. However, Pdlim3 is also expressed in developing cardiac tissue to enhance actin cross-linking, aiding in cardiac muscle mechanical strength/stability and sensing mechanical stress

(Pashmforoush et al., 2001; Hoshijima, 2006). The greater expression in adult

COP rat hearts may indicate compensatory mechanisms for decreased cardiac function such as that which is seen with beta-myosin in heart failure (Lowes et al., 1997). The TM2 domain containing 2 gene (Tm2d2), as of yet, has no known function, which precludes speculation about its role in cardiac performance or

ARC differences in COP and DA rats.

Anaphase promoting complex 2 (Anapc2) was the only differentially expressed gene mapping to the QTL-containing interval on RNO3 that was found to interact with a gene at the other epistatic locus on RNO16. The epistatic locus on RNO16 harbors cell division cycle 16 (Cdc16), which interacts with Anapc2 to mediate ubiquitin-ligase activity of the anaphase promoting complex, and the destruction of cell cycle regulatory proteins shortly before anaphase (Zachariae and Nasmyth, 1999). The ubiquitin-proteasome system is also involved in the regulation of sarcomeric protein turnover and cardiomyocyte size where

132

dysregulation of the system may lead to pathologic cardiac consequences

(Powell, 2006), as suggested by the cardiac performance of COP rats compared to DA. Folypolyglutamate synthetase (Fpgs), located on RNO3, is involved in folate metabolism and purine synthesis in the mitochondria and cytosol of cells

(Lin et al., 1993), although an interacting partner was not identified on RNO16

(Figure 17).

The acyl-CoA synthetase long-chain family member 1 gene (Acsl1) encodes a protein that catalyzes the attachment of long chain fatty acids to coenyzme A, which is the first enzymatic step required for the oxidation of fatty acids for energy production in the heart and the elongation of fatty acids for storage

(Coleman et al., 2000). Dysregulation of this gene product could lead to altered fat metabolism in the heart and may help explain the observations in COP cardiac lipid metabolism during exhaustive exercise, although the apparent lack of fatigue in COP rats may explain this also (Figure 14, H and I). Insulin receptor substrate 2 (Irs2) has been shown to bind to p85β, the protein product of Pik3r2, and regulate cardiac glycolysis, protein synthesis, and gene expression independent of p85α, the protein product of the highly differentially expressed gene Pik3r1, which mediates cardiac GLUT4 translocation (Kessler et al., 2001).

Significantly greater Pik3r1 mRNA expression in the left ventricles of DA rats suggests more efficient cardiac glucose uptake, which could lead to catecholamine-induced increases in myocardial performance in the DA rat as suggested by the greater utilization of cardiac glycogen than in COP rats (Figure

133

Figure 17. Summary of all genes identified as potential candidates and their map position relative to the ARC QTLs represented by the congenic/consomic strains. Map distances are listed in megabases (Mb). The microsatellite markers shown either represent QTL peaks (bold and underlined) or the boundaries of the congenic strains and 1-LOD intervals (shown as thicker lines on the maps). Those genes with accession numbers attached to them are those that were identified as being differentially expressed. Those genes without accession numbers are those that were either identified as interacting with differentially expressed genes, alternate isoforms of differentially expressed genes, or those proposed as previous candidates in the COP/DA model or in humans and sequenced (excepting Cdc16).

14, D), although significant differences in protein expression were not observed

(Figure 13). The Irs2 and Pik3r2 genes, then, may be involved in other signaling

pathways that moderate cardiac function and ARC, as Irs2 has been proposed to

be a candidate gene for an intrinsic maximal power output QTL in humans (Rico-

Sanz et al., 2004). The Acsl1, Irs2, and Pik3r1 genes were also present in the 134

topmost network, for which the most probable functions include lipid metabolism, endocrine disorders, and metabolic disease (Table IX). Since these functions were suggested as potential mechanisms for the differences in cardiac performance observed between the DA.COP(chr 16) consomic strain and DA rats (Table VI) and for the cardiac substrate preferences during exercise observed between COP and DA rats (Figure 14, D and H), Acsl1, Irs2, and

Pik3r1 probably represent the most superior candidate genes/pathways for the

ARC strain differences observed between COP and DA rats. However, the results of sequencing the protein-coding regions for Acsl1, Irs2, and Pik3r2 suggest otherwise.

While the rat ortholog of human myosin light chain 2a gene (Mlc2a) does not map to an ARC QTL-containing interval, its high degree of differential mRNA expression in DA compared to COP rats may help explain the strain differences in ARC and cardiac function described here and in previous studies (Koch et al.,

1999; Chen et al., 2001; Ways et al., 2002). Indeed, decreased myosin heavy chain α/β isoform expression ratio (at both the RNA and protein levels) was associated with decreased cardiac performance in Buffalo rats, another low performing inbred strain, compared to DA rats (Barbato et al., 1998; Barbato et al., 2002).

Proving that any of the candidate genes described above contribute to the

ARC QTLs requires identification of strain-specific allelic differences and altered protein expression or function in key organs. Unfortunately, none of the genes

135

proposed as candidates displayed protein-coding sequence differences that would support a change in function. That does not mean, however, that the other genes proposed do not harbor relevant sequence variants that might explain strain differences in ARC, cardiac function, or metabolism/body composition

(Figure 17). Rat chromosome 16 has a gene density of ~9/Mb (~14/cM), which means that of the 90 Mb containing >800 genes that constitute RNO16, about

1.5% (12 genes) have been considered in the present study, of which only five have been sequenced in the COP and DA rats. Aerobic running capacity is an exceptionally complex trait, dependent upon the interplay of numerous tissues and organs. Therefore, the expression differences observed in this study may not arise solely because of differences in myocardial mRNA expression.

Although the genes in Figure 17 were carefully chosen based on chromosomal map position, the physical mapping boundaries of the congenic/consomic strains, and known relationships with physiological/biochemical pathways that contribute to the overall aerobic performance phenotype, they serve only to speculate about the underlying genetic determinants of ARC. Potential ARC candidate genes may have been missed for several reasons: 1) expression of all rat genes were not assessed because all genes were not present on the available microarray chips, 2) the identification and function of most genes/ESTs remain poorly understood at best, unknown at worst, and therefore some are not present in available molecular databases, and 3) of all tissues associated with ARC QTLs in this study, only the left ventricles were analyzed for gene expression.

136

Perspectives/Future Work

Transferring both RNO16 and RNO3 QTL-containing regions into the same genetic background, whether that be COP or DA rats, would allow confirmation of

the epistatic interaction observed in the ARC genome scan, as has been done for other rat phenotypes (Rapp et al., 1998; Monti et al., 2003; Van Dijk et al., 2006).

Construction of these double congenic strains would also facilitate examination of the summed effects of the alleles that were introgressed into the COP.DA(chr 16) and COP.DA(chr 3), for ARC and the other measured phenotypes. Furthermore, the double congenic strains would allow for the separation of the two interacting loci from the main QTL at D16Rat17. Such an approach is the only sure way to determine whether the phenotypes that were observed to be significantly different between the congenic/consomic strains and their respective recipient parental strains are indeed linked to the confirmed ARC QTLs. Whether this confirms current results, denies them, or brings about new observations remains to be determined.

137

CONCLUSIONS

1. The COP.DA(chr 16) consomic strain confirmed at least one ARC QTL

responsible for the observed ARC strain differences between COP and

DA rats. The lack of significant ARC improvement in COP.DA(chr 3) rats

affirmed that the QTL harbored there exerted only very weak effects on its

own and is consistent with the expectation that an interaction is indeed

needed with DA alleles on RNO16 to observe a significant increase in

running performance. The DA.COP(chr 16) consomic strain corroborated

the effects of the main QTL and probably the effects of the interaction

between QTLs on RNO16 and RNO3 on running performance.

2. The significantly greater amount of subcutaneous abdominal fat, lower

fasting plasma triglyceride concentrations, and greater capacity to release

free fatty acids during exercise observed in COP.DA(chr 16) and DA rats

compared to COP rats suggest that genes regulating energy balance and

ARC are present in the same RNO16 consomic interval.

3. Cardiac function of the DA.COP(chr 16) consomic strain and COP rats

compared to DA rats helped confirmed that contractility, diastolic function,

overcoming peripheral resistance to blood flow, and cardiac output are

138

dependant on RNO16 QTLs that harbor genes responsible for ARC strain

differences between COP and DA rats.

4. The COP and DA rats exhibit different exercise-induced changes in

energy substrates that may be related to differences in cardiac

performance. These results suggest that DA rats also have a greater

capacity to utilize fats as an energy source during exercise compared to

COP rats. The lower concentration of circulating lactate in COP rats

compared to DA rats also suggests that COP rats are not metabolically

exhausted by the end of the treadmill running test and other factors are

contributing to their low treadmill running performance.

5. Twelve candidate genes were proposed to be the underlying genetic

determinants of the ARC strain differences observed between COP and

DA rats, five of which have been sequenced in protein-coding regions. No

sequence polymorphisms that would alter the amino acid sequence and

lead to functional protein differences were identified. Therefore, these five

genes are less likely to be superior candidate genes for the ARC QTLs.

6. At present, we cannot distinguish whether the phenotypic differences

identified in this study stem from the same alleles that affect ARC, or

whether these differences represent distinct QTLs [and gene(s)], present

139

on RNO16 and RNO3. Analysis of congenic substrains for the RNO16 and RNO3 QTLs, which would carry smaller regions of donor-transferred alleles, could be used to determine whether or not ARC and the energy balance phenotypes co-localize.

140

SUMMARY

Understanding how exercise relates to overall fitness is essential to

understanding health and disease susceptibility. Whether an innately high level

of fitness or regular physical activity is more important to health is still under

debate, but even the most physically fit person is still susceptible to risk factors associated with inactivity and can significantly reduce risk susceptibility by

participating in a regular exercise program. About 50% of variation in exercise

performance is attributable to genetic factors. Identifying and characterizing

these genetic factors will lead to a more comprehensive understanding of the

relationship between genes, physical fitness, and overall health. The work

presented here is the continuing progress of genetically characterizing a rodent

model of exercise capacity (Barbato et al., 1998; Koch et al., 1999; Chen et al.,

2001; Walker et al., 2002; Ways et al., 2002; Lee et al., 2005; Ways et al., 2007)

and represents the first congenic, and consomic, strains for exercise capacity in

rats. Since much of the data generated coincides with that which has been found

in mice and humans, this model is relevant to human conditions and will likely

lead to the discovery of human genes that link aerobic fitness to overall health.

141

REFERENCES

Abiola, O., Angel, J. M., Avner, P., Bachmanov, A. A., Belknap, J. K., Bennett, B.,

Blankenhorn, E. P., Blizard, D. A., Bolivar, V., Brockmann, G. A., Buck, K. J.,

Bureau, J. F., Casley, W. L., Chesler, E. J., Cheverud, J. M., Churchill, G. A.,

Cook, M., Crabbe, J. C., Crusio, W. E., Darvasi, A., de Haan, G., Dermant, P.,

Doerge, R. W., Elliot, R. W., Farber, C. R., Flaherty, L., Flint, J., Gershenfeld,

H., Gibson, J. P., Gu, J., Gu, W., Himmelbauer, H., Hitzemann, R., Hsu, H.

C., Hunter, K., Iraqi, F. F., Jansen, R. C., Johnson, T. E., Jones, B. C.,

Kempermann, G., Lammert, F., Lu, L., Manly, K. F., Matthews, D. B.,

Medrano, J. F., Mehrabian, M., Mittlemann, G., Mock, B. A., Mogil, J. S.,

Montagutelli, X., Morahan, G., Mountz, J. D., Nagase, H., Nowakowski, R. S.,

O'Hara, B. F., Osadchuk, A. V., Paigen, B., Palmer, A. A., Peirce, J. L., Pomp,

D., Rosemann, M., Rosen, G. D., Schalkwyk, L. C., Seltzer, Z., Settle, S.,

Shimomura, K., Shou, S., Sikela, J. M., Siracusa, L. D., Spearow, J. L.,

Teuscher, C., Threadgill, D. W., Toth, L. A., Toye, A. A., Vadasz, C., Van

Zant, G., Wakeland, E., Williams, R. W., Zhang, H. G. and Zou, F. (2003).

The nature and identification of quantitative trait loci: a community's view. Nat

Rev Genet, 4, 911-6.

Ahn, J., Varagic, J., Slama, M., Susic, D. and Frohlich, E. D. (2004). Cardiac

structural and functional responses to salt loading in SHR. Am J Physiol Heart

Circ Physiol, 287, H767-72.

142

Aitman, T. J., Glazier, A. M., Wallace, C. A., Cooper, L. D., Norsworthy, P. J.,

Wahid, F. N., Al-Majali, K. M., Trembling, P. M., Mann, C. J., Shoulders, C.

C., Graf, D., St Lezin, E., Kurtz, T. W., Kren, V., Pravenec, M., Ibrahimi, A.,

Abumrad, N. A., Stanton, L. W. and Scott, J. (1999). Identification of Cd36

(Fat) as an insulin-resistance gene causing defective fatty acid and glucose

metabolism in hypertensive rats. Nat Genet, 21, 76-83.

Aitman, T. J., Gotoda, T., Evans, A. L., Imrie, H., Heath, K. E., Trembling, P. M.,

Truman, H., Wallace, C. A., Rahman, A., Dore, C., Flint, J., Kren, V., Zidek,

V., Kurtz, T. W., Pravenec, M. and Scott, J. (1997). Quantitative trait loci for

cellular defects in glucose and fatty acid metabolism in hypertensive rats. Nat

Genet, 16, 197-201.

Ashe, M. C. and Khan, K. M. (2004). Exercise Prescription. J Am Acad Orthop

Surg, 12, 21-27.

Baines, C. P., Kaiser, R. A., Purcell, N. H., Blair, N. S., Osinska, H., Hambleton,

M. A., Brunskill, E. W., Sayen, M. R., Gottlieb, R. A., Dorn, G. W., Robbins, J.

and Molkentin, J. D. (2005). Loss of cyclophilin D reveals a critical role for

mitochondrial permeability transition in cell death. Nature, 434, 658-62.

Baldwin, J. E. and Krebs, H. (1981). The evolution of metabolic cycles. Nature,

291, 381-2.

Baldwin, K. M., Cooke, D. A. and Cheadle, W. G. (1977). Time course

adaptations in cardiac and skeletal muscle to different running programs. J

Appl Physiol, 42, 267-272.

143

Barbato, J. C., Huang, Q. Q., Hossain, M. M., Bond, M. and Jin, J. P. (2005).

Proteolytic N-terminal truncation of cardiac troponin I enhances ventricular

diastolic function. J Biol Chem, 280, 6602-9.

Barbato, J. C., Koch, L. G., Darvish, A., Cicila, G. T., Metting, P. J. and Britton, S.

L. (1998). Spectrum of aerobic endurance running performance in eleven

inbred strains of rats. J Appl Physiol, 85, 530-536.

Barbato, J. C., Lee, S. J., Koch, L. G. and Cicila, G. T. (2002). Myocardial

function in rat genetic models of low and high aerobic running capacity. Am J

Physiol Regulatory Integrative Comp Physiol, 282, R721-726.

Barton, N. H. and Keightley, P. D. (2002). Understanding quantitative genetic

variation. Nat Rev Genet, 3, 11-21.

Bennett, B., Downing, C., Parker, C. and Johnson, T. E. (2006). Mouse genetic

models in alcohol research. Trends Genet, 22, 367-74.

Billat, V. L., Mouisel, E., Roblot, N. and Melki, J. (2005). Inter- and intrastrain

variation in mouse critical running speed. J Appl Physiol, 98, 1258-63.

Blair, S. N., Cheng, Y. and Holder, J. S. (2001). Is physical activity or physical

fitness more important in defining health benefits? Med Sci Sports Exerc, 33,

S379-99; discussion S419-20.

Blair, S. N., Kampert, J. B., Kohl, H. W., 3rd, Barlow, C. E., Macera, C. A.,

Paffenbarger, R. S., Jr. and Gibbons, L. W. (1996). Influences of

cardiorespiratory fitness and other precursors on cardiovascular disease and

all-cause mortality in men and women. Jama, 276, 205-10.

144

Blair, S. N., Kohl, H. W., 3rd, Barlow, C. E., Paffenbarger, R. S., Jr., Gibbons, L.

W. and Macera, C. A. (1995). Changes in physical fitness and all-cause

mortality. A prospective study of healthy and unhealthy men. Jama, 273,

1093-8.

Booth, F. W. and Lees, S. J. (2007). Fundamental questions about genes,

inactivity, and chronic diseases. Physiol Genomics, 28, 146-57.

Bottger, A., van Lith, H. A., Kren, V., Krenova, D., Bila, V., Vorlicek, J., Zidek, V.,

Musilova, A., Zdobinska, M., Wang, J. M., van Zutphen, B. F., Kurtz, T. W.

and Pravenec, M. (1996). Quantitative trait loci influencing cholesterol and

phospholipid phenotypes map to chromosomes that contain genes regulating

blood pressure in the spontaneously hypertensive rat. J Clin Invest, 98, 856-

62.

Bouchard, C., An, P., Rice, T., Skinner, J. S., Wilmore, J. H., Gagnon, J.,

Perusse, L., Leon, A. S. and Rao, D. C. (1999). Familial aggregation of

VO(2max) response to exercise training: results from the HERITAGE Family

Study. J Appl Physiol, 87, 1003-8.

Bouchard, C., Daw, E. W., Rice, T., Perusse, L., Gagnon, J., Province, M. A.,

Leon, A. S., Rao, D. C., Skinner, J. S. and Wilmore, J. H. (1998). Familial

resemblance for VO2max in the sedentary state: the HERITAGE family study.

Med Sci Sports Exerc, 30, 252-8.

Bouchard, C., Despres, J. P. and Mauriege, P. (1993). Genetic and nongenetic

determinants of regional fat distribution. Endocr Rev, 14, 72-93.

145

Bouchard, C., Leon, A. S., Rao, D. C., Skinner, J. S., Wilmore, J. H. and Gagnon,

J. (1995). The HERITAGE family study. Aims, design, and measurement

protocol. Med Sci Sports Exerc, 27, 721-9.

Bouchard, C., Lesage, R., Lortie, G., Simoneau, J. A., Hamel, P., Boulay, M. R.,

Perusse, L., Theriault, G. and Leblanc, C. (1986). Aerobic performance in

brothers, dizygotic and monozygotic twins. Med Sci Sports Exerc, 18, 639-46.

Bouchard, C., Rankinen, T., Chagnon, Y. C., Rice, T., Perusse, L., Gagnon, J.,

Borecki, I., An, P., Leon, A. S., Skinner, J. S., Wilmore, J. H., Province, M.

and Rao, D. C. (2000). Genomic scan for maximal oxygen uptake and its

response to training in the HERITAGE Family Study. J Appl Physiol, 88, 551-

Bray, M. S. (2000). Genomics, genes, and environmental interaction: the role of

exercise. J Appl Physiol, 88, 788-92.

Britton, S. L. and Koch, L. G. (2001). Animal genetic models for complex traits of

physical capacity. Exerc Sport Sci Rev, 29, 7-14.

Britton, S. L. and Koch, L. G. (2005). Animal models of complex diseases: an

initial strategy. IUBMB Life, 57, 631-8.

Brooks, G. A. (1998). Mammalian fuel utilization during sustained exercise.

Comp Biochem Physiol B Biochem Mol Biol, 120, 89-107.

Brooks, G. A., Fahey, T.D., Baldwin, K.M. (2005). Exercise Physiology: Human

Bioenergetics and Its Applications. New York, McGraw Hill.

146

Brooks, G. A. and Mercier, J. (1994). Balance of carbohydrate and lipid utilization

during exercise: the "crossover" concept. J Appl Physiol, 76, 2253-61.

Brouns, F. and van der Vusse, G. J. (1998). Utilization of lipids during exercise in

human subjects: metabolic and dietary constraints. Br J Nutr, 79, 117-28.

Bruce, R. A., Blackmon, J. R., Jones, J. W. and Strait, G. (1963). Exercising

Testing in Adult Normal Subjects and Cardiac Patients. Pediatrics, 32,

SUPPL 742-56.

Chagnon, Y. C., Rice, T., Perusse, L., Borecki, I. B., Ho-Kim, M. A., Lacaille, M.,

Pare, C., Bouchard, L., Gagnon, J., Leon, A. S., Skinner, J. S., Wilmore, J. H.,

Rao, D. C. and Bouchard, C. (2001). Genomic scan for genes affecting body

composition before and after training in Caucasians from HERITAGE. J Appl

Physiol, 90, 1777-87.

Chen, J., Feller, G. M., Barbato, J. C., Periyasamy, S., Xie, Z. J., Koch, L. G.,

Shapiro, J. I. and Britton, S. L. (2001). Cardiac performance in inbred rat

genetic models of low and high running capacity. J Physiol, 535, 611-7.

Cheng, C. P., Freeman, G. L., Santamore, W. P., Constantinescu, M. S. and

Little, W. C. (1990). Effect of loading conditions, contractile state, and heart

rate on early diastolic left ventricular filling in conscious dogs. Circ Res, 66,

814-23.

Cheng, C. P., Igarashi, Y. and Little, W. C. (1992). Mechanism of augmented rate

of left ventricular filling during exercise. Circ Res, 70, 9-19.

147

Coleman, R. A., Lewin, T. M. and Muoio, D. M. (2000). Physiological and

nutritional regulation of enzymes of triacylglcerol synthesis. Ann. Rev. Nutr.,

20, 77-103.

Colle, E., Guttmann, R. D., Seemayer, T. A. and Michel, F. (1983). Spontaneous

diabetes mellitus syndrome in the rat. IV. Immunogenetic interactions of MHC

and non-MHC components of the syndrome. Metabolism, 32, 54-61.

Costill, D. L., Branam, G., Eddy, D. and Sparks, K. (1971). Determinants of

Marathon running success. Int Z Angew Physiol, 29, 249-54.

Costill, D. L., Thomason, H. and Roberts, E. (1973). Fractional utilization of the

aerobic capacity during distance running. Med Sci Sports, 5, 248-52.

Crabbe, J. C., Wahlsten, D. and Dudek, B. C. (1999). Genetics of mouse

behavior: interactions with laboratory environment. Science, 284, 1670-2.

Cuocolo, A., Sax, F. L., Brush, J. E., Maron, B. J., Bacharach, S. L. and Bonow,

R. O. (1990). Left ventricular hypertrophy and impaired diastolic filling in

essential hypertension. Diastolic mechanisms for systolic dysfunction during

exercise. Circulation, 81, 978-86.

Dahan, M., Aubry, N., Baleynaud, S., Ferreira, B., Yu, J. and Gourgon, R. (1995).

Influence of preload reserve on stroke volume response to exercise in

patients with left ventricular systolic dysfunction: a Doppler echocardiographic

study. J Am Coll Cardiol, 25, 680-6.

148

Dahl, L. K., Heine, M. and Tassinari, L. (1962a). Effects of chronic excess salt

ingestion. Evidence that genetic factors play an important role in susceptibility

to experimental hypertension. J Exp Med, 115, 1173-90.

Dahl, L. K., Heine, M. and Tassinari, L. (1962b). Role of genetic factors in

susceptibility to experimental hypertension due to chronic excess salt

ingestion. Nature, 194, 480-2.

Darvasi, A. (1998). Experimental strategies for the genetic dissection of complex

traits in animal models. Nat Genet, 18, 19-24.

Darvasi, A., Weinreb, A., Minke, V., Weller, J. I. and Soller, M. (1993). Detecting

marker-QTL linkage and estimating QTL gene effect and map location using a

saturated genetic map. Genetics, 134, 943-951.

Dempsey, J. A. and Wagner, P. D. (1999). Exercise-induced arterial hypoxemia.

J Appl Physiol, 87, 1997-2006.

Di Bello, V., Santoro, G., Talarico, L., Di Muro, C., Caputo, M. T., Giorgi, D.,

Bertini, A., Bianchi, M. and Giusti, C. (1996). Left ventricular function during

exercise in athletes and in sedentary men. Med Sci Sports Exerc, 28, 190-6.

Dietrich, W. F., Lander, E. S., Smith, J. S., Moser, A. R., Gould, K. A., Luongo,

C., Borenstein, N. and Dove, W. (1993). Genetic identification of Mom-1, a

major modifier locus affecting Min-induced intestinal neoplasia in the mouse.

Cell, 75, 631-9.

149

DiPetrillo, K., Tsaih, S. W., Sheehan, S., Johns, C., Kelmenson, P., Gavras, H.,

Churchill, G. A. and Paigen, B. (2004). Genetic analysis of blood pressure in

C3H/HeJ and SWR/J mice. Physiol Genomics, 17, 215-20.

DiPetrillo, K., Wang, X., Stylianou, I. M. and Paigen, B. (2005). Bioinformatics

toolbox for narrowing rodent quantitative trait loci. Trends Genet, 21, 683-92.

Doerge, R. W. (2002). Mapping and analysis of quantitative trait loci in

experimental populations. Nat Rev Genet, 3, 43-52.

Dragani, T. A. (2003). 10 Years of Mouse Cancer Modifier Loci: Human

Relevance. Cancer Res, 63, 3011-3018.

Dudley, G. A., Abraham, W. M. and Terjung, R. L. (1982). Influence of exercise

intensity and duration on biochemical adaptations in skeletal muscle. J Appl

Physiol, 53, 844-50.

Falconer, D. S. and Mackay, T. F. C. (1996). Introduction to Quantitative

Genetics. Harlow, U.K., Prentice Hall.

Feldman, M. D., Erikson, J. M., Mao, Y., Korcarz, C. E., Lang, R. M. and

Freeman, G. L. (2000). Validation of a mouse conductance system to

determine LV volume: comparison to echocardiography and crystals. Am J

Physiol Heart Circ Physiol, 279, H1698-707.

Flint, J. and Mott, R. (2001). Finding the molecular basis of quantitative traits:

successes and pitfalls. Nat Rev Genet, 2, 437-45.

Foley, T. E., Greenwood, B. N., Day, H. E., Koch, L. G., Britton, S. L. and

Fleshner, M. (2006). Elevated central monoamine receptor mRNA in rats bred

150

for high endurance capacity: implications for central fatigue. Behav Brain Res,

174, 132-42.

Frantz, S., Clemitson, J. R., Bihoreau, M. T., Gauguier, D. and Samani, N. J.

(2001). Genetic dissection of region around the Sa gene on rat chromosome

1: evidence for multiple loci affecting blood pressure. Hypertension, 38, 216-

21.

Ganong, W. F. (1999). Review of Medical Physiology, Stamford, Appleton, and

Lange.

Garenc, C., Perusse, L., Bergeron, J., Gagnon, J., Chagnon, Y. C., Borecki, I. B.,

Leon, A. S., Skinner, J. S., Wilmore, J. H., Rao, D. C. and Bouchard, C.

(2001). Evidence of LPL gene-exercise interaction for body fat and LPL

activity: the HERITAGE Family Study. J Appl Physiol, 91, 1334-40.

Garrett, M. R., Dene, H., Walder, R., Zhang, Q. Y., Cicila, G. T., Assadnia, S.,

Deng, A. Y. and Rapp, J. P. (1998). Genome scan and congenic strains for

blood pressure QTL using Dahl salt-sensitive rats. Genome Res, 8, 711-23.

Garrett, M. R. and Rapp, J. P. (2002a). Multiple blood pressure QTL on rat

Chromosome 2 defined by congenic Dahl rats. Mamm Genome, 13, 41-4.

Garrett, M. R. and Rapp, J. P. (2002b). Two closely linked interactive blood

pressure QTL on rat chromosome 5 defined using congenic Dahl rats. Physiol

Genomics, 8, 81-6.

151

Garrett, M. R., Zhang, X., Dukhanina, O. I., Deng, A. Y. and Rapp, J. P. (2001).

Two linked blood pressure quantitative trait loci on chromosome 10 defined

by dahl rat congenic strains. Hypertension, 38, 779-85.

Gibbs, R. A., Weinstock, G. M., Metzker, M. L., Muzny, D. M., Sodergren, E. J.,

Scherer, S., Scott, G., Steffen, D., Worley, K. C., Burch, P. E., Okwuonu, G.,

Hines, S., Lewis, L., DeRamo, C., Delgado, O., Dugan-Rocha, S., Miner, G.,

Morgan, M., Hawes, A., Gill, R., Celera, Holt, R. A., Adams, M. D.,

Amanatides, P. G., Baden-Tillson, H., Barnstead, M., Chin, S., Evans, C. A.,

Ferriera, S., Fosler, C., Glodek, A., Gu, Z., Jennings, D., Kraft, C. L., Nguyen,

T., Pfannkoch, C. M., Sitter, C., Sutton, G. G., Venter, J. C., Woodage, T.,

Smith, D., Lee, H. M., Gustafson, E., Cahill, P., Kana, A., Doucette-Stamm,

L., Weinstock, K., Fechtel, K., Weiss, R. B., Dunn, D. M., Green, E. D.,

Blakesley, R. W., Bouffard, G. G., De Jong, P. J., Osoegawa, K., Zhu, B.,

Marra, M., Schein, J., Bosdet, I., Fjell, C., Jones, S., Krzywinski, M.,

Mathewson, C., Siddiqui, A., Wye, N., McPherson, J., Zhao, S., Fraser, C. M.,

Shetty, J., Shatsman, S., Geer, K., Chen, Y., Abramzon, S., Nierman, W. C.,

Havlak, P. H., Chen, R., Durbin, K. J., Egan, A., Ren, Y., Song, X. Z., Li, B.,

Liu, Y., Qin, X., Cawley, S., Worley, K. C., Cooney, A. J., D'Souza, L. M.,

Martin, K., Wu, J. Q., Gonzalez-Garay, M. L., Jackson, A. R., Kalafus, K. J.,

McLeod, M. P., Milosavljevic, A., Virk, D., Volkov, A., Wheeler, D. A., Zhang,

Z., Bailey, J. A., Eichler, E. E., Tuzun, E., Birney, E., Mongin, E., Ureta-Vidal,

A., Woodwark, C., Zdobnov, E., Bork, P., Suyama, M., Torrents, D.,

152

Alexandersson, M., Trask, B. J., Young, J. M., Huang, H., Wang, H., Xing, H.,

Daniels, S., Gietzen, D., Schmidt, J., Stevens, K., Vitt, U., Wingrove, J.,

Camara, F., Mar Alba, M., Abril, J. F., Guigo, R., Smit, A., Dubchak, I., Rubin,

E. M., Couronne, O., Poliakov, A., Hubner, N., Ganten, D., Goesele, C.,

Hummel, O., Kreitler, T., Lee, Y. A., Monti, J., Schulz, H., Zimdahl, H.,

Himmelbauer, H., Lehrach, H., Jacob, H. J., Bromberg, S., Gullings-Handley,

J., Jensen-Seaman, M. I., Kwitek, A. E., Lazar, J., Pasko, D., Tonellato, P. J.,

Twigger, S., Ponting, C. P., Duarte, J. M., Rice, S., Goodstadt, L., Beatson, S.

A., Emes, R. D., Winter, E. E., Webber, C., Brandt, P., Nyakatura, G.,

Adetobi, M., Chiaromonte, F., Elnitski, L., Eswara, P., Hardison, R. C., Hou,

M., Kolbe, D., Makova, K., Miller, W., Nekrutenko, A., Riemer, C., Schwartz,

S., Taylor, J., Yang, S., Zhang, Y., Lindpaintner, K., Andrews, T. D.,

Caccamo, M., Clamp, M., Clarke, L., Curwen, V., Durbin, R., Eyras, E.,

Searle, S. M., Cooper, G. M., Batzoglou, S., Brudno, M., Sidow, A., Stone, E.

A., Venter, J. C., Payseur, B. A., Bourque, G., Lopez-Otin, C., Puente, X. S.,

Chakrabarti, K., Chatterji, S., Dewey, C., Pachter, L., Bray, N., Yap, V. B.,

Caspi, A., Tesler, G., Pevzner, P. A., Haussler, D., Roskin, K. M., Baertsch,

R., Clawson, H., Furey, T. S., Hinrichs, A. S., Karolchik, D., Kent, W. J.,

Rosenbloom, K. R., Trumbower, H., Weirauch, M., Cooper, D. N., Stenson, P.

D., Ma, B., Brent, M., Arumugam, M., Shteynberg, D., Copley, R. R., Taylor,

M. S., Riethman, H., Mudunuri, U., Peterson, J., Guyer, M., Felsenfeld, A.,

153

Old, S., Mockrin, S. and Collins, F. (2004). Genome sequence of the Brown

Norway rat yields insights into mammalian evolution. Nature, 428, 493-521.

Glazier, A. M., Nadeau, J. H. and Aitman, T. J. (2002). Finding genes that

underlie complex traits. Science, 298, 2345-9.

Goldblatt, H., Lynch, J., Hanzal, R. F. and Summerville, W. W. (1934). Studies

On Experimental Hypertension: I. The Production Of Persistent Elevation Of

Systolic Blood Pressure By Means Of Renal Ischemia. J. Exp. Med., 59, 347-

379.

Gonzalez, N. C., Howlett, R. A., Henderson, K. K., Koch, L. G., Britton, S. L.,

Wagner, H. E., Favret, F. and Wagner, P. D. (2006a). Systemic oxygen

transport in rats artificially selected for running endurance. Respir Physiol

Neurobiol, 151, 141-50.

Gonzalez, N. C., Kirkton, S. D., Howlett, R. A., Britton, S. L., Koch, L. G.,

Wagner, H. E. and Wagner, P. D. (2006b). Continued divergence in VO2 max

of rats artificially selected for running endurance is mediated by greater

convective blood O2 delivery. J Appl Physiol, 101, 1288-1296.

Goodwin, G. W. and Taegtmeyer, H. (2000). Improved energy homeostasis of

the heart in the metabolic state of exercise. Am J Physiol Heart Circ Physiol,

279, H1490-501.

Greenhouse, D. D., Festing, M. F. W., Hasan, S. and Cohen, A. L. (1990).

Genetic monitoring of inbred strains of rats. Stuttgart, Gustav Fisher.

154

Guazzi, M., Brenner, D. A., Apstein, C. S. and Saupe, K. W. (2001). Exercise

intolerance in rats with hypertensive heart disease is associated with impaired

diastolic relaxation. Hypertension, 37, 204-8.

Guryev, V., Berezikov, E. and Cuppen, E. (2005). CASCAD: a database of

annotated candidate single nucleotide polymorphisms associated with

expressed sequences. BMC Genomics, 6, 10.

Gyurko, R., Kuhlencordt, P., Fishman, M. C. and Huang, P. L. (2000). Modulation

of mouse cardiac function in vivo by eNOS and ANP. Am J Physiol Heart Circ

Physiol, 278, H971-81.

Halestrap, A. P., Kerr, P. M., Javadov, S. and Woodfield, K. Y. (1998).

Elucidating the molecular mechanism of the permeability transition pore and

its role in reperfusion injury of the heart. Biochim. Biophys. Acta, 1366, 79-94.

Hammond, H. K. and Froelicher, V. F. (1984). Exercise testing for

cardiorespiratory fitness. Sports Med, 1, 234-9.

Hansen, C. and Spuhler, K. (1984). Development of the National Institutes of

Health genetically heterogeneous rat stock. Alcohol Clin Exp Res, 8, 477-9.

Harant, I., Marion-Latard, F., Crampes, F., de Glisezinski, I., Berlan, M., Stich, V.

and Riviere, D. (2002). Effect of a long-duration physical exercise on fat cell

lipolytic responsiveness to adrenergic agents and insulin in obese men. Int J

Obes Relat Metab Disord, 26, 1373-8.

Harries, M. (1994). ABC of sports medicine. Pulmonary limitations to

performance in sport. Bmj, 309, 113-5.

155

Hayward, C. P., Cramer, E. M., Song, Z., Zheng, S., Fung, R., Masse, J. M.,

Stead, R. H. and Podor, T. J. (1998). Studies of multimerin in human

endothelial cells. Blood, 91, 1304-17.

Henderson, K. K., Wagner, H., Favret, F., Britton, S. L., Koch, L. G., Wagner, P.

D. and Gonzalez, N. C. (2002). Determinants of maximal O2 uptake in rats

selectively bred for endurance running capacity. J Appl Physiol, 93, 1265-

1274.

Hiatt, W. R. (1991). Exercise physiology in cardiovascular diseases. Curr Opin

Cardiol, 6, 745-9.

Hoit, B. D., Kiatchoosakun, S., Restivo, J., Kirkpatrick, D., Olszens, K., Shao, H.,

Pao, Y. H. and Nadeau, J. H. (2002). Naturally occurring variation in

cardiovascular traits among inbred mouse strains. Genomics, 79, 679-85.

Hoppeler, H., Altpeter, E., Wagner, M., Turner, D. L., Hokanson, J., Konig, M.,

Stalder-Navarro, V. P. and Weibel, E. R. (1995). Cold acclimation and

endurance training in guinea pigs: changes in lung, muscle and brown fat

tissue. Respir Physiol, 101, 189-98.

Hoppeler, H. and Weibel, E. R. (1998). Limits for oxygen and substrate transport

in mammals. J Exp Biol, 201, 1051-64.

Hoppeler, H. and Weibel, E. R. (2000). Structural and functional limits for oxygen

supply to muscle. Acta Physiol Scand, 168, 445-56.

156

Hoshijima, M. (2006). Mechanical stress-strain sensors embedded in cardiac

cytoskeleton: Z disk, titin, and associated structures. Am J Physiol Heart Circ

Physiol, 290, H1313-25.

Howlett, R. A., Gonzalez, N. C., Wagner, H. E., Fu, Z., Britton, S. L., Koch, L. G.

and Wagner, P. D. (2003). Genetic Models in Applied Physiology: Selected

Contribution: Skeletal muscle capillarity and enzyme activity in rats selectively

bred for running endurance. J Appl Physiol, 94, 1682-1688.

Hurley, B. F., Nemeth, P. M., Martin, W. H., 3rd, Hagberg, J. M., Dalsky, G. P.

and Holloszy, J. O. (1986). Muscle triglyceride utilization during exercise:

effect of training. J Appl Physiol, 60, 562-7.

Hussain, S. O., Barbato, J. C., Koch, L. G., Metting, P. J. and Britton, S. L.

(2001). Cardiac function in rats selectively bred for low- and high-capacity

running. Am J Physiol Regulatory Integrative Comp Physiol, 281, R1787-

1791.

Hyne, V., Kearsey, M. J., Pike, D. J. and Snape, J. W. (1995). QTL analysis:

Unreliability and bias in estimation procedures. Molec. Breeding, 1, 273-282.

Inukai, K., Funaki, M., Ogihara, T., Katagiri, H., Kanda, A., Anai, M., Fukushima,

Y., Hosaka, T., Suzuki, M., Shin, B. C., Takata, K., Yazaki, Y., Kikuchi, M.,

Oka, Y. and Asano, T. (1997). p85alpha gene generates three isoforms of

regulatory subunit for phosphatidylinositol 3-kinase (PI 3-Kinase), p50alpha,

p55alpha, and p85alpha, with different PI 3-kinase activity elevating

responses to insulin. J Biol Chem, 272, 7873-82.

157

Jacob, H. J. (1999). Functional genomics and rat models. Genome Res, 9, 1013-

Jacob, H. J. and Kwitek, A. E. (2002). Rat genetics: attaching physiology and

pharmacology to the genome. Nat Rev Genet, 3, 33-42.

James, M. R. and Lindpaintner, K. (1997). Why map the rat? Trends Genet, 13,

171-3.

Jansen, R. C. and Nap, J. P. (2001). Genetical genomics: the added value from

segregation. Trends Genet, 17, 388-91.

Janssen, I., Katzmarzyk, P. T., Ross, R., Leon, A. S., Skinner, J. S., Rao, D. C.,

Wilmore, J. H., Rankinen, T. and Bouchard, C. (2004). Fitness alters the

associations of BMI and waist circumference with total and abdominal fat.

Obes Res, 12, 525-37.

Jegger, D., Jeanrenaud, X., Nasratullah, M., Chassot, P. G., Mallik, A.,

Tevaearai, H., von Segesser, L. K., Segers, P. and Stergiopulos, N. (2006).

Noninvasive Doppler-derived myocardial performance index in rats with

myocardial infarction: validation and correlation by conductance catheter. Am

J Physiol Heart Circ Physiol, 290, H1540-8.

Jennings, G. L., Deakin, G., Dewar, E., Laufer, E. and Nelson, L. (1989).

Exercise, cardiovascular disease and blood pressure. Clin Exp Hypertens A,

11, 1035-52.

Jensen, M. D. (2006). Adipose tissue as an endocrine organ: implications of its

distribution on free fatty acid metabolism. Eur Heart J Suppl, 8, B13-19.

158

Jiang, W. H., Ma, A. Q., Zhang, Y. M., Han, K., Liu, Y., Zhang, Z. T., Wang, T. Z.,

Huang, X. and Zheng, X. P. (2005). Migration of intravenously grafted

mesenchymal stem cells to injured heart in rats. Sheng Li Xue Bao, 57, 566-

72.

Joe, B. (2007). Quest for arthritis-causative genetic factors in the rat. Physiol.

Genomics, 27, 1-11.

Joe, B. a. G., M. R. (2005). Substitution mapping: using congenic strains to

detect genes controlling blood pressure, Humana Press.

Joyner, M. J. (1991). Modeling: optimal marathon performance on the basis of

physiological factors. J Appl Physiol, 70, 683-7.

Karp, C. L., Grupe, A., Schadt, E., Ewart, S. L., Keane-Moore, M., Cuomo, P. J.,

Kohl, J., Wahl, L., Kuperman, D., Germer, S., Aud, D., Peltz, G. and Wills-

Karp, M. (2000). Identification of complement factor 5 as a susceptibility locus

for experimental allergic asthma. Nat Immunol, 1, 221-6.

Kayser, B. (2003). Exercise starts and ends in the brain. Eur J Appl Physiol, 90,

411-9.

Kelly, R. P., Ting, C. T., Yang, T. M., Liu, C. P., Maughan, W. L., Chang, M. S.

and Kass, D. A. (1992). Effective arterial elastance as index of arterial

vascular load in humans. Circulation, 86, 513-21.

Kessler, A., Uphues, I., Ouwens, D. M., Till, M. and Eckel, J. (2001).

Diversification of cardiac insulin signaling involves the p85 alpha/beta

159

subunits of phosphatidylinositol 3-kinase. Am J Physiol Endocrinol Metab,

280, E65-74.

King, G. A., Fitzhugh, E. C., Bassett, D. R., Jr., McLaughlin, J. E., Strath, S. J.,

Swartz, A. M. and Thompson, D. L. (2001). Relationship of leisure-time

physical activity and occupational activity to the prevalence of obesity. Int J

Obes Relat Metab Disord, 25, 606-12.

Koch, L. G. (2007). University of Michigan, Personal Communication.

Koch, L. G. and Britton, S. L. (2001). Artificial selection for intrinsic aerobic

endurance running capacity in rats. Physiol Genomics, 5, 45-52.

Koch, L. G. and Britton, S. L. (2005). Divergent Selection for Aerobic Capacity in

Rats as a Model for Complex Disease. Integr. Comp. Biol., 45, 405-415.

Koch, L. G., Britton, S. L., Barbato, J. C., Rodenbaugh, D. W. and Dicarlo, S. E.

(1999). Phenotypic differences in cardiovascular regulation in inbred rat

models of aerobic capacity. Physiol Genomics, 1, 63-69.

Koch, L. G., Green, C. L., Lee, A. D., Hornyak, J. E., Cicila, G. T. and Britton, S.

L. (2005). Test of the principle of initial value in rat genetic models of exercise

capacity. Am J Physiol Regulatory Integrative Comp Physiol, 288, R466-472.

Koch, L. G., Meredith, T. A., Fraker, T. D., Metting, P. J. and Britton, S. L. (1998).

Heritability of treadmill running endurance in rats. Am J Physiol Regulatory

Integrative Comp Physiol, 275, R1455-1460.

Kovacs, P. and Kloting, I. (1998). Quantitative trait loci on chromosomes 1 and 4

affect lipid phenotypes in the rat. Arch Biochem Biophys, 354, 139-43.

160

Lander, E. S., Linton, L. M., Birren, B., Nusbaum, C., Zody, M. C., Baldwin, J.,

Devon, K., Dewar, K., Doyle, M., FitzHugh, W., Funke, R., Gage, D., Harris,

K., Heaford, A., Howland, J., Kann, L., Lehoczky, J., LeVine, R., McEwan, P.,

McKernan, K., Meldrim, J., Mesirov, J. P., Miranda, C., Morris, W., Naylor, J.,

Raymond, C., Rosetti, M., Santos, R., Sheridan, A., Sougnez, C., Stange-

Thomann, N., Stojanovic, N., Subramanian, A., Wyman, D., Rogers, J.,

Sulston, J., Ainscough, R., Beck, S., Bentley, D., Burton, J., Clee, C., Carter,

N., Coulson, A., Deadman, R., Deloukas, P., Dunham, A., Dunham, I., Durbin,

R., French, L., Grafham, D., Gregory, S., Hubbard, T., Humphray, S., Hunt,

A., Jones, M., Lloyd, C., McMurray, A., Matthews, L., Mercer, S., Milne, S.,

Mullikin, J. C., Mungall, A., Plumb, R., Ross, M., Shownkeen, R., Sims, S.,

Waterston, R. H., Wilson, R. K., Hillier, L. W., McPherson, J. D., Marra, M. A.,

Mardis, E. R., Fulton, L. A., Chinwalla, A. T., Pepin, K. H., Gish, W. R.,

Chissoe, S. L., Wendl, M. C., Delehaunty, K. D., Miner, T. L., Delehaunty, A.,

Kramer, J. B., Cook, L. L., Fulton, R. S., Johnson, D. L., Minx, P. J., Clifton, S.

W., Hawkins, T., Branscomb, E., Predki, P., Richardson, P., Wenning, S.,

Slezak, T., Doggett, N., Cheng, J. F., Olsen, A., Lucas, S., Elkin, C.,

Uberbacher, E., Frazier, M., Gibbs, R. A., Muzny, D. M., Scherer, S. E.,

Bouck, J. B., Sodergren, E. J., Worley, K. C., Rives, C. M., Gorrell, J. H.,

Metzker, M. L., Naylor, S. L., Kucherlapati, R. S., Nelson, D. L., Weinstock, G.

M., Sakaki, Y., Fujiyama, A., Hattori, M., Yada, T., Toyoda, A., Itoh, T.,

Kawagoe, C., Watanabe, H., Totoki, Y., Taylor, T., Weissenbach, J., Heilig,

161

R., Saurin, W., Artiguenave, F., Brottier, P., Bruls, T., Pelletier, E., Robert, C.,

Wincker, P., Smith, D. R., Doucette-Stamm, L., Rubenfield, M., Weinstock, K.,

Lee, H. M., Dubois, J., Rosenthal, A., Platzer, M., Nyakatura, G., Taudien, S.,

Rump, A., Yang, H., Yu, J., Wang, J., Huang, G., Gu, J., Hood, L., Rowen, L.,

Madan, A., Qin, S., Davis, R. W., Federspiel, N. A., Abola, A. P., Proctor, M.

J., Myers, R. M., Schmutz, J., Dickson, M., Grimwood, J., Cox, D. R., Olson,

M. V., Kaul, R., Raymond, C., Shimizu, N., Kawasaki, K., Minoshima, S.,

Evans, G. A., Athanasiou, M., Schultz, R., Roe, B. A., Chen, F., Pan, H.,

Ramser, J., Lehrach, H., Reinhardt, R., McCombie, W. R., de la Bastide, M.,

Dedhia, N., Blocker, H., Hornischer, K., Nordsiek, G., Agarwala, R., Aravind,

L., Bailey, J. A., Bateman, A., Batzoglou, S., Birney, E., Bork, P., Brown, D.

G., Burge, C. B., Cerutti, L., Chen, H. C., Church, D., Clamp, M., Copley, R.

R., Doerks, T., Eddy, S. R., Eichler, E. E., Furey, T. S., Galagan, J., Gilbert, J.

G., Harmon, C., Hayashizaki, Y., Haussler, D., Hermjakob, H., Hokamp, K.,

Jang, W., Johnson, L. S., Jones, T. A., Kasif, S., Kaspryzk, A., Kennedy, S.,

Kent, W. J., Kitts, P., Koonin, E. V., Korf, I., Kulp, D., Lancet, D., Lowe, T. M.,

McLysaght, A., Mikkelsen, T., Moran, J. V., Mulder, N., Pollara, V. J., Ponting,

C. P., Schuler, G., Schultz, J., Slater, G., Smit, A. F., Stupka, E.,

Szustakowski, J., Thierry-Mieg, D., Thierry-Mieg, J., Wagner, L., Wallis, J.,

Wheeler, R., Williams, A., Wolf, Y. I., Wolfe, K. H., Yang, S. P., Yeh, R. F.,

Collins, F., Guyer, M. S., Peterson, J., Felsenfeld, A., Wetterstrand, K. A.,

Patrinos, A., Morgan, M. J., de Jong, P., Catanese, J. J., Osoegawa, K.,

162

Shizuya, H., Choi, S. and Chen, Y. J. (2001). Initial sequencing and analysis

of the human genome. Nature, 409, 860-921.

Lazar, J., Moreno, C., Jacob, H. J. and Kwitek, A. E. (2005). Impact of genomics

on research in the rat. Genome Res, 15, 1717-28.

Leather, H. A., Segers, P., Sun, Y. Y., De Ruyter, H. A., Vandermeersch, E. and

Wouters, P. F. (2002). The limitations of preload-adjusted maximal power as

an index of right ventricular contractility. Anesth Analg, 95, 798-804, table of

contents.

Lee, C. D., Blair, S. N. and Jackson, A. S. (1999). Cardiorespiratory fitness, body

composition, and all-cause and cardiovascular disease mortality in men. Am J

Clin Nutr, 69, 373-80.

Lee, S. J. and Cicila, G. T. (2002). Functional genomics in rat models of

hypertension: using differential expression and congenic strains to identify

and evaluate candidate genes. Crit Rev Eukaryot Gene Expr, 12, 297-316.

Lee, S. J., Liu, J., Qi, N., Guarnera, R. A., Lee, S. Y. and Cicila, G. T. (2003).

Use of a panel of congenic strains to evaluate differentially expressed genes

as candidate genes for blood pressure quantitative trait loci. Hypertens Res,

26, 75-87.

Lee, S. J., Ways, J. A., Barbato, J. C., Essig, D., Pettee, K., DeRaedt, S. J.,

Yang, S., Weaver, D. A., Koch, L. G. and Cicila, G. T. (2005). Gene

expression profiling of the left ventricles in a rat model of intrinsic aerobic

running capacity. Physiol Genomics, 23, 62-71.

163

Leimeister, C., Steidl, C., Schumacher, N., Erhard, S. and Gessler, M. (2002).

Developmental expression and biochemical characterization of Emu family

members. Dev Biol, 249, 204-18.

Lerman, I., Harrison, B. C., Freeman, K., Hewett, T. E., Allen, D. L., Robbins, J.

and Leinwand, L. A. (2002). Genetic variability in forced and voluntary

endurance exercise performance in seven inbred mouse strains. J Appl

Physiol, 92, 2245-55.

Libonati, J. R. (1999). Myocardial diastolic function and exercise. Med Sci Sports

Exerc, 31, 1741-7.

Libonati, J. R., Colby, A. M., Caldwell, T. M., Kasparian, R. and Glassberg, H. L.

(1999). Systolic and diastolic cardiac function time intervals and exercise

capacity in women. Med Sci Sports Exerc, 31, 258-63.

Lightfoot, J. T., Turner, M. J., Debate, K. A. and Kleeberger, S. R. (2001).

Interstrain variation in murine aerobic capacity. Med Sci Sports Exerc, 33,

2053-7.

Lightfoot, J. T., Turner, M. J., Kleinfehn, A. M., Jedlicka, A. E., Oshimura, T.,

Marzec, J., Gladwell, W., Leamy, L. J. and Kleeberger, S. R. (2007).

Quantitative Trait Loci (Qtl) Associated with Maximal Exercise Endurance in

Mice. J Appl Physiol.

Lin, B. F., Huang, R. F. and Shane, B. (1993). Regulation of folate and one-

carbon metabolism in mammalian cells. III. Role of mitochondrial folylpoly-

gamma-glutamate synthetase. J Biol Chem, 268, 21674-9.

164

Lindblad-Toh, K. (2004). Genome sequencing: three's company. Nature, 428,

475-6.

Lindsey, J. R. (1979). The Laboratory Rat. Burlington, Academic Press.

Lindstedt, S. L., Wells, D. J., Jones, J. H., Hoppeler, H. and Thronson, H. A., Jr.

(1988). Limitations to aerobic performance in mammals: interaction of

structure and demand. Int J Sports Med, 9, 210-7.

Lingappa, V. R. and Farey, K. (2000). Physiological Medicine: a clinical

approach to basic medical physiology, McGraw-Hill.

Little, W. C. (1985). The left ventricular dP/dtmax-end-diastolic volume relation in

closed-chest dogs. Circ Res, 56, 808-15.

Little, W. C., Kitzman, D. W. and Cheng, C. P. (2000). Diastolic dysfunction as a

cause of exercise intolerance. Heart Fail Rev, 5, 301-6.

Liu, W. M., Mei, R., Di, X., Ryder, T. B., Hubbell, E., Dee, S., Webster, T. A.,

Harrington, C. A., Ho, M. H., Baid, J. and Smeekens, S. P. (2002). Analysis of

high density expression microarrays with signed-rank call algorithms.

Bioinformatics, 18, 1593-9.

Lowes, B. D., Minobe, W., Abraham, W. T., Rizeq, M. N., Bohlmeyer, T. J.,

Quaife, R. A., Roden, R. L., Dutcher, D. L., Robertson, A. D., Voelkel, N. F.,

Badesch, D. B., Groves, B. M., Gilbert, E. M. and Bristow, M. R. (1997).

Changes in gene expression in the intact human heart. Downregulation of

alpha-myosin heavy chain in hypertrophied, failing ventricular myocardium. J

Clin Invest, 100, 2315-24.

165

Lujan, H. L., Britton, S. L., Koch, L. G. and DiCarlo, S. E. (2006). Reduced

susceptibility to ventricular tachyarrhythmias in rats selectively bred for high

aerobic capacity. Am J Physiol Heart Circ Physiol, 291, H2933-2941.

Markel, P., Shu, P., Ebeling, C., Carlson, G. A., Nagle, D. L., Smutko, J. S. and

Moore, K. J. (1997). Theoretical and empirical issues for marker-assisted

breeding of congenic mouse strains. Nat Genet, 17, 280-4.

Martin, W. H., 3rd, Dalsky, G. P., Hurley, B. F., Matthews, D. E., Bier, D. M.,

Hagberg, J. M., Rogers, M. A., King, D. S. and Holloszy, J. O. (1993). Effect

of endurance training on plasma free fatty acid turnover and oxidation during

exercise. Am J Physiol, 265, E708-14.

Masinde, G. L., Li, X., Gu, W., Davidson, H., Hamilton-Ulland, M., Wergedal, J.,

Mohan, S. and Baylink, D. J. (2002). Quantitative trait loci (QTL) for lean body

mass and body length in MRL/MPJ and SJL/J F(2) mice. Funct Integr

Genomics, 2, 98-104.

McArdle, W. D., Katch, F. I. and Katch, V. L. (1996). Exercise Physiology:

Energy, Nutrition and Human Performance. Baltimore, Wiliams & Willkins.

McMurray, R. G. and Hackney, A. C. (2005). Interactions of metabolic hormones,

adipose tissue and exercise. Sports Med, 35, 393-412.

Minato, K. (1997). Effect of endurance training on pancreatic enzyme activity in

rats. Eur J Appl Physiol Occup Physiol, 76, 491-4.

Minato, K., Shiroya, Y., Nakae, Y. and Kondo, T. (2000). The effect of chronic

exercise on the rat pancreas. Int J Pancreatol, 27, 151-6.

166

Mitchell, J. H. and Blomqvist, G. (1971). Maximal oxygen uptake. N Engl J Med,

284, 1018-22.

Mitchell, J. H., Sproule, B. J. and Chapman, C. B. (1958). The physiological

meaning of the maximal oxygen intake test. J Clin Invest, 37, 538-47.

Monti, J., Plehm, R., Schulz, H., Ganten, D., Kreutz, R. and Hubner, N. (2003).

Interaction between blood pressure quantitative trait loci in rats in which trait

variation at chromosome 1 is conditional upon a specific allele at

chromosome 10. Hum Mol Genet, 12, 435-9.

Murakami, T., Shimomura, Y., Fujitsuka, N., Sokabe, M., Okamura, K. and

Sakamoto, S. (1997). Enlargement glycogen store in rat liver and muscle by

fructose-diet intake and exercise training. J Appl Physiol, 82, 772-5.

Myers, J., Prakash, M., Froelicher, V., Do, D., Partington, S. and Atwood, J. E.

(2002). Exercise capacity and mortality among men referred for exercise

testing. N Engl J Med, 346, 793-801.

Nabika, T., Kobayashi, Y. and Yamori, Y. (2000). Congenic rats for hypertension:

how useful are they for the hunting of hypertension genes? Clin Exp

Pharmacol Physiol, 27, 251-6.

Nadeau, J. H., Singer, J. B., Matin, A. and Lander, E. S. (2000). Analysing

complex genetic traits with chromosome substitution strains. Nat Genet, 24,

221-5.

167

Naumova, A. V., Weiss, R. G. and Chacko, V. P. (2003). Regulation of murine

myocardial energy metabolism during adrenergic stress studied by in vivo

31P NMR spectroscopy. Am J Physiol Heart Circ Physiol, 285, H1976-9.

Nemoto, S., DeFreitas, G., Mann, D. L. and Carabello, B. A. (2002). Effects of

changes in left ventricular contractility on indexes of contractility in mice. Am J

Physiol Heart Circ Physiol, 283, H2504-10.

Noonan, V. and Dean, E. (2000). Submaximal exercise testing: clinical

application and interpretation. Phys Ther, 80, 782-807.

Nowak, T. J., Handford, A.G. (1994). Essentials of Pathophysiology: Concepts

and Applications for Health Care Professionals, Wm. C. Brown Publishers.

Pashmforoush, M., Pomiès, P., Peterson, K., Kubalak, S., Ross, J., Jr., Hefti, A.,

Aebi, U., Beckerle, M. and Chien, K. (2001). Adult mice deficient in actinin -

associated LIM-domain protein reveal a developmental pathway for right

ventricular cardiomyopathy. Nat. Med., 7, 591-597.

Paul, P., Letteboer, T., Gelbert, L., Groden, J., White, R. and Coppes, M. J.

(1993). Identical APC exon 15 mutations result in a variable phenotype in

familial adenomatous polyposis. Hum. Mol. Genet., 2, 925-931.

Perusse, L., Rankinen, T., Rauramaa, R., Rivera, M. A., Wolfarth, B. and

Bouchard, C. (2003). The human gene map for performance and health-

related fitness phenotypes: the 2002 update. Med Sci Sports Exerc, 35, 1248-

64.

168

Pinet, F., Poirier, F., Fuchs, S., Tharaux, P. L., Caron, M., Corvol, P., Michel, J.

B. and Joubert-Caron, R. (2004). Troponin T as a marker of differentiation

revealed by proteomic analysis in renal arterioles. Faseb J, 18, 585-6.

Plante, E., Lachance, D., Drolet, M. C., Roussel, E., Couet, J. and Arsenault, M.

(2005). Dobutamine stress echocardiography in healthy adult male rats.

Cardiovasc Ultrasound, 3, 34.

Pollock, M. L., Gaesser, G. A., Butcher, J. D., Després, J., Dishman, R. K.,

Franklin, B. A. and Garber, C. E. (1998). American College of Sports

Medicine Position Stand. The recommended quantity and quality of exercise

for developing and maintaining cardiorespiratory and muscular fitness, and

flexibility in healthy adults. Med Sci Sports Exerc, 30, 975-91.

Powell, S. R. (2006). The ubiquitin-proteasome system in cardiac physiology and

pathology. Am J Physiol Heart Circ Physiol, 291, H1-H19.

Pravenec, M. and Kurtz, T. W. (2007). Molecular genetics of experimental

hypertension and the metabolic syndrome: from gene pathways to new

therapies. Hypertension, 49, 941-52.

Rahkila, P., Soimajarvi, J., Karvinen, E. and Vihko, V. (1980). Lipid metabolism

during exercise. II. Respiratory exchange ratio and muscle glycogen content

during 4 h bicycle ergometry in two groups of healthy men. Eur J Appl Physiol

Occup Physiol, 44, 245-54.

Ranallo, R. F. and Rhodes, E. C. (1998). Lipid metabolism during exercise.

Sports Med, 26, 29-42.

169

Rankinen, T., An, P., Rice, T., Sun, G., Chagnon, Y. C., Gagnon, J., Leon, A. S.,

Skinner, J. S., Wilmore, J. H., Rao, D. C. and Bouchard, C. (2001a). Genomic

scan for exercise blood pressure in the Health, Risk Factors, Exercise

Training and Genetics (HERITAGE) Family Study. Hypertension, 38, 30-7.

Rankinen, T., Bray, M. S., Hagberg, J. M., Perusse, L., Roth, S. M., Wolfarth, B.

and Bouchard, C. (2006). The human gene map for performance and health-

related fitness phenotypes: the 2005 update. Med Sci Sports Exerc, 38, 1863-

88.

Rankinen, T., Perusse, L., Borecki, I., Chagnon, Y. C., Gagnon, J., Leon, A. S.,

Skinner, J. S., Wilmore, J. H., Rao, D. C. and Bouchard, C. (2000a). The

Na(+)-K(+)-ATPase alpha2 gene and trainability of cardiorespiratory

endurance: the HERITAGE family study. J Appl Physiol, 88, 346-51.

Rankinen, T., Perusse, L., Gagnon, J., Chagnon, Y. C., Leon, A. S., Skinner, J.

S., Wilmore, J. H., Rao, D. C. and Bouchard, C. (2000b). Angiotensin-

converting enzyme ID polymorphism and fitness phenotype in the HERITAGE

Family Study. J Appl Physiol, 88, 1029-35.

Rankinen, T., Perusse, L., Rauramaa, R., Rivera, M. A., Wolfarth, B. and

Bouchard, C. (2001b). The human gene map for performance and health-

related fitness phenotypes. Med Sci Sports Exerc, 33, 855-67.

Rankinen, T., Perusse, L., Rauramaa, R., Rivera, M. A., Wolfarth, B. and

Bouchard, C. (2002). The human gene map for performance and health-

170

related fitness phenotypes: the 2001 update. Med Sci Sports Exerc, 34, 1219-

33.

Rankinen, T., Perusse, L., Rauramaa, R., Rivera, M. A., Wolfarth, B. and

Bouchard, C. (2004). The human gene map for performance and health-

related fitness phenotypes: the 2003 update. Med Sci Sports Exerc, 36, 1451-

69.

Rapp, J. P. (1983). A paradigm for identification of primary genetic causes of

hypertension in rats. Hypertension, 5, I198-203.

Rapp, J. P. (1995). The Search for the Genetic Basis of Blood Pressure Variation

in Rats. Hypertension: Pathophysiology, Diagnosis, and Management.

Rapp, J. P. (2000). Genetic analysis of inherited hypertension in the rat. Physiol

Rev, 80, 135-72.

Rapp, J. P. and Deng, A. Y. (1995). Detection and positional cloning of blood

pressure quantitative trait loci: is it possible? Identifying the genes for genetic

hypertension. Hypertension, 25, 1121-8.

Rapp, J. P., Garrett, M. R. and Deng, A. Y. (1998). Construction of a double

congenic strain to prove an epistatic interaction on blood pressure between

rat chromosomes 2 and 10. J Clin Invest, 101, 1591-5.

Rice, T., Hong, Y., Perusse, L., Despres, J. P., Gagnon, J., Leon, A. S., Skinner,

J. S., Wilmore, J. H., Bouchard, C. and Rao, D. C. (1999). Total body fat and

abdominal visceral fat response to exercise training in the HERITAGE Family

171

Study: evidence for major locus but no multifactorial effects. Metabolism, 48,

1278-86.

Rico-Sanz, J., Rankinen, T., Rice, T., Leon, A. S., Skinner, J. S., Wilmore, J. H.,

Rao, D. C. and Bouchard, C. (2004). Quantitative trait loci for maximal

exercise capacity phenotypes and their responses to training in the

HERITAGE Family Study. Physiol Genomics, 16, 256-60.

Romijn, J. A., Coyle, E. F., Sidossis, L. S., Gastaldelli, A., Horowitz, J. F., Endert,

E. and Wolfe, R. R. (1993). Regulation of endogenous fat and carbohydrate

metabolism in relation to exercise intensity and duration. Am J Physiol, 265,

E380-91.

Ross, R., Freeman, J. A. and Janssen, I. (2000). Exercise alone is an effective

strategy for reducing obesity and related comorbidities. Exerc Sport Sci Rev,

28, 165-70.

Ross, R., Janssen, I., Dawson, J., Kungl, A. M., Kuk, J. L., Wong, S. L., Nguyen-

Duy, T. B., Lee, S., Kilpatrick, K. and Hudson, R. (2004). Exercise-induced

reduction in obesity and insulin resistance in women: a randomized controlled

trial. Obes Res, 12, 789-98.

Rossini, A. A., Like, A. A., Chick, W. L., Appel, M. C. and Cahill, G. F., Jr. (1977).

Studies of streptozotocin-induced insulitis and diabetes. Proc Natl Acad Sci U

S A, 74, 2485-9.

172

Russell, L. B. and Russell, W. L. (1992). Frequency and nature of specific-locus

mutations induced in female mice by radiations and chemicals: a review.

Mutat Res, 296, 107-27.

Saad, Y., Garrett, M. R. and Rapp, J. P. (2001). Multiple blood pressure QTL on

rat chromosome 1 defined by Dahl rat congenic strains. Physiol Genomics, 4,

201-14.

Sakane, N., Yoshida, T., Umekawa, T., Kogure, A., Takakura, Y. and Kondo, M.

(1997). Effects of Trp64Arg mutation in the beta 3-adrenergic receptor gene

on weight loss, body fat distribution, glycemic control, and insulin resistance

in obese type 2 diabetic patients. Diabetes Care, 20, 1887-90.

Saltin, B. (1985). Malleability of the system in overcoming limitations: functional

elements. J Exp Biol, 115, 345-54.

Schenk, S., Popovic, Z. B., Ochiai, Y., Casas, F., McCarthy, P. M., Starling, R.

C., Kopcak, M. W., Jr., Dessoffy, R., Navia, J. L., Greenberg, N. L., Thomas,

J. D. and Fukamachi, K. (2004). Preload-adjusted right ventricular maximal

power: concept and validation. Am J Physiol Heart Circ Physiol, 287, H1632-

40.

Schertel, E. R. (1998). Assessment of left-ventricular function. Thorac

Cardiovasc Surg, 46 Suppl 2, 248-54.

Schnermann, J. (2002). Exercise. Am J Physiol Regul Integr Comp Physiol, 283,

R2-6.

173

Schrader, J. (1990). Adenosine. A homeostatic metabolite in cardiac energy

metabolism. Circulation, 81, 389-91.

Segel, L. D. and Rendig, S. V. (1986). Sodium pentobarbital effects on cardiac

function and response to dobutamine. J Cardiovasc Pharmacol, 8, 392-7.

Shephard, R. J., Allen, C., Benade, A. J., Davies, C. T., Di Prampero, P. E.,

Hedman, R., Merriman, J. E., Myhre, K. and Simmons, R. (1968). The

maximum oxygen intake. An international reference standard of

cardiorespiratory fitness. Bull World Health Organ, 38, 757-64.

Shibasaki, F., Homma, Y. and Takenawa, T. (1991). Two types of

phosphatidylinositol 3-kinase from bovine thymus. Monomer and heterodimer

form. J Biol Chem, 266, 8108-14.

Shima, K., Zhu, M., Noma, Y., Mizuno, A., Murakami, T., Sano, T. and Kuwajima,

M. (1997). Exercise training in Otsuka Long-Evans Tokushima Fatty rat, a

model of spontaneous non-insulin-dependent diabetes mellitus: effects on the

B-cell mass, insulin content and fibrosis in the pancreas. Diabetes Res Clin

Pract, 35, 11-9.

Silver, L. M. (1995). Mouse Genetics: Concepts and Applications. Oxford,

Oxford University Press.

Snell, G. D. (1948). Methods for the study of histocompatibility genes. J Genet,

49, 87-103.

Spargo, F. J., McGee, S. L., Dzamko, N., Watt, M. J., Kemp, B. E., Britton, S. L.,

Koch, L. G., Hargreaves, M. and Hawley, J. A. (2007). Dysregulation of

174

muscle lipid metabolism in rats selectively bred for low aerobic running

capacity. Am J Physiol Endocrinol Metab, 292, E1631-6.

Stanley, W. C., Recchia, F. A. and Lopaschuk, G. D. (2005). Myocardial

substrate metabolism in the normal and failing heart. Physiol Rev, 85, 1093-

129.

Stevens, J., Cai, J., Evenson, K. R. and Thomas, R. (2002). Fitness and fatness

as predictors of mortality from all causes and from cardiovascular disease in

men and women in the lipid research clinics study. Am J Epidemiol, 156, 832-

41.

Stryer, L. (1995). Biochemistry. New York, Freeman.

Sumimoto, T., Jikuhara, T., Hattori, T., Yuasa, F., Kaida, M., Hikosaka, M.,

Takehana, K., Tamura, T., Sugiura, T. and Iwasaka, T. (1997). Importance of

left ventricular diastolic function on maintenance of exercise capacity in

patients with systolic dysfunction after anterior myocardial infarction. Am

Heart J, 133, 87-93.

Suto, J. and Sekikawa, K. (2003). Quantitative trait locus analysis of plasma

cholesterol and triglyceride levels in KK x RR F2 mice. Biochem Genet, 41,

325-41.

Troxell, M. L., Britton, S. L. and Koch, L. G. (2003). Genetic Models in Applied

Physiology: Selected Contribution: Variation and heritability for the

adaptational response to exercise in genetically heterogeneous rats. J Appl

Physiol, 94, 1674-1681.

175

Twigger, S. N., Pasko, D., Nie, J., Shimoyama, M., Bromberg, S., Campbell, D.,

Chen, J., dela Cruz, N., Fan, C., Foote, C., Harris, G., Hickmann, B., Ji, Y.,

Jin, W., Li, D., Mathis, J., Nenasheva, N., Nigam, R., Petri, V., Reilly, D.,

Ruotti, V., Schauberger, E., Seiler, K., Slyper, R., Smith, J., Wang, W., Wu,

W., Zhao, L., Zuniga-Meyer, A., Tonellato, P. J., Kwitek, A. E. and Jacob, H.

J. (2005). Tools and strategies for physiological genomics: the Rat Genome

Database. Physiol Genomics, 23, 246-56.

Ueki, K., Fruman, D. A., Yballe, C. M., Fasshauer, M., Klein, J., Asano, T.,

Cantley, L. C. and Kahn, C. R. (2003). Positive and negative roles of p85

alpha and p85 beta regulatory subunits of phosphoinositide 3-kinase in insulin

signaling. J Biol Chem, 278, 48453-66.

Van Dijk, S. J., Specht, P. A., Lazar, J., Jacob, H. J. and Provoost, A. P. (2006).

Synergistic QTL interactions between Rf-1 and Rf-3 increase renal damage

susceptibility in double congenic rats. Kidney Int, 69, 1369-76.

Vanoverschelde, J. J., Essamri, B., Vanbutsele, R., d'Hondt, A., Cosyns, J. R.,

Detry, J. R. and Melin, J. A. (1993). Contribution of left ventricular diastolic

function to exercise capacity in normal subjects. J Appl Physiol, 74, 2225-33.

Venter, J. C., Adams, M. D., Myers, E. W., Li, P. W., Mural, R. J., Sutton, G. G.,

Smith, H. O., Yandell, M., Evans, C. A., Holt, R. A., Gocayne, J. D.,

Amanatides, P., Ballew, R. M., Huson, D. H., Wortman, J. R., Zhang, Q.,

Kodira, C. D., Zheng, X. H., Chen, L., Skupski, M., Subramanian, G.,

Thomas, P. D., Zhang, J., Gabor Miklos, G. L., Nelson, C., Broder, S., Clark,

176

A. G., Nadeau, J., McKusick, V. A., Zinder, N., Levine, A. J., Roberts, R. J.,

Simon, M., Slayman, C., Hunkapiller, M., Bolanos, R., Delcher, A., Dew, I.,

Fasulo, D., Flanigan, M., Florea, L., Halpern, A., Hannenhalli, S., Kravitz, S.,

Levy, S., Mobarry, C., Reinert, K., Remington, K., Abu-Threideh, J., Beasley,

E., Biddick, K., Bonazzi, V., Brandon, R., Cargill, M., Chandramouliswaran, I.,

Charlab, R., Chaturvedi, K., Deng, Z., Di Francesco, V., Dunn, P., Eilbeck, K.,

Evangelista, C., Gabrielian, A. E., Gan, W., Ge, W., Gong, F., Gu, Z., Guan,

P., Heiman, T. J., Higgins, M. E., Ji, R. R., Ke, Z., Ketchum, K. A., Lai, Z., Lei,

Y., Li, Z., Li, J., Liang, Y., Lin, X., Lu, F., Merkulov, G. V., Milshina, N., Moore,

H. M., Naik, A. K., Narayan, V. A., Neelam, B., Nusskern, D., Rusch, D. B.,

Salzberg, S., Shao, W., Shue, B., Sun, J., Wang, Z., Wang, A., Wang, X.,

Wang, J., Wei, M., Wides, R., Xiao, C., Yan, C., Yao, A., Ye, J., Zhan, M.,

Zhang, W., Zhang, H., Zhao, Q., Zheng, L., Zhong, F., Zhong, W., Zhu, S.,

Zhao, S., Gilbert, D., Baumhueter, S., Spier, G., Carter, C., Cravchik, A.,

Woodage, T., Ali, F., An, H., Awe, A., Baldwin, D., Baden, H., Barnstead, M.,

Barrow, I., Beeson, K., Busam, D., Carver, A., Center, A., Cheng, M. L.,

Curry, L., Danaher, S., Davenport, L., Desilets, R., Dietz, S., Dodson, K.,

Doup, L., Ferriera, S., Garg, N., Gluecksmann, A., Hart, B., Haynes, J.,

Haynes, C., Heiner, C., Hladun, S., Hostin, D., Houck, J., Howland, T.,

Ibegwam, C., Johnson, J., Kalush, F., Kline, L., Koduru, S., Love, A., Mann,

F., May, D., McCawley, S., McIntosh, T., McMullen, I., Moy, M., Moy, L.,

Murphy, B., Nelson, K., Pfannkoch, C., Pratts, E., Puri, V., Qureshi, H.,

177

Reardon, M., Rodriguez, R., Rogers, Y. H., Romblad, D., Ruhfel, B., Scott, R.,

Sitter, C., Smallwood, M., Stewart, E., Strong, R., Suh, E., Thomas, R., Tint,

N. N., Tse, S., Vech, C., Wang, G., Wetter, J., Williams, S., Williams, M.,

Windsor, S., Winn-Deen, E., Wolfe, K., Zaveri, J., Zaveri, K., Abril, J. F.,

Guigo, R., Campbell, M. J., Sjolander, K. V., Karlak, B., Kejariwal, A., Mi, H.,

Lazareva, B., Hatton, T., Narechania, A., Diemer, K., Muruganujan, A., Guo,

N., Sato, S., Bafna, V., Istrail, S., Lippert, R., Schwartz, R., Walenz, B.,

Yooseph, S., Allen, D., Basu, A., Baxendale, J., Blick, L., Caminha, M.,

Carnes-Stine, J., Caulk, P., Chiang, Y. H., Coyne, M., Dahlke, C., Mays, A.,

Dombroski, M., Donnelly, M., Ely, D., Esparham, S., Fosler, C., Gire, H.,

Glanowski, S., Glasser, K., Glodek, A., Gorokhov, M., Graham, K., Gropman,

B., Harris, M., Heil, J., Henderson, S., Hoover, J., Jennings, D., Jordan, C.,

Jordan, J., Kasha, J., Kagan, L., Kraft, C., Levitsky, A., Lewis, M., Liu, X.,

Lopez, J., Ma, D., Majoros, W., McDaniel, J., Murphy, S., Newman, M.,

Nguyen, T., Nguyen, N., Nodell, M., Pan, S., Peck, J., Peterson, M., Rowe,

W., Sanders, R., Scott, J., Simpson, M., Smith, T., Sprague, A., Stockwell, T.,

Turner, R., Venter, E., Wang, M., Wen, M., Wu, D., Wu, M., Xia, A., Zandieh,

A. and Zhu, X. (2001). The sequence of the human genome. Science, 291,

1304-51.

Wakeland, E., Morel, L., Achey, K., Yui, M. and Longmate, J. (1997). Speed

congenics: a classic technique in the fast lane (relatively speaking). Immunol

Today, 18, 472-7.

178

Walker, J. P., Barbato, J. C. and Koch, L. G. (2002). Cardiac adenosine

production in rat genetic models of low and high exercise capacity. Am J

Physiol Regulatory Integrative Comp Physiol, 283, R168-173.

Wallace, D. C. (2005). A mitochondrial paradigm of metabolic and degenerative

diseases, aging, and cancer: a dawn for evolutionary medicine. Annu Rev

Genet, 39, 359-407.

Walsh, B., Hooks, R. B., Hornyak, J. E., Koch, L. G., Britton, S. L. and Hogan, M.

C. (2006). Enhanced mitochondrial sensitivity to creatine in rats bred for high

aerobic capacity. J Appl Physiol, 100, 1765-1769.

Warburton, D. E., Nicol, C. W. and Bredin, S. S. (2006). Health benefits of

physical activity: the evidence. Cmaj, 174, 801-9.

Watanabe, T. K., Bihoreau, M. T., McCarthy, L. C., Kiguwa, S. L., Hishigaki, H.,

Tsuji, A., Browne, J., Yamasaki, Y., Mizoguchi-Miyakita, A., Oga, K., Ono, T.,

Okuno, S., Kanemoto, N., Takahashi, E., Tomita, K., Hayashi, H., Adachi, M.,

Webber, C., Davis, M., Kiel, S., Knights, C., Smith, A., Critcher, R., Miller, J.,

Thangarajah, T., Day, P. J., Hudson, J. R., Jr., Irie, Y., Takagi, T., Nakamura,

Y., Goodfellow, P. N., Lathrop, G. M., Tanigami, A. and James, M. R. (1999).

A radiation hybrid map of the rat genome containing 5,255 markers. Nat

Genet, 22, 27-36.

Waterston, R. H., Lindblad-Toh, K., Birney, E., Rogers, J., Abril, J. F., Agarwal,

P., Agarwala, R., Ainscough, R., Alexandersson, M., An, P., Antonarakis, S.

E., Attwood, J., Baertsch, R., Bailey, J., Barlow, K., Beck, S., Berry, E., Birren,

179

B., Bloom, T., Bork, P., Botcherby, M., Bray, N., Brent, M. R., Brown, D. G.,

Brown, S. D., Bult, C., Burton, J., Butler, J., Campbell, R. D., Carninci, P.,

Cawley, S., Chiaromonte, F., Chinwalla, A. T., Church, D. M., Clamp, M.,

Clee, C., Collins, F. S., Cook, L. L., Copley, R. R., Coulson, A., Couronne, O.,

Cuff, J., Curwen, V., Cutts, T., Daly, M., David, R., Davies, J., Delehaunty, K.

D., Deri, J., Dermitzakis, E. T., Dewey, C., Dickens, N. J., Diekhans, M.,

Dodge, S., Dubchak, I., Dunn, D. M., Eddy, S. R., Elnitski, L., Emes, R. D.,

Eswara, P., Eyras, E., Felsenfeld, A., Fewell, G. A., Flicek, P., Foley, K.,

Frankel, W. N., Fulton, L. A., Fulton, R. S., Furey, T. S., Gage, D., Gibbs, R.

A., Glusman, G., Gnerre, S., Goldman, N., Goodstadt, L., Grafham, D.,

Graves, T. A., Green, E. D., Gregory, S., Guigo, R., Guyer, M., Hardison, R.

C., Haussler, D., Hayashizaki, Y., Hillier, L. W., Hinrichs, A., Hlavina, W.,

Holzer, T., Hsu, F., Hua, A., Hubbard, T., Hunt, A., Jackson, I., Jaffe, D. B.,

Johnson, L. S., Jones, M., Jones, T. A., Joy, A., Kamal, M., Karlsson, E. K.,

Karolchik, D., Kasprzyk, A., Kawai, J., Keibler, E., Kells, C., Kent, W. J., Kirby,

A., Kolbe, D. L., Korf, I., Kucherlapati, R. S., Kulbokas, E. J., Kulp, D.,

Landers, T., Leger, J. P., Leonard, S., Letunic, I., Levine, R., Li, J., Li, M.,

Lloyd, C., Lucas, S., Ma, B., Maglott, D. R., Mardis, E. R., Matthews, L.,

Mauceli, E., Mayer, J. H., McCarthy, M., McCombie, W. R., McLaren, S.,

McLay, K., McPherson, J. D., Meldrim, J., Meredith, B., Mesirov, J. P., Miller,

W., Miner, T. L., Mongin, E., Montgomery, K. T., Morgan, M., Mott, R.,

Mullikin, J. C., Muzny, D. M., Nash, W. E., Nelson, J. O., Nhan, M. N., Nicol,

180

R., Ning, Z., Nusbaum, C., O'Connor, M. J., Okazaki, Y., Oliver, K., Overton-

Larty, E., Pachter, L., Parra, G., Pepin, K. H., Peterson, J., Pevzner, P.,

Plumb, R., Pohl, C. S., Poliakov, A., Ponce, T. C., Ponting, C. P., Potter, S.,

Quail, M., Reymond, A., Roe, B. A., Roskin, K. M., Rubin, E. M., Rust, A. G.,

Santos, R., Sapojnikov, V., Schultz, B., Schultz, J., Schwartz, M. S.,

Schwartz, S., Scott, C., Seaman, S., Searle, S., Sharpe, T., Sheridan, A.,

Shownkeen, R., Sims, S., Singer, J. B., Slater, G., Smit, A., Smith, D. R.,

Spencer, B., Stabenau, A., Stange-Thomann, N., Sugnet, C., Suyama, M.,

Tesler, G., Thompson, J., Torrents, D., Trevaskis, E., Tromp, J., Ucla, C.,

Ureta-Vidal, A., Vinson, J. P., Von Niederhausern, A. C., Wade, C. M., Wall,

M., Weber, R. J., Weiss, R. B., Wendl, M. C., West, A. P., Wetterstrand, K.,

Wheeler, R., Whelan, S., Wierzbowski, J., Willey, D., Williams, S., Wilson, R.

K., Winter, E., Worley, K. C., Wyman, D., Yang, S., Yang, S. P., Zdobnov, E.

M., Zody, M. C. and Lander, E. S. (2002). Initial sequencing and comparative

analysis of the mouse genome. Nature, 420, 520-62.

Ways, J. A., Cicila, G. T., Garrett, M. R. and Koch, L. G. (2002). A genome scan

for Loci associated with aerobic running capacity in rats. Genomics, 80, 13-

20.

Ways, J. A., Smith, B. M., Barbato, J. C., Ramdath, R. S., Pettee, K. M.,

DeRaedt, S. J., Allison, D. C., Koch, L. G., Lee, S. J. and Cicila, G. T. (2007).

Congenic strains confirm aerobic running capacity quantitative trait loci on rat

181

chromosome 16 and identify possible intermediate phenotypes. Physiol

Genomics, 29, 91-7.

Weber, J. L. and May, P. E. (1989). Abundant class of human DNA

polymorphisms which can be typed using the polymerase chain reaction. Am

J Hum Genet, 44, 388-96.

Wei, M., Kampert, J. B., Barlow, C. E., Nichaman, M. Z., Gibbons, L. W.,

Paffenbarger, R. S., Jr. and Blair, S. N. (1999). Relationship between low

cardiorespiratory fitness and mortality in normal-weight, overweight, and

obese men. Jama, 282, 1547-53.

Wendling, P. S., Peters, S. J., Heigenhauser, G. J. and Spriet, L. L. (1996).

Epinephrine infusion does not enhance net muscle glycogenolysis during

prolonged aerobic exercise. Can J Appl Physiol, 21, 271-84.

Wilder, S. P., Bihoreau, M. T., Argoud, K., Watanabe, T. K., Lathrop, M. and

Gauguier, D. (2004). Integration of the rat recombination and EST maps in

the rat genomic sequence and comparative mapping analysis with the mouse

genome. Genome Res, 14, 758-65.

Winder, W. W., Hickson, R. C., Hagberg, J. M., Ehsani, A. A. and McLane, J. A.

(1979). Training-induced changes in hormonal and metabolic responses to

submaximal exercise. J Appl Physiol, 46, 766-71.

Wisloff, U., Najjar, S. M., Ellingsen, O., Haram, P. M., Swoap, S., Al-Share, Q.,

Fernstrom, M., Rezaei, K., Lee, S. J., Koch, L. G. and Britton, S. L. (2005).

182

Cardiovascular risk factors emerge after artificial selection for low aerobic

capacity. Science, 307, 418-420.

Wolfarth, B., Bray, M. S., Hagberg, J. M., Perusse, L., Rauramaa, R., Rivera, M.

A., Roth, S. M., Rankinen, T. and Bouchard, C. (2005). The human gene map

for performance and health-related fitness phenotypes: the 2004 update. Med

Sci Sports Exerc, 37, 881-903.

Wong, S. L., Katzmarzyk, P., Nichaman, M. Z., Church, T. S., Blair, S. N. and

Ross, R. (2004). Cardiorespiratory fitness is associated with lower abdominal

fat independent of body mass index. Med Sci Sports Exerc, 36, 286-91.

Zachariae, W. and Nasmyth, K. (1999). Whose end is destruction: cell division

and the anaphase-promoting complex. Genes Dev., 13, 2039-2058.

Zimdahl, H., Nyakatura, G., Brandt, P., Schulz, H., Hummel, O., Fartmann, B.,

Brett, D., Droege, M., Monti, J., Lee, Y. A., Sun, Y., Zhao, S., Winter, E. E.,

Ponting, C. P., Chen, Y., Kasprzyk, A., Birney, E., Ganten, D. and Hubner, N.

(2004). A SNP map of the rat genome generated from cDNA sequences.

Science, 303, 807.

Zolotareva, A. G. and Kogan, M. E. (1978). Production of experimental occlusive

myocardial infarction in mice. Cor Vasa, 20, 308-14.

183

APPENDIX A

Alternative representation of Figure 7 showing the physical maps of the congenic regions carried by COP.DA(chr 3) and the two RNO16 consomic strains. Solid bars represent transferred regions in the congenic and consomic strains, open bars represent chromosomal regions where crossovers have occurred marking the boundaries of the congenic region, and solid lines represent the length of the entire chromosome. Markers located at or near QTL peaks, as described in Ways et al. (2002), are underlined and in bold. Map distances expressed in megabases (Mb). The complete set of autosomes and where the congenic strains are located in relation to the rest of the genome are displayed.

184

APPENDIX B

All Variables Measured for the COP, DA, and Consomic Hemodynamic Response to Dobutamine

COP COP.DA(chr 16) DA DA.COP(chr 16) F Strain (n = 8) (n = 8) (n = 10) (n = 10) Statistic HR (bpm) 11.3 ± 2.7 15.3 ± 2.2 6.8 ± 1.2 7.9 ± 0.7 4.62

SV (µL) 16.6 ± 7.6 32.8 ± 8.8 26.6 ± 5.3 34.8 ± 6.6 1.31

SVI-bw (µL/kg) 16.6 ± 7.6 32.8 ± 8.8 26.6 ± 5.3 34.8 ± 6.6 1.31

SVI-hw (µl/g) 16.6 ± 7.6 32.8 ± 8.8 26.6 ± 5.3 34.8 ± 6.6 1.31

CO (mL/min) 30.0 ± 9.9 51.8 ± 8.7 35.5 ± 6.6 45.6 ± 7.3 1.39

CI-bw (µl/min/g) 30.0 ± 9.9 51.8 ± 8.7 35.5 ± 6.6 45.6 ± 7.3 1.39 CI-hw 30.0 ± 9.9 51.8 ± 8.7 35.5 ± 6.6 45.6 ± 7.3 1.39 (µl/min/mg) ESP (mmHg) #1 -7.5 ± 8.4 -13.9 ± 5.6 -8.2 ± 2.1 -10.7 ± 3.0 0.35

ESV (µL) #1 -22.3 ± 11.9 -49.4 ± 10.0 -33.6 ± 9.2 -43.0 ± 10.6 1.13

SW (mmHg·mL) 19.2 ± 8.9 31.6 ± 10.7 33.5 ± 5.9 42.2 ± 7.4 1.34 SWI-bw 19.2 ± 8.9 31.6 ± 10.7 33.5 ± 5.9 42.2 ± 7.4 1.34 (mmHg·µl/g) SWI-hw 19.2 ± 8.9 31.6 ± 10.7 33.5 ± 5.9 42.2 ± 7.4 1.34 (mmHg·µl/mg) EF (%) #1 3.1 ± 1.6 19.2 ± 5.1** 17.7 ± 4.5 15.8 ± 5.2 2.10 +dP/dt 29.0 ± 10.2 45.6 ± 11.3 42.0 ± 7.7 52.5 ± 6.9 1.20 (mmHg/sec) Cont. Index 2 30.9 ± 12.1 46.9 ± 12.9 38.7 ± 8.6 49.2 ± 6.0 0.70 (mmHg /sec) V@+dP/dt (µL) 8.2 ± 4.3 14.0 ± 9.2 8.3 ± 2.2 18.8 ± 3.8*** 1.05 PWRmax #2, #3 27.5 ± 8.3 15.4 ± 12.2 37.8 ± 8.3 49.5 ± 8.5 2.44 (mWatts) PAMP 2 26.3 ± 9.5 19.8 ± 19.2 32.6 ± 10.4 16.6 ± 10.2 0.35 (mWatts/µL ) EDP (mmHg) 10.5 ± 8.9 2.1 ± 4.4 2.9 ± 2.9 15.0 ± 4.5 1.45

EDV (µL) 1.8 ± 3.7 8.8 ± 8.7 5.7 ± 2.0 15.1 ± 3.9 1.36 -dP/dt -7.9 ± 7.2 1.1 ± 10.5 -8.2 ± 4.0 -14.2 ± 5.4 0.85 (mmHg/sec) Tau (msec) -10.8 ± 5.8 -18.1 ± 2.2 -7.6 ± 2.7 -6.7 ± 2.4 2.26

Ea (mmHg/µL) #1 -20.9 ± 9.0 -32.5 ± 5.6 -25.9 ± 4.2 -32.1 ± 4.4 0.88

185

ABSTRACT

Aerobic exercise tests measure the integrative ability of multiple physiological

systems to adapt to acute aerobic exercise and are often used to determine physical fitness, assess overall health, and predict mortality. While the associations between physical fitness, aerobic performance, and overall health are well known, the underlying genetic factors involved are poorly understood.

Identifying these underlying genetic factors is therefore an essential step toward a greater understanding of the relationship between physical fitness and health.

Two inbred rat strains divergent for treadmill aerobic running capacity (ARC)

were previously identified, the low performing Copenhagen (COP) and the high

performing DA rats. An F2(COPxDA) population was used to identify ARC

quantitative trait loci (QTLs) on rat chromosome 16 (RNO16) and the proximal

portion of rat chromosome 3 (RNO3). Three congenic rat strains were bred to

further investigate these ARC QTLs by transferring RNO16 and the proximal

portion of RNO3 from DA rats and RNO16 from COP rats into the respective

genetic backgrounds of COP and DA rats. These rat strains were named

COP.DA(chr 16), COP.DA(chr 3), and DA.COP(chr 16), respectively.

COP.DA(chr 16) and DA.COP(chr 16) rats had significantly different ARC

compared to COP and DA rats, respectively. COP.DA(chr 3) rats had increased,

although not significant, ARC compared to COP rats. Subcutaneous abdominal

fat, fasting triglyceride concentrations, cardiac diastolic function, and exercise-

induced changes in energy substrates were identified as potential intermediate

phenotypes to help explain the differences in exercise performance between

186

COP and DA strains. While the colocalization of these phenotypes and the ARC

QTLs may be coincidental, it is also possible that these differences in energy balance and cardiac function may be associated with the superior running performance of DA rats. Gene expression and molecular network analysis identified twelve potential gene candidates as being the underlying genetic determinants of the intermediate phenotype and ARC strain differences observed between COP, DA, and their respective congenic strains.

187