The influence of host diversity, transmission and microbiome on virulence evolution

The harm caused to a host (virulence) is an important aspect to any pathogenic infection and is influenced by many different factors. Here I seek to understand how three of these factors, host genetic diversity, transmission, and gut microbial diversity, influence the virulence of a murine specific retrovirus, Friend Virus Complex (FVC). Chapter 1 explores the effect of major histocompatibility complex (MHC) diversity on virulence. Using serial passage of FVC, through either MHC similar or MHC dissimilar mice, I show that there is a significant reduction of both fitness and virulence of FVC when a dissimilar genotype is seen than when FVC is passaged through the same genotype; suggesting that MHC diversity is an impediment to virulence evolution. Furthermore, the alternating patterns reemerged after infection with a virus adapted to resistant animals that initially swamped the alternating effect, providing evidence for negative frequency-dependent selection maintaining MHC diversity in host populations. Chapter 2 elucidates the influences natural transmission, sex, and social status have on virulence using wild-derived contact (initially uninfected) and index (initially infected) animals in seminatural enclosures. Male-male transmission is the predominant mode of transmission as minimal female transmission and no vertical transmission was observed. Moreover, natural transmission is an impediment to FVC replication as infected contact animals had lower viral titer and virulence than index animals. Finally, though dominant and nondominant males contract the virus at similar rates and experience similar virulence, nondominant animals have higher titers. Chapter 3 seeks to understand how the microbiome influences pathogen virulence. After antibiotic treatment, animals of two different MHC congenic genotypes were reconstituted with gut microbiota from a donor of their own MHC genotype (native) or from a donor with a different MHC genotype (novel). After challenge with FVC, significantly higher titers were seen in animals receiving novel microbiota than animals receiving native microbiomes. There was only a shift down in total T-lymphocyte number in novel groups as no other cell subsets tested showed a change in abundance. The work presented here allows us to gain a better understanding of how virulence is impacted by a multitude of different forces, and that many different aspects need to be taken into account when trying to determine the evolution of virulence.

THE INFLUENCE OF HOST DIVERSITY, TRANSMISSION AND MICROBIOME ON VIRULENCE EVOLUTION by Douglas H. Cornwall A dissertation submitted to the faculty of The University of Utah in partial fulfillment of the requirements for the degree of Doctor of Philosophy Department of Biology The University of Utah August 2017 Copyright © Douglas H. Cornwall 2017 All Rights Reserved The University of Utah Graduate School STATEMENT OF DISSERTATION APPROVAL The dissertation of Douglas H. Cornwall has been approved by the following supervisory committee members: Wayne K. Potts and by , Chair 5-3-2017 Frederick R. Adler , Member 5-3-2017 Sarah Elizabeth Bush , Member 5-3-2017 Kim Hasenkrug , Member 5-3-2017 Jon Seger , Member 5-3-2017 M. Denise Dearing the Department/College/School of and by David B. Kieda, Dean of The Graduate School. Date Approved Date Approved Date Approved Date Approved Date Approved , Chair/Dean of Biology ABSTRACT The harm caused to a host (virulence) is an important aspect to any pathogenic infection and is influenced by many different factors. Here I seek to understand how three of these factors, host genetic diversity, transmission, and gut microbial diversity, influence the virulence of a murine specific retrovirus, Friend Virus Complex (FVC). Chapter 1 explores the effect of major histocompatibility complex (MHC) diversity on virulence. Using serial passage of FVC, through either MHC similar or MHC dissimilar mice, I show that there is a significant reduction of both fitness and virulence of FVC when a dissimilar genotype is seen than when FVC is passaged through the same genotype; suggesting that MHC diversity is an impediment to virulence evolution. Furthermore, the alternating patterns reemerged after infection with a virus adapted to resistant animals that initially swamped the alternating effect, providing evidence for negative frequency-dependent selection maintaining MHC diversity in host populations. Chapter 2 elucidates the influences natural transmission, sex, and social status have on virulence using wild-derived contact (initially uninfected) and index (initially infected) animals in seminatural enclosures. Male-male transmission is the predominant mode of transmission as minimal female transmission and no vertical transmission was observed. Moreover, natural transmission is an impediment to FVC replication as infected contact animals had lower viral titer and virulence than index animals. Finally, though dominant and nondominant males contract the virus at similar rates and experience similar virulence, nondominant animals have higher titers. Chapter 3 seeks to understand how the microbiome influences pathogen virulence. After antibiotic treatment, animals of two different MHC congenic genotypes were reconstituted with gut microbiota from a donor of their own MHC genotype (native) or from a donor with a different MHC genotype (novel). After challenge with FVC, significantly higher titers were seen in animals receiving novel microbiota than animals receiving native microbiomes. There was only a shift down in total T-lymphocyte number in novel groups as no other cell subsets tested showed a change in abundance. The work presented here allows us to gain a better understanding of how virulence is impacted by a multitude of different forces, and that many different aspects need to be taken into account when trying to determine the evolution of virulence. iv Dedicated to my wife Kylie and my family. Thank you for always being there. TABLE OF CONTENTS ABSTRACT....................................................................................................................... iii LIST OF TABLES ........................................................................................................... viii LIST OF FIGURES ........................................................................................................... ix ACKNOWLEDGMENTS ................................................................................................. xi Chapters 1 INTRODUCTION ........................................................................................................... 1 1.1 Chapter Summaries .............................................................................................. 6 1.2 References .......................................................................................................... 10 2 EXPERIMENTAL MANIPULATION OF POPULATION-LEVEL MHC DIVERSITY CONTROLS PATHOGEN VIRULENCE EVOLUTION IN MUS MUSCULUS ........... 14 2.1 Abstract .............................................................................................................. 14 2.2 Introduction ........................................................................................................ 15 2.3 Methods.............................................................................................................. 18 2.4 Results ................................................................................................................ 23 2.5 Discussion .......................................................................................................... 25 2.6 References .......................................................................................................... 31 2.7 Supplementary Material ..................................................................................... 40 3 NATURAL HORIZONTAL TRANSMISSION OF A MURINE RETROVIRUS WITHIN SEMINATURAL POPULATIONS OF MICE ................................................. 43 3.1 Abstract .............................................................................................................. 43 3.2 Introduction ........................................................................................................ 44 3.3 Methods and Materials ....................................................................................... 48 3.4 Results ................................................................................................................ 56 3.5 Discussion .......................................................................................................... 58 3.6 References .......................................................................................................... 63 3.7 Supplementary Material ..................................................................................... 72 4 MICROBIAL TRANSPLANT INCREASES SUSCEPTIBILITY TO A SYSTEMIC VIRAL INFECTION ........................................................................................................ 81 4.1 Abstract .............................................................................................................. 81 4.2 Introduction ........................................................................................................ 82 4.3 Methods.............................................................................................................. 85 4.4 Results ................................................................................................................ 88 4.5 Discussion .......................................................................................................... 89 4.6 References .......................................................................................................... 92 Appendices A: SERIAL INFECTION OF DIVERSE HOST (MUS) GENOTYPES RAPIDLY IMPEDES FITNESS AND VIRULENCE ..................................................................... 101 B: SEXUAL SELECTION CONSTRAINS BODY MASS OF FEMALE BUT NOT MALE MICE .................................................................................................................. 111 C: MHC HETEROZYGOTE DISADVANTAGE DURING INFECTION OF FRIEND VIRUS COMPLEX ........................................................................................................ 117 vii LIST OF TABLES 2S.1. Linear mixed model comparing treatment groups (Pure and Alternated), controlling for host genotype ………………………………..…………..……… 42 3S.1 Summary of LM results concerning viral titers and spleen mass of caged animals ……………………………………………………………….…... 75 3S.2 Summary of LMMs concerning the virulence and titer of F-MuLV, of contact mice by population design ……………………………………….…. 76 3S.3 Summary of GLMM results for FVC transmission dynamics in seminatural enclosures ………………………………………………………………………. 77 3S.4 Summary of LMM results concerning titer and virulence measures between index and contact animals …………………………………….….... 78 3S.5 Summary table of correlation of titer and virulence in contact animals ………....79 3S.6 Summary of GLMMs of male social status on F-MuLV titer and viruelnce …....80 4.1 Linear models comparing treatment groups (native and novel), and host genotype ………………………………………………………………….....99 4.2 Linear models comparing treatment groups (native and novel), and host genotype of proportion immune cell subsets ………………….…….………. 100 LIST OF FIGURES 2.1. Serial passage of FVC through MHC congenic mice ...….………………...……. 36 2.2 Virulence and titer of FVC in test phase among all experimental treatments .…….…………………………………………..…………………….. 37 2.3 Alternating MHC genotypes reduces both titer and virulence of FVC Complex ………………………………………………….……………………... 38 2.4 Viral titer and virulence using the Balb/cbb adapted virus ……………………. 39 2S.1 Correlation between SFFV and F-MuLV titers ………………………………... 40 2S.2 Virulence and titer of SFFV in the test phase among all experimental treatments ……..……………………………………………………….………... 41 3.1 Seminatural enclosure and population design ………………………………... 68 3.2 Transmission dynamics of F-MuLV and SFFV in seminatural enclosures by contact and index mouse sex ..…………………………………... 69 3.3 Spleen mass, F-MuLV titers, and their correlation in female and male mice from seminatural enclosures ………………………………………... 70 3.4 Influence of male social dominance on F-MuLV and SFFV contraction, virulence, and titer in seminatural enclosures……………………... 71 3S.1 Gel showing the ~170bp SFFV fragment …………..……………………......... 72 3S.2 Spleen mass and F-MuLV titer in female and male caged mice ……………..... 73 3S.3 Influence of population design on virulence and titer .……………………….... 74 4.1 Microbiota transfer experimental design ……...………………………………... 96 4.2 Virulence and fitness of FVC in hosts receiving either native of novel microbiomes ………………………….…………………….…………... 97 4.3 Immune cell subset proportions of total cells or total lymphocytes……….…... 98 C.1 Spleen mass and proviral estimates by treatment group in round 1 passages ……………………………………………………………..….……... 122 C.2 Virulence estimates in the second and third round of infections …...….……... 123 x ACKNOWLEDGMENTS First, I would like to thank my advisor Wayne Potts for taking a chance on a less than stellar undergraduate and giving me the opportunity to learn and become the scientist that I am. Second, all of the lab members past and present that helped me through and were always available when I hit a road block in my science. I would especially like to thank James Ruff, Jason Kubinak, and Earl Middlebrook, I really don't know if I would have made it through without your help. I would also like to give a special thanks to Linda Morrison who always had an answer for me when something was going terribly wrong. I would also like to thank my committee, Drs. Frederick Adler, Sarah Bush, Kim Hasenkrug, and Jon Seger, each one of them made my science better and challenged me to think and write more like a scientist. Finally, I would like to thank all of the undergraduates, especially Elisabeth Zachary and Sophie Trujillo whom I was able to mentor, they all made my life a lot easier and I cannot express my gratitude for what they were all able to accomplish. This work was supported by the National Institutes of Health. CHAPTER 1 INTRODUCTION Virulence is an important component of any pathogenic infection, as it is the harm caused to the host by the pathogen. Moreover, there are many factors that lead to virulence, or lack of virulence, within a host; however, many of these components are still unknown or do not have sufficient testing in the literature (Alizon et al., 2009). Genetic diversity in a host population, transmission patterns, and most recently, microbial diversity have all been suggested to control virulence of pathogens and are the focus of this dissertation (Ebert & Bull, 2008; Koch & Schmid-Hempel, 2012; Buffie & Pamer, 2013). Conversely, the evolutionary pressures leading to virulence have also been thought to influence diversity in host populations through antagonistic interactions between host and pathogen (Salathé et al., 2008; Brockhurst et al., 2014). Since the discovery of the major histocompatibility complex (MHC) in the late 1940s, there has been strong interest in the unprecedented genetic diversity found at the classic peptide-antigen presenting genes within the complex (Snell, 1948). JBS Haldane hypothesized that the purpose of the diversity of these cell surface proteins may be to combat the constant barrage of infectious pathogens (Haldane, 1949). Haldane's hypothesis was later refined and dubbed the Red Queen (RQ) hypothesis (Van Valen, 1973; Bodmer et al., 1986). The RQ hypothesizes that as pathogens adapt to a single 2 host, they become more specialized on that host genotype. It follows that those pathogens will then have a more difficult time replicating in hosts with dissimilar host genotypes. Through these interactions, and with varying amounts of allele frequencies in host populations, rare beneficial alleles (initially derived through mutations) will rise in frequency due to the decline of common alleles to which pathogens are adapting. This cyclical process is termed frequency-dependent selection and is one hypothesis for how MHC diversity is maintained in host populations (Apanius et al., 1997; Sutton et al., 2011; Brockhurst et al., 2014). During MHC-specific RQ coevolution, the pathogen has a selective advantage when mutations that abolish MHC binding to a particular peptide occur. However, if there are more alleles capable of recognizing a higher number of viral peptides and these alleles are distributed throughout a population, this MHC diversity can contribute to counteracting the viral fitness advantage; because escape from one allele does not mean escape from all alleles (Finlay & McFadden, 2006). Classical MHC genes have been shown to be associated with pathogen load, where an increased number of alleles in a host population correlates with a lower number of parasites and pathogens affecting that population (Jeffery & Bangham, 2000; Radwan et al., 2010). Furthermore, serial passage studies have shown that the two main tenets of the RQ hold true: genotype specific adaption occurs and there is a tradeoff in pathogen virulence when new genotypes are seen (Ebert, 1998; Kubinak et al., 2012, 2013). MHC diversity in a population has also been suggested to have a major role in defense against pathogen adaptation (Kubinak et al., 2013). Nonetheless, it is still unknown how rapid this host genotype specific adaptation is and to what extent MHC plays a role when adaptation time is limited, as 3 when a pathogen is moving through a population. The vertebrate immune system consists of two different, but interconnected responses, the innate and adaptive. Innate immune responses consist of an immediate broad response to the foreign invader, usually within a few hours of infection (Abbas et al., 2012). The adaptive response, on the other hand, is a highly specific response usually driven by T-lymphocyte recognition of MHC presenting a foreign peptide (protein segments that have been degraded by cellular machinery) (Germain, 1994). After intracellular binding of MHC to a peptide, the MHC-peptide complex is presented on the outside of the cell to be interrogated by circulating T-lymphocytes (T-cells). Following recognition by a T-cell receptor (TCR), that T-cell clonally proliferates and becomes activated for cytopathic activities, allowing for recognition and destruction of other cells presenting the same MHC-peptide complex (Watts, 1997). If a viral pathogen mutates part of the peptide that the MHC molecule is binding (anchor positions), it will not be presented and will therefore escape immune recognition through that foreign peptide, usually leading to an increase in viral fitness (Finlay & McFadden, 2006). The studies in this dissertation utilize Friend Virus Complex (FVC) as a mousespecific pathogen to test various hypotheses concerning factors that influence virulence evolution. FVC is a murine specific gamma retrovirus that preferentially infects dividing erythroblasts, leading to enlargement of the spleen and liver (Friend, 1957). Given that one site of FVC replication is the spleen, it is ideal to use for serial passage studies as it is easily titered and passaged. FVC is referred to as a complex because it consists of two viral genomes that work together synergistically to cause disease. The first genome, Friend Murine Leukemia Virus (F-MuLV) is a replication competent retrovirus that can 4 complete its entire life cycle independent of other genomes or pathogens (Hasenkrug & Chesebro, 1997). Spleen Focus Forming Virus (SFFV), on the other hand, cannot complete its life cycle without the help of F-MuLV, as the envelope of the F-MuLV complements SFFV and allows SFFV to exit the cell. Though SFFV cannot produce the necessary proteins that allow it to form a capsid and exit the cell, it can stimulate the erythropoietin receptor and causes uncontrolled proliferation of the infected erythroblast (Ruscetti, 1999). In this way, SFFV creates more target cells for F-MuLV, which in turn complements SFFV, allowing for a synergistic interaction that leads to higher fitness for both viral genomes. Inbred laboratory mice are powerful tools to understand how pathogen virulence is influenced by a multitude of factors. MHC congenic strains are identical throughout the entire genome with the exception of the three megabase gene region of the MHC. Utilizing three strains of MHC congenic mice (Balb/cdd, Balb/ckk, and Balb/cbb) allowed interrogation of the relationship between MHC, pathogen adaptation, and the influence of gut microbiota on virulence. Though inbred congenic animals are a powerful tool in controlled laboratory settings, they do not allow for natural transmission of FVC at a level useful for experimentation, due to impaired functional behaviors (Nelson et al., 2013; Chalfin et al., 2014). Therefore, I also utilized wild-derived mice that engage in natural behaviors when placed in seminatural enclosures (Potts et al., 1991; Meagher et al., 2000; Ruff et al., 2012). Wild-derived mice were useful in understanding the natural history of FVC, as well as allowing investigation of how natural transmission influences virulence of FVC. This dissertation tests three hypotheses related to the virulence of a murine 5 retrovirus. First, if diversity throughout the genome is capable of impeding pathogen virulence in just one round, it follows that a single alternation of an MHC host genotype will impede FVC fitness and virulence when compared to a passage occurring through the same MHC host genotype due to its role in virulence evolution (Kubinak et al., 2013, 2015a). The second set of hypotheses regard how natural transmission, host sex and social status influence virulence of FVC. FVC is thought to be transmitted through bodily fluids (Mirand et al., 1966; Portis et al., 1987), suggesting there may be differential transmission efficiency between the sexes with males transmitting more due to their antagonistic encounters. In addition, natural transmission barriers are expected to reduce overall rates of virulence evolution of FVC relative to investigator transmission. The third hypothesis regards microbiome diversity. It has been recently shown that MHC genotype sculpts the composition of the host gut microbiota and that transferring microbiota between two animals can shift susceptibility to enteric pathogens (Stecher et al., 2005; Kubinak et al., 2015b). Furthermore, transferring gut microbiome causes changes in some immune cell subsets (Cebra et al., 1998; Mazmanian et al., 2005; Zaph et al., 2008; Hensley-McBain et al., 2016). These findings suggest two possible hypotheses; (1) when gut microbiota is transferred between two different individuals with differing MHC genotypes there will be an increase in virulence of a systemic pathogen, such as FVC, due to a energetics resource tradeoff occurring within immune system because the novel transplant contains many new antigens that may be responded to hindering the ability of a systemic immune response, or (2) microbiota will interact with the host immune system as it did in the original host leading to susceptibility shifts that correspond with the resistance of the donor host. Microbial transfers have large 6 implications in human health as the recent literature has shown them to be successful in treatment of some enteric pathogens (Bakken et al., 2011); however, it is still unknown what other physiological effects occur when there are changes in the gut microbiome. 1.1 Chapter Summaries 1.1.1 Chapter 2 The MHC is the most genetically diverse region in the vertebrate genome with over 2,500 Class I alleles known in the human population alone (Robinson et al., 2015). The MHC region is vitally important in immunity due to its capability to recognize foreign peptides (Germain, 1994). It has been suggested that increased allelic diversity of MHC in host populations can have a negative influence on the presence of pathogenic infection and that there only needs to be one host change between two different genotypes to see this negative effect (Brockhurst et al., 2014; Kubinak et al., 2015a). In this chapter, I test the hypothesis that when FVC is passaged through hosts with alternating MHC genotypes, there will be a reduction in fitness and virulence compared to viruses passaged through a single host genotype (pure passage). These pure passages consisted of FVC being serially passaged three times through a series of genetically identical mice. The alternated passages consisted of FVC being serially passaged through three rounds of mice, but with a different MHC genotype in the second round. This protocol allowed determination of the effect of a single MHC genotype change on virulence evolution of FVC. There is a significant reduction in both fitness and virulence when the virus is passed through an alternating genotype, leading to the conclusion that increased MHC diversity is an impediment to pathogen adaptation, as exposure to just 7 one new MHC genotype can significantly hinder viral adaptation. Due to the resistance of one of the MHC genotypes, I was also able to test the hypothesis that negative frequency-dependent selection is one possible mechanism by which MHC diversity is maintained in host populations. Initially, the resistant MHC genotype did not show the predicted effect of lower viral fitness and virulence when the virus was passed through alternating hosts; which may be due to the inherent resistance masking the reduction in fitness and virulent seen when MHC alternating passages are done because the resistance keeps the virus from replicating. To test this hypothesis, I modified the resistance of this MHC by using a strain of FVC that had been given the opportunity to adapt to the resistant genotype over ten rounds of passage. Using the adapted FVC virus allowed me to see a significant reduction in both fitness and virulence when the virus was passed through a dissimilar genotype, exactly what one would expect if negative frequency-dependent selection was occurring in the experimental population. 1.1.2 Chapter 3 Transmission is an important aspect of life history for any natural pathogen, as it is the process required for pathogen propagation (Antonovics et al., 2017; Mccallum et al., 2017). In order to determine the influence transmission, host sex and social status have on virulence, I used wild-derived animals in seminatural enclosures. The 30 m2 enclosures allowed the mice to interact naturally, leading to competition for mates and territories between the males and selective mate choice by females (Potts et al., 1991; Carroll & Potts, 2007). Since FVC is believed to transmit through bodily fluids, the natural behavior witnessed in the seminatural enclosures promotes transmission of the 8 virus through increasing interactions between mice. To test the hypotheses that males are the main source of transmission, that natural transmission hinders virulence and that social status can influence virulence, three different infection-treatment populations were created using both index (infected) and contact (initially uninfected) animals. The three treatments were: female only index, male only index, and both female and male index populations. This design allowed me to determine the relative transmission frequencies via each sex. Furthermore, by measuring the virulence and titer of FVC in each contact animal, I was able to determine how natural transmission affected virulence relative to the experimentally infected index animals. I found that populations that have both male and female index animals, or male only index animals have the highest rates of transmission, whereas the populations with only females have the lowest rates of transmission. These results suggest that males are the main source of transmission events in this system. Even though there was a significant increase in viral titers in male contact animals compared to female contact animals, the higher rates of transmission did not lead to increased spleen mass in males over females. Not only are males transmitting the virus more, but they are also harboring more virus per unit of spleen after contracting the virus, providing evidence that FVC may be male adapted. Moreover, as natural transmission occurs in these populations, there is a significant reduction in overall virulence of FVC in contact animals compared to index animals. Likewise, index animals had significantly higher titers than contact animals suggesting that natural transmission is an impediment to replication as well. 9 1.1.3 Chapter 4 Fecal microbiome transfers (FMT) are showing promise in the treatment of enteric pathogens (Bakken et al., 2011; Vrieze et al., 2013). Though there have been many studies elucidating the effects of microbiome on enteric function, there has been little information published about how transferring the microbiome influences systemic immunity of the host (Sekirov et al., 2010). Using established microbial transfer methods with antibiotic treatment, followed by recolonization with new microbial constituents, one could inquire whether fecal transplants from different host genotypes influence systemic virulence (Kubinak et al., 2015b). I propose two hypotheses; (1) that due to a trade-off in immune function created when a host must respond to the new antigens present when encountering a novel microbiome, a systemic viral pathogen will have higher virulence than it would if the host had a native microbiome and (2) the resistance to FVC of the donor microbiome will be recapitulated in the host receiving the microbial transplant due to the cross talk between the gut microbiome and the host immune system. To test these hypotheses, I created both treatment (novel) and control (native) groups. In the treatment groups, novel microbiomes were transferred between two different MHC congenic mouse strains (Balb/cdd and Balb/cbb), while in the control groups each animal received native microbiome from a host with the same MHC genotype. Following reconstitution each animal was challenged with a stock FVC. The test resulted in virus from the novel treatment group having significantly higher titer and virulence than the native group, partially confirming hypothesis one. This may be due to an immune tradeoff where the host is shuttling more immune resources to the gut and the systemic compartment is left with fewer resources capable of responding 10 to the pathogen (Lochmiller & Deerenberg, 2000; Bordes & Morand, 2011). However, only a reduction of overall lymphocyte number was seen in cell immune phenotyping in novel groups and there were no other significant effects. This suggests that a wider variety of immune cell subsets need to be tested in order to fully understand how microbiome transfers influence susceptibility to systemic pathogens. The finding that fecal transplants increase susceptibility to pathogens has large implications in the human health field, as using this line of treatment for enteric pathogens may make the patient more susceptible to systemic infection. 1.2 References Abbas, A.K., Lichtman, A.H. & Pillai, S. 2012. Properties and overview of immune responses. Cellular and Molecular Immunology, pp. 1-14. Alizon, S., Hurford, a, Mideo, N. & Van Baalen, M. 2009. Virulence evolution and the trade-off hypothesis: history, current state of affairs and the future. J. Evol. Biol. 22: 245-59. Antonovics, J., Wilson, A.J., Forbes, M.R., Hauffe, H.C., Kallio, E.R., Leggett, H.C., et al. 2017. The evolution of transmission mode. Philos. Trans. R. Soc. B 372. 1-12. Apanius, V., Penn, D., Slev, P.R., Ruff, L.R. & Potts, W.K. 1997. The nature of selection on the major histocompatibility complex. Crit. Rev. Immunol. 17: 179-224. Bakken, J.S., Borody, T., Brandt, L.J., Brill, J. V., Demarco, D.C., Franzos, M.A., et al. 2011. Treating clostridium difficile infection with fecal microbiota transplantation. Clin. Gastroenterol. Hepatol. 9: 1044-1049. Elsevier Inc. Bodmer, W.F., Trowsdale, J., Young, J. & Bodmer, J. 1986. Gene clusters and the evolution of the major histocompatibility system. Philos. Trans. R. Soc. Lond. B. Biol. Sci. 312: 303-315. Bordes, F. & Morand, S. 2011. The impact of multiple infections on wild animal hosts: a review. Infect. Ecol. Epidemiol. 1: 1-10. Brockhurst, M.A., Chapman, T., King, K.C., Mank, J.E., Paterson, S. & Hurst, G.D.D. 2014. Running with the Red Queen: the role of biotic conflicts in evolution. Proc. R. 11 Soc. B 281: 1-9. Buffie, C.G. & Pamer, E.G. 2013. Microbiota-mediated colonization resistance against intestinal pathogens. Nat Rev Immunol 13: 790-801. Nature Publishing Group. Carroll, L.S. & Potts, W.K. 2007. Sexual selection: Using social ecology to determine fitness differences. Rodent societies: An ecological & Evolutionary Perspective (J. O. Wolff & P. W. Sherman, eds), pp. 58-67. The University of Chicago Press, Chicago, IL. Cebra, J.J., Periwal, S.B., Lee, G., Lee, F.A.N. & Shroff, K.E. 1998. Development and maintenance of the gut-associated lymphoid tissue (GALT): The roles of enteric bacteria and viruses. Dev. Immunol. 6: 13-18. Chalfin, L., Dayan, M., Levy, D.R., Austad, S.N., Miller, R. a., Iraqi, F. a., et al. 2014. Mapping ecologically relevant social behaviours by gene knockout in wild mice. Nat. Commun. 5: 1-10. Nature Publishing Group. Ebert, D. 1998. Experimental evolution of parasites. Science (80-. ). 282: 1432-1436. Ebert, D. & Bull, J. 2008. The evolution and expression of virulence. In: Evolution in health and disease, pp. 153-167. Finlay, B.B. & McFadden, G. 2006. Anti-immunology: Evasion of the host immune system by bacterial and viral pathogens. Cell 124: 767-82. Friend, C. 1957. Cell-free transmission in adult swiss mice of a disease having the character of a leukemia. J. Exp. Med. 105. 307-325 Germain, R.N. 1994. MHC-dependent antigen processing and peptide presentation: Providing ligands for T lymphocyte activation. Cell 76: 287-299. Haldane, J.B.S. 1949. Disease and evolution. La Ric. Scentifica Suppl. 19: 1-11. Hasenkrug, K.J. & Chesebro, B. 1997. Immunity to retroviral infection : The Friend virus model. Proc. Natl. Acad. Sci. U. S. A. 94: 7811-7816. Hensley-McBain, T., Zevin, A.S., Manuzak, J., Smith, E., Gile, J., Miller, C., et al. 2016. The effects of fecal microbial transplantation on microbiome and immunity in SIVinfected macaques. J. Virol. 90: JVI.00099-16. Jeffery, K.J. & Bangham, C.R. 2000. Do infectious diseases drive MHC diversity? Microbes Infect. 2: 1335-41. Koch, H. & Schmid-Hempel, P. 2012. Gut microbiota instead of host genotype drive the specificity in the interaction of a natural host-parasite system. Ecol. Lett. 15: 1095- 12 1103. Kubinak, J.L., Cornwall, D.H., Hasenkrug, K.J., Adler, F.R. & Potts, W.K. 2015a. Serial infection of diverse host (Mus) genotypes rapidly impedes pathogen fitness and virulence. Proc. Biol. Sci. 282. Kubinak, J.L., Ruff, J.S., Cornwall, D.H., Middlebrook, E.A., Hasenkrug, K.J. & Potts, W.K. 2013. Experimental viral evolution reveals major histocompatibility complex polymorphisms as the primary host factors controlling pathogen adaptation and virulence. Genes Immun. 14: 365-72. Nature Publishing Group. Kubinak, J.L., Ruff, J.S., Whitney, C., Slev, P.R. & Potts, W.K. 2012. Experimental viral evolution to specific host MHC genotypes reveals fitness and virulence trade-offs in alternative MHC types. Proc. Natl. Acad. Sci. U. S. A. 109: 3422-3427. Kubinak, J.L., Stephens, W.Z., Soto, R., Petersen, C., Chiaro, T., Gogokhia, L., et al. 2015b. MHC variation sculpts individualized microbial communities that control susceptibility to enteric infection. Nat. Commun. 6: 8642. Lochmiller, R.L. & Deerenberg, C. 2000. Trade-offs in evolutionary immunology: Just what is the cost of immunity? Oikos 88: 87-98. Mazmanian, S.K., Liu, C.H., Tzianabos, A.O. & Kasper, D.L. 2005. An immunomodulatory molecule of symbiotic bacteria directs maturation of the host immune system. Cell 122: 107-118. Mccallum, H., Fenton, A., Hudson, P.J., Lee, B., Levick, B., Norman, R., et al. 2017. Breaking beta: Deconstructing the parasite transmission function. Philos. Trans. R. Soc. B 372. 1-11. Meagher, S., Penn, D.J. & Potts, W.K. 2000. Male-male competition magnifies inbreeding depression in wild house mice. Proc. Natl. Acad. Sci. U. S. A. 97: 3324- 3329. Mirand, E. a., Buffett, R.F. & Grace, J.T. 1966. Mode of transmission of Friend virus disease. Exp. Biol. Med. 121: 970-976. Nelson, A.C., Cauceglia, J.W., Merkley, S.D., Youngson, N. a, Oler, A.J., Nelson, R.J., et al. 2013. Reintroducing domesticated wild mice to sociality induces adaptive transgenerational effects on MUP expression. Proc. Natl. Acad. Sci. U. S. A. 110: 19848-53. Portis, J.L., McAtee, F.J. & Hayes, S.F. 1987. Horizontal transmission of murine retroviruses. J. Virol. 61: 1037-44. Potts, W., Manning, C. & Wakeland, E. 1991. Mating patterns in seminatural populations 13 of mice influenced by MHC genotype. Nature 352: 619-621. Radwan, J., Biedrzycka, A. & Babik, W. 2010. Does reduced MHC diversity decrease viability of vertebrate populations? Biol. Conserv. 143: 537-544. Elsevier Ltd. Robinson, J., Halliwell, J.A., Hayhurst, J.D., Flicek, P., Parham, P. & Marsh, S.G.E. 2015. The IPD and IMGT/HLA database: Allele variant databases. Nucleic Acids Res. 43: D423-D431. Ruff, J.S., Nelson, A.C., Kubinak, J.L. & Potts, W.K. 2012. MHC signaling during social communication. Adv. Exp. Med. Biol. 738: 290-313. Ruscetti, S.K. 1999. Deregulation of erythropoiesis by the Friend spleen focus-forming virus. Int. J. Biochem. Cell Biol. 31: 1089-109. Salathé, M., Kouyos, R.D. & Bonhoeffer, S. 2008. The state of affairs in the kingdom of the Red Queen. Trends Ecol. Evol. 23: 439-45. Sekirov, I., Russell, S. & Antunes, L. 2010. Gut microbiota in health and disease. Physiol. Rev. 90: 859-904. Snell, G.D. 1948. Methods for the study of histocompatibility genes. J. Genet. 49: 87- 108. Stecher, B., Macpherson, A.J., Hapfelmeier, S., Kremer, M., Stallmach, T. & Hardt, W. 2005. Comparison of Salmonella enterica Serovar typhimurium colitis in germfree mice and mice pretreated with Streptomycin. Infect. Immun. 73: 3228-3241. Sutton, J.T., Nakagawa, S., Robertson, B.C. & Jamieson, I.G. 2011. Disentangling the roles of natural selection and genetic drift in shaping variation at MHC immunity genes. Mol. Ecol. 20: 4408-20. Van Valen, L. 1973. A new evolutionary theory. Evol. Theory 1: 1-30. Vrieze, A., De Groot, P.F., Kootte, R.S., Knaapen, M., Van Nood, E. & Nieuwdorp, M. 2013. Fecal transplant: A safe and sustainable clinical therapy for restoring intestinal microbial balance in human disease? Best Pract. Res. Clin. Gastroenterol. 27: 127- 137. Elsevier Ltd. Watts, C. 1997. Capture and processing of exogenous antigens for presentation on MHC molecules. Annu. Rev. Immunol. 15: 821-50. Zaph, C., Du, Y., Saenz, S.A., Nair, M.G., Perrigoue, J.G., Taylor, B.C., et al. 2008. Commensal-dependent expression of IL-25 regulates the IL-23 - IL-17 axis in the intestine. J. Exp. Med. 205: 2191-2198. CHAPTER 2 EXPERIMENTAL MANIPULATION OF POPULATION-LEVEL MHC DIVERSITY CONTROLS PATHOGEN VIRULENCE EVOLUTION IN MUS MUSCULUS 2.1 Abstract The virulence levels attained by serial passage of pathogens through similar host genotypes in laboratories are much higher than observed in natural systems, however, it is unknown what keeps natural virulence levels below these empirically demonstrated maximum levels. One hypothesis suggests that host diversity impedes pathogen virulence, because adaptation to one host genotype carries tradeoffs in the ability to replicate and cause disease in other host genotypes. To test this hypothesis, with the simplest level of population diversity within the loci of the major histocompatibility complex (MHC), we serially passaged Friend Virus Complex (FVC) through two rounds in hosts with either the same MHC genotypes (pure passage) or hosts with different MHC genotypes (alternated passage). Alternated passages showed a significant overall reduction of viral titer (41%) and virulence (55%) when compared to pure passages. Furthermore, a resistant host genotype initially overshadowed the effects we expected to see during alternating MHC passages, however, when FVC was allowed to adapt to the resistant host genotype, typical effects emerged in alternating lines, i.e., alternated lines 15 show reduced virulence. These data indicate serial exposure to diverse MHC genotypes is an impediment to pathogen adaptation, suggesting genetic variation at MHC loci is important for limiting virulence in a rapidly evolving pathogen and supports negative frequency-dependent selection as a force maintaining MHC diversity in host populations. 2.2 Introduction Serial passage studies of pathogens have shown order-of-magnitude increases in virulence; however, many of the factors contributing to this virulence evolution are still unknown (Brown et al., 2001; Kawecki et al., 2012; Kubinak et al., 2013). Genotype specific adaptation by the virus may be a major cause of these order-of-magnitude level virulence increases, as each experimental passage was done through similar host genotypes. The hypothesis that between host genetic diversity can influence virulence was tested in a mouse-retrovirus system and when host genotypes are alternated during serial passage, trade-offs occur as both fitness and virulence evolution are severely impeded (Kubinak et al., 2015); this provides an explanation for why the virulence of pathogens in nature, seem to seldom reach the levels observed during serial passage studies in the laboratory. However, the previous study used mouse genotypes that differed across the entire genome, so that study was unable to identify any regions of particular importance for retarding virulence evolution. Classical major histocompatibility complex (MHC) genes are the most polymorphic loci known in vertebrates and play a critical role during immune recognition (Robinson et al., 2015). The high degree of allelic diversity present in host populations is thought to be due to antagonistic coevolution between hosts and their pathogens (Potts et 16 al., 1994; De Boer et al., 2004). Often referred to as Red Queen (RQ) dynamics, these interactions predict reciprocal patterns of selection whereby adaptations that benefit one species will favor counter-adaptations in the interacting species (Van Valen, 1973). Furthermore, RQ dynamics may lead to maintenance of MHC diversity in populations through negative frequency-dependent selection; in turn, controlling virulence in the host population because of the trade-offs associated with genotype specific adaptation, i.e., specialization (Carius et al., 2001; Meyer-Lucht & Sommer, 2005; Sutton et al., 2011). MHC class I and class II genes allow for discrimination between self and nonself peptide antigens by encoding cell surface molecules that bind and present intracellular and extracellular peptides on the cell surface for interrogation by circulating Tlymphocytes (T-cells) (Jeffery & Bangham, 2000; Wegner et al., 2003; Spurgin & Richardson, 2010). Recognition by a T-cell expressing a cognate T-cell receptor to a given MHC-peptide complex, with co-stimulating factors, triggers an adaptive immune response (Watts, 1997). Consequently, circumventing antigen presentation by MHC molecules may be a primary means by which pathogens increase within-host growth rates (fitness) (Fruh et al., 1999). For example, over 50% of genetic variants seen in Human Immunodeficiency Virus (HIV) occur in MHC presented epitopes (Pereyra et al., 2011; Henn et al., 2012); viral adaptation to escape specific MHCs could lead to trade-offs in pathogen fitness in hosts expressing different MHC molecules, due to the specialization of the viral pathogen on the first host leading to maladaptation in the second host. Multiple studies in mice have demonstrated that MHC polymorphisms influence the severity of disease associated with pathogen infection (i.e., virulence) (Meyer-Lucht & Sommer, 2005; Wedekind et al., 2006; Kubinak & Potts, 2013). Furthermore, trade-offs 17 in pathogen fitness associated with adaptation to specific MHC genotypes between multiple hosts may explain the association between virulence and MHC polymorphism (McClelland et al., 2003). Friend Virus Complex (FVC) is a mouse specific gamma-retrovirus that has been used extensively in immunological studies and more recently in experimental evolution studies (Kubinak et al., 2012b, 2013, 2015). The complex is made up of two viral genomes that work synergistically to cause an erythro-proliferative disorder in susceptible mouse strains: the replication-competent Friend Murine Leukemia Virus (FMuLV) and the replication-incompetent Spleen Focus Forming Virus (SFFV) (Friend, 1957; Ruscetti, 1999). FVC has been shown to adapt to specific host MHC genotypes through serial passage, which results in a reduction of fitness and virulence when hosts possessing different MHC genotypes are encountered (Kubinak et al., 2012b). A recent study also suggests that loci within the MHC region account for a majority of the reduction in pathogen fitness and virulence associated with exposure to a novel host genotype (Kubinak et al., 2013); however, the previously mentioned studies were performed by serially exposing FVC to a single MHC genotype through ten rounds of passage, which is unlikely to happen in wild host populations because of existing genetic variation at these loci. Here we test the simplest case of a population where we alternate virus between two different host MHC genotypes to determine whether MHC polymorphisms between two individuals is sufficient to reduce viral fitness and virulence. We hypothesize that passage of FVC through two hosts possessing unique MHC alleles will result in significantly lower viral fitness and virulence than virus serially exposed to the same 18 MHC genotype. The work presented here builds upon previous evidence showing benefits of genome-wide genetic diversity and suggests that between-host MHC diversity may be a major factor reducing pathogen virulence while simultaneously being maintained through selective pressures imposed by pathogens interacting with host genotypes (Kubinak et al., 2012b, 2013, 2015). 2.3 Methods 2.3.1 Mouse Model The Mus musculus animals used for the study were 2 to 6 month old females purchased from Jackson Laboratories and bred under specific-pathogen-free conditions. Three different MHC congenic strains of BALB/c mice were used: BALB/cByJ (Balb/cdd), C.C3-H2k (Balb/ckk), and C.B10-H2b (Balb/cbb). Each independent experiment consisted of three viral passage lines with ten replicates per line, followed by a test phase again using ten replicates per passage line. All of the strains used were susceptible to FV infection; however, they had different susceptibility profiles with BALB/cbb being most resistant, BALB/cdd being least resistant, and BALB/ckk being of intermediate resistance (Chesebro et al., 1990; Kubinak & Potts, 2013). All animals were housed and used in accordance with the University of Utah Institutional Animal Care and Use Committee (protocol # 04-08022). 2.3.2 Pathogen Model A previously described biological clone of an NB-tropic strain of Friend Virus Complex (FVC) was maintained in an NIH3T3 fibroblast cell line (biological clone) and 19 used to generate all viral stocks used in the experiments (Kubinak et al., 2012b, 2013, 2015; Kubinak & Potts, 2013). Unpassaged viral stocks were generated by harvesting viral cell supernatants from an infected NIH3T3 fibroblast cell line; this bioclone stock was passaged for two rounds through BALB/cdd animals in order to obtain a BALB/cdd adapted stock virus; furthermore, due to the high mutation rates and large population sizes viruses are known to have, this two round passage also increases viral diversity of the stock virus allowing evolution to act (Lauring & Andino, 2010; Domingo et al., 2012). The two-round adapted stock was used to start all passage lines for each individual experiment, excluding the Balb/cbb adapted two round experiment. The two-round adapted stock was used to maintain sustained animal-to-animal passage in more resistant Balb/cbb and Balb/ckk genotypes. Spleen supernatants from the 14 animals infected in the second round were pooled in order to obtain enough volume to last the duration of the experiment. The pooled spleen supernatants were aliquoted and stored at -80°C. All viral stocks were kept on ice and each aliquot was only thawed once. 2.3.3 Virulence Estimates FVC causes an erythro-proliferative disorder leading to gross enlargement of the spleen (splenomegaly) during acute infection. Thus, disease virulence (harm to host) was quantified by measuring spleen mass to the nearest hundredth of a gram. 2.3.4 Viral Fitness Estimates Viral titer (retroviral genomes integrated into the host genome) was used as a proxy for viral fitness. Due to the correlated relationship between F-MuLV and SFFV (p 20 = 0.014, r2 = 0.41) we only F-MuLV titer data was fully generated (Figure 2S.1); however, SFFV titer data was also generated on a subset of the animals, showing a similar effect to F-Mulv (Figure 2S.2). Viral titer was measured by using the Roche Lightcycler 2.0 platform with primers previously published for this system (Kubinak et al., 2012b). Detected F-MuLV viral genomes were normalized to glyceraldehyde-3phospate dehydrogenase (GAPD) copies in order to get a standardized ratio of viral genomes per host genome copies. The viral genome by GAPD genome ratios were then multiplied by the spleen mass in order to obtain estimates of viral genomes per spleen. 2.3.4 Serial Passage Design The initial passage rounds were created by using either a two-round Balb/cdd passaged stock virus or a Balb/cbb adapted stock virus (described below) to infect, through intraperitoneal (ip) injection, 10 individual animals of each mouse strain. At 12 days postinfection each animal was sacrificed, infected spleens were collected for viral load estimates, and measurements of disease were made. After organ mass was recorded a portion of the spleen was saved in RNAlater™ and the rest combined in a 1:1 ratio of 1x Phosphate Buffered Saline (PBS) (i.e., 1g of spleen was diluted in 1ml PBS). The spleen-PBS mixture was mechanically homogenized. Homogenates were then centrifuged at 5000 RPM for 5 min and virus-laden supernatants were collected and frozen at -80oC. In order to control for dose of the virus in the test phase each viral inoculum was standardized by titering the supernatant and diluting it to the average titer of the alternating treatment group with the lowest average titer of its replicates; any inoculum under the average was used undiluted at a full dose (200uL) while anything 21 above the average was diluted before injection. Furthermore, it has been previously shown that there is no graft effect of viral proteins, from a previous animal seen, when the spleen supernatant is injected into a novel host (Kubinak et al., 2015). 2.3.5 Stock Two-Round Alternating Experiments Experiments consisted of three independent sub-experiments carried out in each BALB/c MHC congenic mouse strain (Balb/cdd, Balb/ckk, and Balb/cbb). In each, virus was passaged twice through animals possessing the same MHC genotype or twice through animals possessing different MHC genotypes (Figure 2.1). 200 uL of the stock virus was used to inoculate 10 animals of each MHC congenic strain (Round 1). After the infection period, the animals were sacrificed and 20 uL of each individual spleen supernatant was used to infect an animal from each of the three MHC congenic host strains (Round 2). After the second infection period, the spleen supernatant was standardized and used to inoculate the original strain of passage (Test Phase). Three subexperiments were done in total with each congenic strain of Mus acting as the control pure passage with their corresponding alternated passage lines through the other two MHC genotypes. In total, there were 70 infections with 30 independent passage lines in each of the three subexperiments, for a total of 210 infections. 2.3.5 Balb/cbb Adapted Two-Round Alternating Experiment Due to the unexpected result shown in the Balb/cbb adapted experiment, where fitness and virulence was reduced in all treatments in the test phase, we conducted another sub-experiment with a FVC strain adapted to the Balb/cbb genotype. These 22 additional experiments were done in order to test if the expected pattern where passages through alternating MHC genotypes reduces titer and virulence, holds true when the virus is adapted to the Balb/cbb genotype to overcome some of the inherent host resistance. We used a previously generated virus that was adapted to the Balb/cbb host genotype through ten rounds of passage in order to obtain an adapted virus to the resistant genotype (Kubinak et al., 2012b). The previously described Balb/cbb experiment was repeated with the new Balb/cbb stock adapted through ten rounds of passage. Again, three different passage treatments were conducted, in which the virus was passaged through either pure passage lines (Balb/cbb in each round) or alternating passage lines (different MHC genotype in the second round of passage, Balb/ckk or Balb/cdd). Again, a test phase was conducted in the original host strain of passage with the same dilution scheme, in order to test for differences between passage treatments. 2.3.6 Statistics All statistical analyses were completed in R (3.3.1) (R Core Team, 2016).Titer data was multiplied by spleen mass in order to obtain total viral titer per spleen estimates and spleen mass and titer data was log10 transformed due to its non-normal nature. Data pooled between the initial stock two round experiments were analyzed using linear mixed model in the lme4 package with response variables of either viral titer or spleen mass and a fixed effect of treatment (Bates et al., 2014). A random effect of host genotype was included in order to control for variation due to different resistance patterns in the host strains between experiments, p-values were obtained using the R package lmerTest (2.033) (Table 2S.1) (Kuznetsova et al., 2016). Furthermore, each individual stock two round 23 experiment was analyzed by a one-way ANOVA with corresponding Tukey post-hoc tests to determine differences between each treatment group. The Balb/cbb adapted experiment was analyzed separately from the two round alternating experiments and was also analyzed by a one-way ANOVA with post-hoc Tukey tests to determine differences between passage lines. All data in figures represent mean ± SEM. All data will be uploaded to an online repository, such as DRYAD, upon acceptance. 2.4 Results 2.4.1 Stock Two-Round MHC Alternating Passages Overall, MHC experiments show a significant reduction in viral fitness and virulence when the virus went through an alternated passage regime versus a pure passage regime. FVC stocks derived from alternated passage demonstrated a 41% reduction in titers compared to pure passaged FVC (LMM; t = -3.72, p < 0.001) (Figure 2.2A). SFFV titer showed a similar overall pattern where alternating passages have lower titers than pure passage lines (LMM; t = -3.53, p = 0.001) (Figure 2S.2). Furthermore, virulence of FVC was reduced by 55% in alternating passages when compared to pure passage lines (LMM; t = -3.42, p < 0.001) (Figure 2.2B). Viral titer measures demonstrated a statistically significant overall reduction in fitness in alternated passages when the three two-round stock experiments are pooled, however, not all comparisons for viral titer in each individual experiment were significantly reduced (Figure 2.3A-C). The Balb/cdd experiment showed a significant difference in viral titer between the experimental groups (Figure 2.3A) (ANOVA; F(2, 31) = 4.9, p = 0.01). Specifically, there was a significant reduction of FVC titer when 24 passaged through both alternated Balb/cbb (Tukey, p = 0.02), and alternated Balb/ckk lines (Tukey, p = 0.05) relative to pure passage lines. Similarly, when FVC was passaged in the Balb/ckk experiment there were significant differences between viral titer of experimental groups (Figure 2.3B) (ANOVA; F(2, 21) = 11.4, p < 0.001). In this case, a significant reduction in viral titer was observed only when FVC was passaged through the Balb/cbb genotype when compared to pure passage lines (Tukey, Balb/cbb, p < 0.001; Balb/cdd, p = 0.12). The Balb/cbb experiment did not show the predicted effect and there was actually an increased viral titer when FVC was passaged through the alternating Balb/cdd genotype (Figure 2.3C) (Fitness, ANOVA, F(2, 24) = 10.4, p < 0.001; Tukey, Balb/cdd, p = 0.02, Balb/ckk, p = 0.27). Spleen mass showed similar results to the titer data; the overall data showed alternating passages had a statistically significant virulence reduction compared to pure passages, however, the trend was not consistent among sub-experiments. The Balb/cdd sub-experiment showed a statistically significant relationship between groups (Figure 2.3D) (ANOVA; F(2, 39) = 4.9, p = 0.013), though, there was only a statistically significant reduction in virulence when FVC was passaged through the alternated Balb/cbb lines (Tukey, p = 0.009) and only slightly reduced in the alternated Balb/ckk lines (Tukey, p = 0.225). The Balb/ckk sub-experiment also showed a statistically significant relationship between groups (ANOVA; F(2, 25) = 8.1, p = 0.002) with each of the alternated lines showing reduced virulence, but only the Balb/cbb lines showing statistically significant reductions (Figure 2.3E) (Balb/cbb Tukey, p = 0.001; Balb/cdd Tukey, p = 0.270). Again, the Balb/cbb stock experiment did not show a reduction in virulence in alternating passages and there was an increase when the virus went through an alternating Balb/cdd 25 passage (Figure 2.3F) (ANOVA, F(2,27) = 10.9, p < 0.001; Tukey; Balb/cdd, p < 0.001). 2.4.2 Two-Round Balb/cbb Adapted Experiment In the experiment using a Balb/cbb adapted virus, a reduction in both titer and virulence was observed when FVC was passaged through animals with alternated MHC genotypes, though again not statistically significant in each alternating passage (Fitness, ANOVA; F(2, 23) = 20.3, p < 0.0001; Virulence, ANOVA; F(2, 27) = 19.7, p < 0.0001; Figure 2.4). Individually, there was only a significant reduction in titer and virulence when passaged through the Balb/ckk haplotype (Titer, Tukey p < 0.0001; Virulence, Tukey, p < 0.0001) (Figure 2.4A-B). Passaging FVC through the Balb/cdd genotype did show a reduction in the means in both titer and virulence from pure passage, however, this reduction was not significant (Tukey; titer, p = 0.93; virulence, p = 0.97). 2.5 Discussion The data presented here support the hypothesis that increased MHC diversity in host populations impedes pathogen adaptation as evidenced by reduced fitness and virulence. Though we do not see significant reduction of viral titer and virulence in each individual alternating passage, in general, there is an overall 41% reduction of viral titer and 55% reduction in virulence when FVC is passaged through alternated hosts compared to pure-passage hosts when variation between experiments is controlled in a LMM. The reductions in virulence observed here provide a mechanism for how pathogens, forced to adapt to multiple MHC genotypes encountered in natural populations, are not able to attain the virulence that has been observed in serial passages through isogenic hosts. 26 The initial two round stock Balb/cbb experiment did not show the predicted pattern; one explaination may be that due to the high resistance of the Balb/cbb, the resistance overshadows the effects that are expecting in alternating passages (Kubinak & Potts, 2013). If true, then a virus more adapted Balb/cbb might allow the predicted pattern to emerge. In order to test this hypothesis, we repeated the Balb/cbb experiment with a Balb/cbb adapted virus serially passaged through ten rounds. When FVC was able to adapt to the Balb/cbb genotype, our hypothesized pattern emerges: there is a reduction of fitness and virulence in the alternated lines when compared to the pure passage lines, though only significant in the Balb/ckk alternated lines. Interestingly, the virus alternated through Balb/cdd had significantly higher virulence and titer in the test phase than pure passage suggesting the susceptibility of MHC genotypes is an important interacting variable. These experimental data are consistent with other correlative studies showing a reduction in pathogen load when many different MHC alleles are present in a population (MeyerLucht & Sommer, 2005; Sin et al., 2014). Furthermore, by allowing the virus to adapt to the most resistant MHC genotype, making these hosts more susceptible, we experimentally support the negative frequency-dependent predictions of the RQ hypothesis, demonstrating the dynamic character of such coevolutionary interactions (Borghans et al., 2004; De Boer et al., 2004). Though the results presented here do not completely rule out other hypotheses that may be maintaining MHC diversity in populations, it does show that negative frequency-dependent selection is one force maintaining diversity and that more work needs to be completed. MHC diversity is thought to be maintained in vertebrate populations through three nonmutually exclusive mechanisms: non-random mate preference, heterozygote 27 advantage and negative frequency-dependent selection (Kubinak et al., 2012a; Brockhurst et al., 2014). Here we provide evidence for negative-frequency dependent selection in two ways. First, as MHC genotypes are alternated during serial passage of the virus, they supply a different selective pressure than the pure passage causing the virus to experience a fitness cost when each new MHC genotype is adapted to by the virus. These tradeoffs are essential for establishing molecular arms races and give rise to the attendant negative frequency-dependent selection. Second, the initial experiment through the resistant Balb/cbb genotype did not show reduced titer or virulence during alternated passages (compared to pure passages). However, FVC is capable of adapting to the most resistant MHC genotype (Balb/cbb) when passage through ten rounds, thereby making the host more susceptible and allowing the expected patterns of reduced virulence during alternated passages to emerge in subsequent experiments (Kubinak et al., 2013). The adapted Balb/cbb experiment models what might be expected in populations where the rare MHC alleles are initially not the targets of pathogen adaptation, but over time they become more common and consequently less resistant because they are now more frequent targets of pathogen adaptation. However, this dynamic is not seen in the initial experiment because hosts genotypes in their high resistance (rare) phase can dominate and overwhelm typical MHC-specific virulence patterns. These data demonstrate that population-level MHC diversity is beneficial to hosts because it reduces virulence evolution. It follows that a selective advantage will arise in families where both parents are MHC dissimilar because virulence is reduced as it moves through the different MHC genotypes of a family; providing a familial level function for MHC-dissimilar mating preferences (Potts et al., 1991; Eizaguirre et al., 2009; Lenz et 28 al., 2009). In many serial passage studies from Daphnia to mice, virulence has been shown to increase by orders of magnitude (Ebert & Bull, 2003; Kawecki et al., 2012; Kubinak et al., 2012b, 2013); however, the underlying mechanism of this adaptation is unknown. We provide evidence that the increases in virulence of FVC are due at least in part, to specialization of the pathogen through specific MHC genotype adaptation; as MHC plays a major role in the interactions pathogens have with hosts and as the pathogen increases in virulence by adapting to one specific MHC genotype it incurs trades-offs in other MHC genotypes (Kubinak et al., 2012b). The 36% reduction in virulence observed when the pathogen is passaged through two hosts with dissimilar MHC genotypes supports the hypothesis that diversity at MHC loci is a controlling factor for virulence evolution. Increasing the amount of MHC diversity in host populations provides an advantage for each individual in that population as it reduces the probability a pathogen that infects them will have seen a similar MHC genotype during the previous infection. Serial passage is a useful tool in coevolution studies and many have shown an increase in fitness and virulence of various pathogens using these techniques (Ebert, 1998; De Bruin et al., 2008; Mackinnon et al., 2008; Wei et al., 2014). Many of these studies have focused on pathogen interactions with species changes or sexual vs. asexual passages (Coffey & Vignuzzi, 2011; Ciota et al., 2015), but not passages of pathogens through different genotypes of the same vertebrate host species, as was done here. Furthermore, only a few of these serial passage studies have incorporated MHC allelic diversity into their passage schemes; which is thought to cause pathogen fitness tradeoffs through antagonistic relationships when the number of MHC alleles are increased in 29 populations (Radwan et al., 2010). Only Kubinak et. al. provided experimental evidence that between host genetic diversity reduces pathogen fitness and virulence in a vertebrate Mus host, however this work was done with hosts that differed throughout the genome, not just at the MHC region (Kubinak et al., 2015). Confirming strong effects of MHC diversity alone during experimental evolution highlights the importance MHC-specific adaptations have on viral evolution and represents the first experimental demonstration of diversity at specific loci impeding pathogen virulence evolution. Though this study shows a strong relationship between MHC diversity and pathogen evolution, we do not take into account natural transmission or other pathogens. Removal of the cost of transmission could lead to increased virulence of the pathogen (Ebert & Bull, 2003; Alizon et al., 2009), however, evidence suggests this is not causing the effect in our system. As host genotypes are alternated we see the associated costs in the new hosts with the same transmission route suggesting that viral interactions with the host MHC genotype is driving the adaptations and tradeoffs independent of removing the cost of transmission (Figure 2.2). Unfortunately, the relative importance of transmission and host genetics in virulence evolution is not known and needs to be investigated further. Moreover, FVC is a lab strain of virus and may not be found in this form in natural systems; however, other viruses such as Dengue virus have replication incompetent particles that enhance virulence much like SFFV, so this system may enable better understanding of the role of these defectives (Ke et al., 2013). Furthermore, due to its similar life cycle, FVC is a good model to use to understand other ecologically relevant retroviruses, such as HIV, Simian immunodeficiency virus, Bovine leukemia virus, and Koala retrovirus (Pollari et al., 1992; Tarlinton et al., 2006; Myers & 30 Hasenkrug, 2009; Hartmann, 2012). Furthermore, we would like to point out that though we used congenic inbred animals we do not know if each strain is the same across the genome outside of the MHC region. However, if this is the case we would expect the effects to be small compared to the effects of the MHC region as there has been much less time for large changes to occur that influence pathogen defense in the rest of the genome due to the pathogen free nature of these animals. Future work should focus on the relationships between naturally transmitting pathogens and population wide MHC diversity in order to understand the relative importance each plays in their coevolutionary system. Here we demonstrate that RQ dynamics contribute to virulence evolution by showing trade-offs in FVC fitness and virulence when it encounters a host with a novel MHC genotype, though many different factors may contribute to these trade-offs as well. These data suggest that the high virulence seen in serial passage studies may not be seen in nature due to the continual selection placed on the virus with each transmission event into a host with a novel MHC genotype. Furthermore, as the virus adapts to the most resistant genotype, we observed MHC trade-offs emerge when new (alternating) genotypes are encountered, as would be expected through negative-frequency dependent selection. These data shed light on the importance host pathogen co-evolutionary relationships play in the maintenance of MHC diversity and how this diversity is important for virulence evolution; having practical implications for endangered and agricultural species through their respective breeding programs (Frankham, 2010). 31 2.6 References Alizon, S., Hurford, a, Mideo, N. & Van Baalen, M. 2009. Virulence evolution and the trade-off hypothesis: history, current state of affairs and the future. J. Evol. Biol. 22: 245-59. Bates, D., Maechler, M., Bolker, B. & Walker, S. 2014. Fitting linear mixed-effects models using lme4. J. Stat. Softw. arXiv: 1406.5823. Borghans, J.A.M., Beltman, J.B. & De Boer, R.J. 2004. MHC polymorphism under hostpathogen coevolution. Immunogenetics 55: 732-739. Brockhurst, M.A., Chapman, T., King, K.C., Mank, J.E., Paterson, S. & Hurst, G.D.D. 2014. Running with the Red Queen: the role of biotic conflicts in evolution. Proc. R. Soc. B 281: 20141382-. Brown, E.G., Liu, H., Kit, L.C., Baird, S. & Nesrallah, M. 2001. Pattern of mutation in the genome of influenza A virus on adaptation to increased virulence in the mouse lung: identification of functional themes. Proc. Natl. Acad. Sci. U. S. A. 98: 6883-8. Carius, H.J., Little, T.J. & Ebert, D. 2001. Genetic variation in a host-parasite association: potential for coevolution and frequency-dependent selection. Evolution 55: 1136-45. Chesebro, B., Miyazawa, M. & Britt, W. 1990. Host genetic control of spontaneous and induced immunity to Friend murine retrovirus infection. Annu. Rev. Immunol. 8: 477-499. Ciota, A.T., Payne, A.F. & Kramer, L.D. 2015. West Nile virus adaptation to ixodid tick cells is associated with phenotypic trade-offs in primary hosts. Virology 482: 128- 132. Elsevier. Coffey, L.L. & Vignuzzi, M. 2011. Host alternation of chikungunya virus increases fitness while restricting population diversity and adaptability to novel selective pressures. J. Virol. 85: 1025-35. De Boer, R.J., Borghans, J.A.M., Van Boven, M., Keşmir, C. & Weissing, F.J. 2004. Heterozygote advantage fails to explain the high degree of polymorphism of the MHC. Immunogenetics 55: 725-731. De Bruin, A., Ibelings, B.W., Kagami, M., Mooij, W.M. & Van Donk, E. 2008. Adaptation of the fungal parasite Zygorhizidium planktonicum during 200 generations of growth on homogeneous and heterogeneous populations of its host, the diatom Asterionella formosa. J. Eukaryot. Microbiol. 55: 69-74. Domingo, E., Sheldon, J. & Perales, C. 2012. Viral quasispecies evolution. Microbiol. 32 Mol. Biol. Rev. 76: 159-216. Ebert, D. 1998. Experimental evolution of parasites. Science (80-. ). 282: 1432-1436. Ebert, D. & Bull, J.J. 2003. Challenging the trade-off model for the evolution of virulence: is virulence management feasible? Trends Microbiol. 11: 15-20. Eizaguirre, C., Yeates, S.E., Lenz, T.L., Kalbe, M. & Milinski, M. 2009. MHC-based mate choice combines good genes and maintenance of MHC polymorphism. Mol. Ecol. 18: 3316-29. Frankham, R. 2010. Challenges and opportunities of genetic approaches to biological conservation. Biol. Conserv. 143: 1919-1927. Elsevier Ltd. Friend, C. 1957. Cell-free transmission in adult swiss mice of a disease having the character of a leukemia. J. Exp. Med. 105. 307-325 Fruh, K., Gruhler, A., Krishna, R.M. & Schoenhals, G.J. 1999. A comparison of viral immune escape strategies targeting the MHC class I assembly pathway. Immunol. Rev. 168: 157-166. Hartmann, K. 2012. Clinical aspects of feline retroviruses: A review. Viruses 4: 2684- 2710. Henn, M.R., Boutwell, C.L., Charlebois, P., Lennon, N.J., Power, K.A., Macalalad, A.R., et al. 2012. Whole genome deep sequencing of HIV-1 reveals the impact of early minor variants upon immune recognition during acute infection. PLoS Pathog. 8. 114. Jeffery, K.J. & Bangham, C.R. 2000. Do infectious diseases drive MHC diversity? Microbes Infect. 2: 1335-41. Kawecki, T.J., Lenski, R.E., Ebert, D., Hollis, B., Olivieri, I. & Whitlock, M.C. 2012. Experimental evolution. Trends Ecol. Evol. 27: 547-60. Elsevier Ltd. Ke, R., Aaskov, J., Holmes, E.C. & Lloyd-Smith, J.O. 2013. Phylodynamic analysis of the emergence and epidemiological impact of transmissible defective dengue viruses. PLoS Pathog. 9. 1-11. Kubinak, J., Nelson, A., Ruff, J. & Potts, W. 2012a. Trade-offs limiting MHC heterozygosity. In: Ecoimmunology (G. E. Demas & R. J. Nelson, eds), pp. 225- 258. Oxford University press. Kubinak, J.L., Cornwall, D.H., Hasenkrug, K.J., Adler, F.R. & Potts, W.K. 2015. Serial infection of diverse host (Mus) genotypes rapidly impedes pathogen fitness and virulence. Proc. Biol. Sci. 282. 1-9. 33 Kubinak, J.L. & Potts, W.K. 2013. Host resistance influences patterns of experimental viral adaptation and virulence evolution. Virulence 4: 410-8. Kubinak, J.L., Ruff, J.S., Cornwall, D.H., Middlebrook, E.A., Hasenkrug, K.J. & Potts, W.K. 2013. Experimental viral evolution reveals major histocompatibility complex polymorphisms as the primary host factors controlling pathogen adaptation and virulence. Genes Immun. 14: 365-72. Nature Publishing Group. Kubinak, J.L., Ruff, J.S., Whitney, C., Slev, P.R. & Potts, W.K. 2012b. Experimental viral evolution to specific host MHC genotypes reveals fitness and virulence tradeoffs in alternative MHC types. Proc. Natl. Acad. Sci. U. S. A. 109: 3422-3427. Kuznetsova, A., Brockhoff, P.B. & Christensen, R.H.B. 2016. lmerTest: Tests in linear mixed effects models. R package version 2.0-25. Lauring, A.S. & Andino, R. 2010. Quasispecies theory and the behavior of RNA viruses. PLoS Pathog. 6: e1001005. Lenz, T.L., Eizaguirre, C., Scharsack, J.P., Kalbe, M. & Milinski, M. 2009. Disentangling the role of MHC-dependent "good genes" and "compatible genes" in mate-choice decisions of three-spined sticklebacks Gasterosteus aculeatus under seminatural conditions. J. Fish Biol. 75: 2122-42. Mackinnon, M.J., Gandon, S. & Read, a. F. 2008. Virulence evolution in response to vaccination: The case of malaria. Vaccine 26: C42-C52. McClelland, E.E., Penn, D.J. & Potts, W.K. 2003. Major Histocompatibility Complex heterozygote superiority during coinfection. Infect. Immun. 71: 2079-2086. Meyer-Lucht, Y. & Sommer, S. 2005. MHC diversity and the association to nematode parasitism in the yellow-necked mouse (Apodemus flavicollis). Mol. Ecol. 14: 2233-43. Myers, L. & Hasenkrug, K.J. 2009. Retroviral immunology: lessons from a mouse model. Immunol. Res. 43: 160-6. Pereyra, F., Jia, X., Mclaren, P.J., Kadie, C.M., Carlson, J.M., Heckerman, D., et al. 2011. The major genetic determinants of HIV-1 control affect HLA Class I peptide presentation. Science (80-. ). 330: 1551-1557. Pollari, F.L., Wangsuphachart, V.L., DiGiacomo, R.F. & Evermann, J.F. 1992. Effects of bovine leukemia virus infection on production and reproduction in dairy cattle. Can. J. Vet. Res. 56: 289-95. Potts, W., Manning, C. & Wakeland, E. 1991. Mating patterns in seminatural populations of mice influenced by MHC genotype. Nature 352: 619-621. 34 Potts, W., Manning, C. & Wakeland, E. 1994. The role of infectious disease, inbreeding and mating preferences in maintaining MHC genetic diversity: an experimental test. Philos. Trans. R. Soc. Lond. B. Biol. Sci. 346: 369-378. R Core Team. 2016. R: A language and environment for statistical computing. R Foundation for Statistical Computing, Vienna, Austria. Radwan, J., Biedrzycka, A. & Babik, W. 2010. Does reduced MHC diversity decrease viability of vertebrate populations? Biol. Conserv. 143: 537-544. Elsevier Ltd. Robinson, J., Halliwell, J.A., Hayhurst, J.D., Flicek, P., Parham, P. & Marsh, S.G.E. 2015. The IPD and IMGT/HLA database: Allele variant databases. Nucleic Acids Res. 43: D423-D431. Ruscetti, S.K. 1999. Deregulation of erythropoiesis by the Friend spleen focus-forming virus. Int. J. Biochem. Cell Biol. 31: 1089-109. Sin, Y.W., Annavi, G., Dugdale, H.L., Newman, C., Burke, T. & MacDonald, D.W. 2014. Pathogen burden, co-infection and major histocompatibility complex variability in the European badger (Meles meles). Mol. Ecol. 23: 5072-5088. Blackwell Publishing Ltd. Spurgin, L.G. & Richardson, D.S. 2010. How pathogens drive genetic diversity: MHC, mechanisms and misunderstandings. Proc. Biol. Sci. 277: 979-988. Sutton, J.T., Nakagawa, S., Robertson, B.C. & Jamieson, I.G. 2011. Disentangling the roles of natural selection and genetic drift in shaping variation at MHC immunity genes. Mol. Ecol. 20: 4408-20. Tarlinton, R.E., Meers, J. & Young, P.R. 2006. Retroviral invasion of the koala genome. Nature 442: 79-81. Van Valen, L. 1973. A new evolutionary law. Evol. Theory 1: 1-30. Watts, C. 1997. Capture and processing of exogenous antigens for presentation on MHC molecules. Annu. Rev. Immunol. 15: 821-50. Wedekind, C., Walker, M. & Little, T.J. 2006. The separate and combined effects of MHC genotype, parasite clone, and host gender on the course of malaria in mice. BMC Genet. 7: 55. Wegner, K.M., Reusch, T.B.H. & Kalbe, M. 2003. Multiple parasites are driving major histocompatibility complex polymorphism in the wild. J. Evol. Biol. 16: 224-32. Wei, K., Sun, H., Sun, Z., Sun, Y., Kong, W., Pu, J., et al. 2014. Influenza a virus 35 acquires enhanced pathogenicity and transmissibility after serial passages in Swine. J. Virol. 88: 11981-94. 36 Figure 2.1. Serial passage of FVC through MHC congenic mice. Virus was administered intraperitoneally. After 12 days of incubation it was split three ways and was passaged into a host of the similar MHC genotype (round 2 pure) as well as two different MHC genotypes (round 2 alternating). During the test phase, each round two virus was passaged into the original host of passage (round 1 genotype) in order to control for host differences in the test phase. Furthermore, each round was completed with ten replicates in each MHC genotype treatment group, denoted by x10. Each panel represents the independent experiments done in order to test each genotype in each passage scenario. 37 Figure 2.2. Virulence and titer of FVC in the test phase among all experimental treatments. (a) The titer of FVC was significantly reduced when passaged through alternating MHC host genotypes. (b) There was also a significant reduction of virulence (Spleen mass) when FVC was passaged through an alternating host. Symbols represent the experiment the data point came from: Balb/ckk (triangle), Balb/cdd (circle) and Balb/cbb (square). Data represented by mean ± SE. ** Indicates a P value < 0.01, *** < 0.001 38 Figure 2.3. Alternating MHC genotypes reduces both titer and virulence of FV complex. (a-c) There was a significant reduction of viral titer in three out of six alternating passages across each sub-experiment; furthermore, titer was reduced below pure passages in each alternating passage, though not significantly, except the Balb/cbb sub-experiment. (d-f) Virulence of FVC showed a significant reduction in two out of six possible alternating passages, again, virulence was reduced from pure passages in all cases though this reduction was not always significant. The Balb/cbb experiment did not show the expected result with regard to either titer or virulence. Data represented by mean ± SE. * Indicates a P value < 0.05., ** < 0.01, *** < 0.001. 39 Figure 2.4. Viral titer and virulence using the Balb/cbb adapted virus. (a) Alternating the MHC genotype shows a significant reduction in titer between groups, however, much of this significant effect is driven by the virus that is alternated through Balb/ckk genotypes with a minor reduction in titer when passaged through Balb/cdd . (b) Likewise, there was significant reductions between groups in virulence of FVC. Again, the reduction in virulence was only significant when alternated through Balb/ckk MHC haplotypes, with only a slight reduction in Balb/cdd passages. Data represented by mean ± SE. *** Indicates a P value < 0.001. 40 2.7 Supplementary Material Figure 2S.1. Correlation between SFFV and F-MuLV titers. Only qPCR data from FMuLV genomes was collected and reported due to the significant correlation between the two viral genomes that make up FVC: Friend Murine Leukemia Virus (F-MuLV) and Spleen Focus Forming Virus (SFFV). 41 Figure 2S.2. Virulence and titer of SFFV in the test phase among all experimental treatments. (a) The titer of FVC was significantly reduced when passaged through alternating MHC host genotypes. Data represented by mean ± SE. ** Indicates a P value < 0.01. 42 Table 2S.1. Linear mixed models comparing treatment groups (pure and alternated), controlling for host genotype. Virulence (Spleen mass) of test phase animals by treatment LMM (groups: Host = 3; observations 100) Random effects Variance Std. Deviation Host 0.1059 0.3254 Fixed effects Estimate Std. Error t value Pr(>|t|) Intercept (Pure Passage) -0.1607 0.2078 -0.773 0.50458 Treatment (Alternated Passage) -0.3708 0.1086 -3.415 0.00094 *** F-MuLV titers (log transformed) of test phase animals by treatment LMM (groups: Host = 3; observations 85) Random effects Variance Std. Deviation Host 1.4087 1.1869 Fixed effects Estimate Std. Error t value Pr(>|t|) Intercept (Pure Passage) 4.1895 0.7278 5.756 0.01952 * Treatment (Alternated Passage) -1.1970 0.3214 -3.724 0.00036 *** * Indicates a P value < 0.05., *** < 0.001. CHAPTER 3 NATURAL HORIZONTAL TRANSMISSION OF A MURINE RETROVIRUS WITHIN SEMINATURAL POPULATIONS OF MICE 3.1 Abstract Transmission is a key component of a pathogen's life history and more robust systems need to be developed to obtain reliable ecological estimates of the dominant mode(s) of transmission and determine how host traits such as sex and social status influence transmission dynamics. We have developed such a system, using seminatural populations of house mice and use it to evaluate the transmission dynamics of Friend Virus Complex (FVC). Seven seminatural populations were founded with 8 male and 16 female mice that were either "index" (initially infected) or "contact" (initially uninfected). Three experimental population designs were implemented: female index only, male index only, and female and male index. We found that male-male transmission is the predominant mode of transmission with little female transmission and no evidence of vertical transmission. Additionally, infected contact males harbor higher titers than females while suffering a similar amount of virulence; this paired with a strong correlation between titer and virulence in males and no correlation in females suggests FVC may be male adapted. We also found that contact animals had lower titers than 44 index animals at the end of the experiment suggesting FVC replication is reduced. Finally, social status influences infection, as though dominant and nondominant males contract the virus at similar rates and experience similar virulence, nondominant animals have higher titers. These results shed light on the complex life cycle of murine leukemia viruses and their evolutionary history and may also help develop a better understanding of the role of host sex, and social status on retroviral transmission dynamics. 3.2 Introduction Pathogen life history is complex and consists of many barriers that when passed coalesce in infection, and one of these consequential life history challenges is transmission (Mccallum et al., 2017). Therefore, understanding the mode(s) of transmission and how this mode influences pathogen virulence is a large and ongoing area of research (Antonovics et al., 2017). Historically it has been difficult to obtain reliable estimates of transmission, due to the biased nature of collection methods (i.e., only symptomatic individuals measured), and of the many factors controlling successful transmission of pathogens (Antonovics, 2017; Walker et al., 2017). Specifics about pathogen transmission dynamics such as, the transmission efficiencies of various modes and the influence of host sex, social status, and age are difficult to obtain in natural settings due to all the competing factors involved in transmission; addressing these shortcomings requires design of better experimental systems to obtain reliable estimates of the movements of different pathogens under a host of ecological conditions (Arthur et al., 2017). Transmission mode, how a pathogen moves between hosts, influences both the 45 likelihood of successful infection and resulting virulence experienced by the host. Pathogens are often opportunistic and use many different modes to infect subsequent hosts; however, some of these modes may allow for more likely transmission and higher pathogen burden than others (Antonovics et al., 2017). Experimenter transmission, wherein researchers directly infect a host with a pathogen, does not always mimic natural transmission modes (e.g., intraperitoneal injection with Friend Virus Complex (Portis et al., 1987; Ebert, 2013)). Also, it is thought that experimenter transmission removes the cost of transmission, potentially artificially increasing virulence (Ewald, 1983). Therefore, understanding the modes of transmission that a specific pathogen employs, their relative importance, as well as shifting away from experimenter to natural transmission is necessary to characterize the transmission ecology of host-pathogen systems. Through differences in physiology, hormone production, and immunity, host sex may be a powerful force driving differences in the transmission of pathogens (Zuk & McKean, 1996; Duneau & Ebert, 2012; Robinson & Klein, 2012). Males, for example, produce testosterone, which has been suggested to suppress the immune system, thereby increasing susceptibility to parasites and pathogens (Zuk, 1996). Many studies have examined the virulence in females in regards to vertically transmitted parasites (e.g., FVC in mice or Hepatitis C virus in humans) (Mirand et al., 1966; Thaler et al., 1991); however, few have explored differences in host virulence due to horizontal transmission or if transmission rates differ between the sexes. Furthermore, when sex biased horizontal transmission has been tested it has been measured indirectly using measures that correlate with transmission such as pathogen shedding or within host growth rates not successful 46 inoculation of another host (Perkins et al., 2008; Luong et al., 2009). The differences between the sexes in immunity, behavior, and other physiological processes may lead to key differences in transmission dynamics for a given parasite; unfortunately, the influence of host sex and potential sex-specific processes have not had rigorous characterization for most parasites (Klein, 2000). Another important characteristic that can shape transmission dynamics is an individual host's social status. Individuals of a species vary in physical condition, social connectivity, dispersal ecology, or possess heterogeneity in other traits that may bear importantly on their likelihood of contracting or transmitting a parasite (Arthur et al., 2017; Stephenson et al., 2017); these heterogeneities and their influence on transmission dynamics may be associated with belonging to a specific social class, such as dominant or nondominant (Hausfater & Watson, 1976; Hawley et al., 2007). Although specific relationships between social status and parasite transmission/virulence have been documented, general patterns across host and parasite species remain obscure (Arthur et al., 2017). Therefore, in order to understand the transmission ecology of a given pathogen, studies specifically assessing the influence of host social status are required. Friend Virus Complex (FVC) is a murine specific retrovirus that has been widely used in laboratories to understand various aspects of the retroviral life cycle and immunity (reviewed in Hasenkrug & Chesebro, 1997; Myers & Hasenkrug, 2009). FVC has been shown to be transmitted both vertically (through milk) and horizontally (sexually from males to females), at least, within susceptible strains of laboratory mice in artificial caged environments (Mirand et al., 1966; Portis et al., 1987). Additionally, a related wild murine ecotropic virus has been shown to transmit between males in cage 47 trials, though, neither of these studies have focused on disentangling the various modes of natural transmission, even in caged environments, to determine their relative importance (Portis et al., 1987). Moreover, simply demonstrating that transmission can occur via a specific mode does not provide information about which mode of transmission is the most common in nature, or the route to which a pathogen is primarily adapted. The commensal nature of house mice (Mus musculus) provides a unique opportunity to manipulate aspects of transmission ecology (e.g., density, sex-ratio of infected individuals) to determine how their pathogens move through a population in a controlled experimental setting. Experimental infections in house mouse populations inhabiting seminatural enclosures, when paired with appropriate titering, allow for quantifying contraction and transmission rates of pathogens and resulting virulence measures in an ecologically-relevant setting. The use of such enclosures has been advocated by the evolutionary and functional genomics community (Carroll & Potts, 2006, 2007); however, until now their novel attributes have yet to be harnessed for disease ecology and epidemiology. Here we use wild-derived house mouse populations living under seminatural conditions to evaluate the virulence and transmission dynamics of a murine retrovirus (FVC). Specifically, we assess: whether horizontal or vertical is the dominant mode of transmission; which sex, if either, drives horizontal transmission; if there is a corresponding sex-specific virulence and pathogen titer pattern accompanying differences in transmission; if virulence and titer measures resulting from experimenter and natural transmission differ; and if male social status influences the observed virulence, titer, or transmission patterns. Regarding the dominance of horizontal vs. vertical transmission, 48 specific a priori predictions are difficult to formulate due to conflicting cage studies (Mirand et al., 1966; Portis et al., 1987). However, if sexual transmission is the main route of horizontal transmission then female mice should contract the virus at a higher rate than males (while males transmit at higher rates than females (e.g., Padian et al., 1991; Looker et al., 2008), but if aggressive interactions are the main transmission mechanism, then males should have higher rates of both contraction and transmission as they more readily engage in fighting (Singleton & Krebs, 2007). Correspondingly, we predict that if one sex is primarily responsible for transmission then it will experience lower virulence due to infection than the other, according to theory (Anderson & May, 1982). Also, if natural transmission involves more barriers to transmission, such as less opportunity, dose or different routes, than experimenter transmission, we predict that animals naturally contracting the virus will have lower titers and virulence than animals infected intraperitoneally (Ewald, 1983). Finally, if dominant males are in better condition than nondominant males they will have lower titers and virulence; however, if they expend resources defending territories and this results in a tradeoff with immune function they will develop more severe infection (Arthur et al., 2017). 3.3 Methods and Materials 3.3.1 General Methods 3.3.1.1 Animals To assess transmission dynamics, mice with functional, wild type, behaviors are needed and it has been documented that standard laboratory strains have lost many of these behaviors (Nelson et al., 2013a; Chalfin et al., 2014). Therefore, we employed 49 wild-derived genetically outbred house mice (Mus musculus) from the 14th and 16th generation of a colony initially described in reference (Meagher et al., 2000). For cagebased experiments and before release into seminatural enclosures, individuals were provided access to food and water ad libitum and maintained on a 12:12 h light:dark cycle. All procedures were approved by the IACUC of the University of Utah (protocol #14-08022). 3.3.1.2 Pathogen FVC is an NB-Tropic gamma retrovirus that primarily infects erythroblasts causing gross splenomegaly in its murine host. FVC is composed of two separate genomes that work synergistically to cause disease. First, Friend murine Leukemia Virus (F-MuLV) is a prototypic gamma retrovirus which has all the proper machinery to infect and exit erythroblasts (for review see (Hasenkrug & Chesebro, 1997)). The second genome, Spleen Focus Forming Virus (SFFV), is capable of inserting into the genome, but cannot exit a cell without being pseudo-typed (packaged) in an F-MuLV capsid. However, SFFV's truncated envelope gene can stimulate the Erythropoietin receptor (EpoR) causing uncontrolled proliferation of erythroblasts allowing for more target cells for F-MuLV and increasing the probability of the SFFV genome being pseudo-typed and exiting the host cell (Ruscetti, 1999; Kosmider & Moreau-Gachelin, 2006). Due to the erythropoietic nature of FVC the main sight of replication is the spleen (Metcalf et al., 1959). The dramatic increase in the size of the spleen allows for an estimate of the amount of harm (i.e., virulence) the host is experiencing during infection. After sacrifice, the spleen is harvested and the mass measured. 50 F-MuLV viral titers were measured by qPCR amplification of retroviral integrations from the host genome (provirus) using primer sets and protocols described previously (Kubinak et al., 2012). Provirus quantification, as opposed to Ribonucleic Acid (RNA) measurements, was used because it is a measure of productive viral genomes that have inserted into the genome. Quantitative PCR was used to obtain estimates of viral fitness from spleen homogenates using the Lightcycler 2.0 platform. Crossing points above cycle 36 were requantified and if still over that threshold the sample was treated as uninfected per the manufacturer's recommendation due to the unreliability of quantifiable estimates at that number of cycles. Each proviral estimate was normalized to a Glyceraldehyde Phosphate Dehydrogenase (GAPD) reference to obtain relative per cell genome estimates. Furthermore, each proviral GAPD ratio was multiplied by spleen mass in order to get proviral load per spleen estimates; this estimate was used in all analysis except the correlations of F-MuLV proviral load by spleen mass, due to its potential to inflate the significance of the correlation. Only SFFV infection status (positive/negative) was measured due to the high level of endogenous DNA and expressed RNA similar to SFFV found in wild-derived mice. SFFV was amplified from RNA with the following primer sets, Fwd - 5'GTTCGCGCGCTTCTGCTTCC3', Rev - 5'ACAGCGAGACCACGAGTCGG3', using the Verso RT-PCR kit by Thermo Scientific on a Biorad c1000 touch machine used per manufacturer recommendations. PCR products were run on a 1% agarose gel in order to determine the presence or absence of the ~170 bp fragment of SFFV (Figure 3S.1). RNA extractions were completed using the Zymo Quick-RNA MiniPrep kit in order to obtain pure RNA template for PCR. 51 A Bioclone FVC stock was cultured from a transfected a NIH3T3 cell line, culture supernatants were collected, aliquoted and frozen at -70°C until needed. The resulting Bioclone stock was not virulent enough to cause disease in wild-derived mice; therefore, a Mus-adapted stock (as opposed to cell-culture adapted) capable of infecting animals in the study was generated. To create this stock, Bioclone was serially passaged through two rounds of BALB/c female mice. After two rounds of passage, each of the 14 animal's spleens were homogenized and pooled. Supernatants were collected from the pooled spleens and stored at -70°C until needed. The resulting virulent FVC stock did produce larger spleens in wild-derived mice than Bioclone (virulent stock: 0.27 ± 0.07 g, Bioclone: 0.05 ± 0.008 g; WRS; n = 48, W = 168.5, p = 0.019). Furthermore, within cages female mice experienced higher virulence from FVC infections with our virulent FVC stock than did males, but only a marginal increase of F-MuLV titers was observed in females over males. Twelve days postinfection females had spleens averaging 0.45 ± 0.13 g (M ± SEM), which were 3.46 times the mass of male spleens which were 0.11 ± 0.04 g (WSR; n = 28, W = 158.5, p = 0.005; Figure 3S.2A). However, only a marginally statistically significant difference in F-MuLV titers per spleen (F:M; 2.07 ± 0.33: 1.17 ± 0.32) was observed between the sexes (t-test; t = 1.97, p = 0.060 [Figure 3S.2B]). Interestingly, a correlation between spleen mass and titer was not identifiable in this data set (LM; t = -0.55, p = 0.209; Table 3S.1). 3.3.1.3 Seminatural Enclosures Enclosures are 30 m2 and subdivided into six subsections by wire mesh to promote territorial formation (Figure 3.1). Subsections contain ad libitum food and water 52 dispensers associated with a set of nest boxes in one of the four "optimal" territories, enclosed dark nesting sites, or two "suboptimal" territories, with nest boxes exposed to the light. Descriptions of enclosure, including additional photographs, and establishment/maintenance methods have been provided elsewhere in more detail (Ruff et al., 2013, 2015a; b, Gaukler et al., 2014, 2016). Populations are founded by 24 mice (16 females; 10 males); densities and sex ratio were selected to be similar to those reported in the wild (Singleton & Krebs, 2007). Populations in this study were maintained for 7 weeks, at which point founders and any offspring born within enclosures were sacrificed and organs were harvested to assess virulence measures and viral titers. Prior to release in seminatural enclosures both contact and index female founders were genotyped and categorized by a common mitochondrial allelic variant. Offspring collected at the termination of the populations were then typed so that the treatment of their mother could be determined. Mitochondrial types from all 108 offspring were gathered from populations containing female index animals and offspring were classified as having come from either contact (83) or index females (25). Information on primer design and PCR and gel electrophorese protocols have been described previously (Meagher et al., 2000). Prior to release into enclosures, founders of both sexes were implanted with unique passive integrated transponder (PIT) tags (TX1400ST, BioMark, Boise ID). Location data were gathered via a set of PIT antennae and readers (FS2001F-ISO, BioMark, Boise ID) placed at each of the feeders and data are recorded with data-logging software (Minimon, Culver City, CA). Dominance was assigned when a male had more than 80% of all male readings at a single feeder over the majority of the study as in 53 previous studies (Morris et al., n.d.; Nelson et al., 2013b). 3.3.2 Study Design and Statistical Analyses 3.3.2.1 Initial Evaluation of FVC Natural Transmission To determine if natural transmission of FVC is possible within seminatural enclosure populations a single pilot population of 24 mice (16 female and 8 male) was established with both female and male index mice (initially infected via IP injection) as well as contact mice (initially uninfected) of both sexes, resulting in half of the animals being contact and half being index. These mice were from the 14th-generation of the aforementioned colony and were 76.9 ± 12.1 (M ± SD) weeks of age. 3.3.2.2 Relative Importance of Horizontal Versus Vertical Transmission Infection status was assessed in the offspring of index females as described above and the rate of infected offspring was compared to the observed rates of horizontal transmission in adults. As no offspring tested positive for F-MuLV titers a one-sample ttest was used to compare rates of horizontal transmission across populations to the hypothetical value of 0. 3.3.2.3 Influence of Host Sex on Horizontal Transmission Dynamics Once natural transmission of FVC was confirmed six additional populations were established to determine the influence of host sex on FVC transmission dynamics. Two 54 populations were designed as described above with both female and male index founders, two populations had only female index founders, and the last two had only male index founders (Figure 3.1B). All populations contained 30 mice (20 female and 10 male) that were 13.6 ± 2.7 weeks old and study duration was again 7 weeks. To assess the effect of contact (female and male) and index host sex (female, male, and both sexes) on contact FVC infection status, we used a generalized linear mixed model (GLMM) to predict the probability of infection status; this and all other GLMMs were fit in R using the glmer function of the lme4 library (Bates et al., 2015; R Core Team, 2016). As infection status was defined as either infected or uninfected, we used a model assuming a binomial distribution with a logit link. Infection status was assessed in contact animals that survived the duration of the trial (n = 112) and these mice were grouped into seven populations, (the initial pilot population was included) from which individual observations for female and male contact mice were made (observations = 14). Contact and index sex were treated as fixed effects and population was modeled as a random effect. The model intercept was set for female contact animals from populations with only female index animals. 3.3.2.4 Influence of Host Sex and Transmission on Virulence and Pathogen Fitness Comparisons of spleen mass and viral titers were obtained between index and contact animals in order to determine the effect natural transmission has on FVC's ability to replicate and cause disease. All surviving animals were assessed for infection status by spleen mass and viral titer then each measure was pooled in order to obtain overall 55 virulence and titer estimates; in total, 75 observations were assessed out of the seven populations. Comparisons between contact and index virulence and titer were performed using linear mixed models (LMM) from R packages lmer4 and degrees of freedom and resulting p-values were estimated using lmerTest (Kuznetsova et al., 2016). Furthermore, in order to determine the effect host sex has on titer and virulence models were performed on each viral measure of contact animals; host sex and infection status were used as fixed effects and population as a random effect, as was done previously. To ensure that population design (sex of index mice) did not influence virulence and viral titer, and impair the ability to compare virulence and F-MuLV titers of contact and index mice, we performed a LMM. No significant effect of population design was observed (Figure 3S.3, Table 3S.2). 3.3.2.5 Influence of Male Social Status on Transmission, Virulence and Titer of Naturally Transmitted FVC To determine the effect social dominance has on transmission, virulence, and titer, comparisons were made using both GLMMs and LMMs. The influence of social dominance on transmission was determined by comparisons made between the transmission efficiency between surviving dominant and nondominant contact animals (n = 40). Due to the binomial nature of transmission data GLMMs were utilized again assuming a binomial distribution and a logit link. Social status was used as the fixed effect with population again used as a random effect. In order to determine the influence of social status on virulence and titer, LMMs were utilized due to the continuous nature of the data analyzed. Each LMM was conducted on 22 observations throughout the seven 56 populations and each model consisted of social status as the single fixed effect followed by population as the lone random effect. 3.4 Results Overall rates of horizontal transmission to contact founders in seminatural populations, regardless of the sex of index and contact mice, were 47.6 ± 7.6% (M ± SEM) (n = 7). Contrastingly, no offspring (0/25) born to index females had detectable FMuLV titers showing that vertical transmission did not occur. The observed difference in rates between horizontal and vertical transmission were statistically significant (t-test; t = 6.29, p = 0.001). Male mice within seminatural enclosures were the major drivers of viral transmission dynamics as they were both more likely to contract and transmit F-MuLV than females. Across all populations contact males had contraction rates of 61.2 ± 10.3% and were 1.63 times more likely to contract F-MuLV than contact females who were infected with F-MuLV at a rate of 37.5 ± 10.3%; this difference was statistically significant (GLMM; Z = -2.37, P = 0.018; Figure 3.2A; Table 3S.3). Similarly, populations founded with only index males had transmission frequencies of 50.0 ± 7.1% (n = 2) (across both sexes of contacts), which were 2.51 times higher than rates in populations containing only index females (19.9 ± 2.3%; n = 2); again, this difference was found to be statistically significant (GLMM; Z = -3.135, p = 0.002; Figure 3.2B). Further support for male-driven transmission was observed as populations containing both female and male index mice did not have higher rates of transmission (63.5 ± 4.5%; n = 2) than those containing only index males (GLMM; Z = 1.04, p = 0.299; Figure 57 3.2B), despite containing more index individuals in total. Furthermore, SFFV transmission was also male biased, though males were not significantly more likely to contract SFFV than females they are more likely to transmit SFFV. Across all populations contact males had marginally significantly higher rates of SFFV contraction (GLMM; Z = 1.79, p = 0.073; Figure 3.2C; Table 3S.2). Furthermore, there is 2.31 times more transmission in populations founded by only male index mice than those founded by only female index mice (GLMM; Z = 4.14, p < 0.0001; Figure 3.2D). Interestingly, populations founded with both male and female index animals did not have significantly higher SFFV transmission rates than female index populations (GLMM; Z = 1.74, p = .082; Figure 3.2D, 3S.3). Contact males showed a significantly increased F-MuLV titer over contact females (LMM; t =2.61, p = 0.011; Figure 3.3B; Table 3S.4), though there was no difference between the sexes in index animals. A strong correlation exists between virulence and titer in contact males (LM; r2 = 0.398, F1,20 = 13.23, n = 22, Table 3S.5), however, there is no relationship seen between virulence and titer in contact females (LM; r2 = 0.048 , F1,27 = 1.326, n = 29; Figure 3.3C). Furthermore, both male and female index mice had 2.39 times larger spleen mass than contact mice (LMM; t = 3.02, p = 0.004; Figure 3.3A; Table 3S.4). However, there is no significant difference in virulence between the sexes in either contact or index mice (LMM; t = 0.25, p = 0.807; Figure 3.3A). F-MuLV titer was significantly higher in index males and females when compared to contact animals (LMM; t = 4.35, p < 0.0001; Figure 3.3B). No effect of male social status was seen in regards to susceptibility as both socially dominant and nondominant male contact mice contracted F-MuLV at similar 58 frequencies (GLMM; Z = -0.08, p = 0.938; Figure 3.4A; Table 3S.6). Similar patterns were seen for rates of SFFV contraction (GLMM; Z = -0.03, p = 0.973; Figure 3.4B). Of males that contracted F-MuLV, social status did not influence spleen mass (LMM; t = 1.19, p = 0.250; Figure 3.4C; Table 3S.6); However, F-MuLV titers were higher in nondominant relative to dominant males (LMM; t = -2.59, p = 0.018: Figure 3.4D). 3.5 Discussion Within seminatural enclosures we demonstrated that horizontal transmission (and not vertical) is the primary mode of transmission, characterized asymmetrical rates of contraction and transmission between the sexes, that natural transmission results in less virulent infection than direct injection, that males which contract F-MuLV have higher titers than do females despite experiencing similar levels of virulence, and that male social status influences infection severity. Across all contact mice, rates of horizontal transmission of F-MuLV were 47.6%, while no offspring born to index females within enclosures tested positive. Male contact mice were 1.63 times more likely to contract FMuLV than contact females. Though there was no difference in spleen mass between male and female contact animals there was a significant increase (~ 1 order of magnitude) in titer in males over females. Index mice had 6.92 times the spleen mass as did contact mice. Finally, it was observed that social status did not influence contraction rates of FVC, nor the resulting virulence, but nondominant males were found to harbor F-MuLV titers that were ~ 2 orders of magnitude higher than dominant males. FVC horizontal transmission through sex and vertical transmission has been explicitly tested using inbred mouse strains in cages (Mirand et al., 1966; Portis et al., 59 1987); however, these studies did not take into account all modes of transmission simultaneously to determine their relative importance. Horizontal transmission is dominant to vertical transmission in this FVC system, as no evidence of vertical transmission was observed, even though previous studies have shown vertical transmission to occur at ~16% (Mirand et al., 1966). Because of the high rates of male contraction and transmission male-male agonistic encounters emerge as the likely main mode of FVC transmission. If sexual transmission were the main mode, we would predict patterns of high rates of female contraction and high rates of male transmission, which were not observed. The observations that males primarily contract and transmit FVC within seminatural enclosures, while possessing higher titers, but similar levels of virulence to females, is indicative that FVC, and potentially other MuLVs, may have an evolutionary history transmitting through males. Sex-specific transmission and virulence is a key life history characteristic of many parasites and pathogens (Zuk & McKean, 1996). Host sex can be an important selective pressure on pathogens, which are more likely to be transmitted or contracted by one sex than the other due to sex-based differences in behavior and physiology (Duneau & Ebert, 2012; Robinson & Klein, 2012). Sex-specific pathogen adaptation has been documented in invertebrate host-pathogen systems (Høeg, 1995; Werren et al., 2008; Rosenkranz et al., 2010), and in the vertebrate systems of Myxozoa in fish (Whipps & Kent, 2006), and the mite Spinturnix andegavinus in bats (Christe et al., 2007). Experimental demonstration of pathogen adaptation based on host sex is currently limited to invertebrates (e.g., Daphnia (Duneau et al., 2012)). Mammalian models, such as the FVC/Mus system, are needed to explore the generality of 60 these patterns in species with adaptive immunity. Mammalian models will also be better models for human disease, where recent evidence is indicating that sex-specific adaption may be an important phenomenon (Úbeda & Jansen, 2016). The finding that male social status influences F-MuLV titers, but not spleen size or prevalence, highlights the complicated interactions between social status, individual immune function, and virulence (reviewed in Fairbanks & Hawley, 2012). Two main hypothetical frameworks exist to explain the relationship between social status and cost of infection in vertebrates: 1) that dominant individuals expend more resources acquiring a territory and matings, leaving them less capable of investing in immune defenses to ward off infections (Folstad & Karter, 1992); and 2) that dominant individuals are in better condition, and therefore more resistant to pathogens as subordinates undergo stress associated with poor habitat and constant antagonism by dominant individuals (Sapolsky, 2004) (for recent review of both paradigms see Habig et al., 2015). Unfortunately, the unusual pattern between viral titers and spleen size (virulence) in this study can be viewed to support either paradigm. Specifically, dominant males may be getting sicker per unit virus as they have lower titers but similar sized spleens as nondominate males; conversely dominant males may be in better condition than subordinates, and thus able to better regulate their viral load. In regard to prevalence in social vertebrates, typically dominant individuals are more parasitized than subordinates due to more frequent social interactions (Habig et al., 2015); however, in this study both classes of males contracted F-MuLV at similar frequencies possibility due to male-male interactions as the primary driver of transmission (i.e., dominant males have more interaction with females but they do not 61 readily transmit the virus). To further dissect the interplay between social status, immune parameters, and pathogen virulence seminatural systems, like this one, are needed to help merge laboratory measures of immune function with truly natural systems. Although this study advances our understanding of how a model retrovirus transmits in a naturalistic setting, there are a few limitations in interpreting the results. First, variation exists in the infection duration of contact animals (i.e., some contact mice contract FVC earlier in the study than others) likely leading to inflated variation in titers and spleen mass of contact mice; this may cause us to miss patterns that would become apparent with more temporal control. Likewise, comparisons between contact and index mice are temporally disjointed in the sense that index mice are all infected prior to release into enclosures and contacts are infected throughout the study; additionally, index mice are likely infected with higher doses of virus via IP injection than contact mice. Moreover, vertical transmission may not have been seen in this study due to two reasons: first, wild-derived mice are much more resistant to FVC than the inbred mice used previously; second, vertical transmission was measured at 1-4 weeks postbirth in this study and anywhere from 5 to 8 months postbirth in previous studies (Mirand et al., 1966). Finally, the inability to use qPCR-based techniques to quantify SFFV due to expression of endogenous retroviral RNA limits our ability to integrate SFFV fully into our understanding of FVC transmission dynamics. Understanding how MuLVs naturally move through populations is important for more than just Mus populations, as related retroviruses have been found across taxa of terrestrial vertebrates and are thought to be important for emerging infectious diseases (Martin et al., 1999). Furthermore, the transmission of leukemia viruses such as Feline 62 Leukemia Virus in the endangered Florida Panther (Puma concolor coryi) are of great interest to conservationists trying to restore natural populations of these animals not to mention the effects of these viruses on domestic felines (Cunningham et al., 2008; O'Brien et al., 2012). The power of seminatural enclosures allows for the elucidation of the factors contributing to the natural history of murine leukemia like viruses. Furthermore, characterizations of the transmission dynamics of various pathogens under natural or seminatural conditions, such as the one done here, may provide insight into better combating emerging infectious disease, and established wildlife and human pathogens (Antonovics et al., 2017). Using seminatural enclosures and wild-derived mice we have characterized the transmission dynamics of a classic model retrovirus. In doing so, we have demonstrated that the prevailing understanding-that FVC is primarily vertically transmitted through mothers milk, or transmitted horizontally through sex (Mirand et al., 1966; Portis et al., 1987)-is not supported and that male-male transmission during aggressive interactions is likely the predominant mode of transmission. Furthermore, we illustrate that within seminatural enclosures males harbor higher titers than females while suffering similar virulence; collectively, with the observation that there is a strong correlation between nonspleen corrected titer and virulence in males and none in females suggests that FVC is male adapted. Additionally, we demonstrated that male social status influences aspects of infection in an unexpected way: though dominant and nondominant males contract the virus at similar rates, and experience similar levels of virulence, dominant males have lower titers than do nondominant individuals. Understating the natural transmission dynamics of model pathogens will lead to better interpretations of the results of 63 laboratory experiments by framing them in an ecological and evolutionary context, while studies with seminatural host-pathogen systems promise to bridge the gap between laboratory and field studies and potentially shed light on the complex interactions of sociality and disease ecology. 3.6 References Anderson, R.M. & May, R.M. 1982. Coevoltuion of hosts and parasites. Parasitology 85: 411-426. Antonovics, J. 2017. Transmission dynamics: critical questions and challenges. Philos. Trans. R. Soc. B 372. 1-8. Antonovics, J., Wilson, A.J., Forbes, M.R., Hauffe, H.C., Kallio, E.R., Leggett, H.C., et al. 2017. The evolution of transmission mode. Philos. Trans. R. Soc. B 372. 1-12. Arthur, R.F., Gurley, E.S., Salje, H., Arthur, R.F., Gurley, E.S., Salje, H., et al. 2017. Contact structure, mobility, environmental impact and behaviour: the importance of social forces to infectious disease dynamics and disease ecology. Philos. Trans. R. Soc. B 372. 1-9. Bates, D., Maechler, M., Bolker, B.M. & Walker, S. 2015. Fitting linear mixed-effects models using {lme4}. J. Stat. Softw. 67: 1-48. Carroll, L.S. & Potts, W.K. 2006. Functional genomics requires ecology. Adv. Study Behav. 36: 173-215. Carroll, L.S. & Potts, W.K. 2007. Sexual selection: Using social ecology to determine fitness differences. In: Rodent societies: An ecological & evolutionary perspective (J. O. Wolff & P. W. Sherman, eds), pp. 58-67. The University of Chicago Press, Chicago, IL. Chalfin, L., Dayan, M., Levy, D.R., Austad, S.N., Miller, R. a., Iraqi, F. a., et al. 2014. Mapping ecologically relevant social behaviours by gene knockout in wild mice. Nat. Commun. 5: 1-10. Nature Publishing Group. Christe, P., Glaizot, O., Evanno, G., Bruyndonckx, N., Devevey, G., Yannic, G., et al. 2007. Host sex and ectoparasites choice: Preference for, and higher survival on female hosts. J. Anim. Ecol. 76: 703-710. Cunningham, M.W., Brown, M.A., Shindle, D.B., Terrell, S.P., Hayes, K.A., Ferree, 64 B.C., et al. 2008. Epizootiology and management of feline leukemia virus in the florida puma. J. Wildl. Dis. 44: 537-552. Duneau, D. & Ebert, D. 2012. Host sexual dimorphism and parasite adaptation. PLoS Biol. 10: e1001271. Duneau, D., Luijckx, P., Ruder, L. & Ebert, D. 2012. Sex-specific effects of a parasite evolving in a female-biased host population. BMC Biol. 10: 104. BioMed Central Ltd. Ebert, D. 2013. The epidemiology and evolution of symbionts with mixed-mode transmission. Annu. Rev. Ecol. Evol. Syst. 44: 623-643. Ewald, P.W. 1983. Host-parasite relations , vectors, and the evolution of disease severity. Annu. Rev. Ecol. Syst. 14: 465-485. Fairbanks, B.M. & Hawley, D.M. 2012. Interactions between host social behavior, physiology, and disease susceptibility: the role of dominance status and social context. In: Ecoimmunology (G. E. Demas & R. J. Nelson, eds), pp. 440-467. New York, NY. Folstad, I. & Karter, A.J. 1992. Parasites, bright males, and the immunocompetence handicap. Am. Nat. 139: 603-622. Gaukler, S.M., Ruff, J.S., Galland, T., Kandaris, K. a., Underwood, T.K., Liu, N.M., et al. 2014. Low-dose paroxetine exposure causes lifetime declines in male mouse body weight, reproduction and competitive ability as measured by the novel organismal performance assay. Neurotoxicol. Teratol., doi: 10.1016/j.ntt.2014.11.002. Elsevier Inc. Gaukler, S.M., Ruff, J.S., Galland, T., Underwood, T.K., Kandaris, K.A., Liu, N.M., et al. 2016. Quantification of cerivastatin toxicity supports organismal performance assays as an effective tool during pharmaceutical safety assessment. Evol. Appl. 9: 685-696. Habig, B., Archie, E.A. & Habig, B. 2015. Social status, immune response and parasitism in males: a meta-analysis. Philos. Trans. R. Soc. B 370: 20140109. Hasenkrug, K.J. & Chesebro, B. 1997. Immunity to retroviral infection : The Friend virus model. Proc. Natl. Acad. Sci. U. S. A. 94: 7811-7816. Hausfater, G. & Watson, D.F. 1976. Social and reproductive correlates of parasite ova emissions by baboons. Nature 260: 619-21. Hawley, D.M., Jennelle, C.S., Sydenstricker, K. V. & Dhondt, A.A. 2007. Pathogen resistance and immunocompetence covary with social status in house finches 65 (Carpodacus mexicanus). Funct. Ecol. 21: 520-527. Høeg, J.T. 1995. The biology and life cycle of the Rhizocephala (Cirripedia). J. Mar. Biol. Assoc. United Kingdom 75: 517-550. Klein, S.L. 2000. The effects of hormones on sex differences in infection: from genes to behavior. Neurosci. Biobehav. Rev. 24: 627-638. Kosmider, O. & Moreau-Gachelin, F. 2006. From mice to human: The "two-hit model" of leukemogenesis. Cell Cycle 5: 569-570. Kubinak, J.L., Ruff, J.S., Whitney, C., Slev, P.R. & Potts, W.K. 2012. Experimental viral evolution to specific host MHC genotypes reveals fitness and virulence trade-offs in alternative MHC types. Proc. Natl. Acad. Sci. U. S. A. 109: 3422-3427. Kuznetsova, A., Brockhoff, P.B. & Christensen, R.H.B. 2016. lmerTest: Tests in linear mixed effects models. R package version 2.0-25. Looker, K.J., Garnett, G.P. & Schmid, G.P. 2008. An estimate of the global prevalence and incidence of herpes simplex virus type 2 infection. Bull World Heal. Organ 86: 805-812. Luong, L.T., Grear, D.A. & Hudson, P.J. 2009. Male hosts are responsible for the transmission of a trophically transmitted parasite, Pterygodermatites peromysci, to the intermediate host in the absence of sex-biased infection. Int. J. Parasitol. 39: 1263-1268. Australian Society for Parasitology Inc. Martin, J., Herniou, E., Cook, J. & Neill, R.W.O. 1999. Interclass transmission and phyletic host tracking in murine leukemia virus-related retroviruses. J. Virol. 73: 2442-2449. Mccallum, H., Fenton, A., Hudson, P.J., Lee, B., Levick, B., Norman, R., et al. 2017. Breaking beta: deconstructing the parasite transmission function. Philos. Trans. R. Soc. B 372. 1-11. Meagher, S., Penn, D.J. & Potts, W.K. 2000. Male-male competition magnifies inbreeding depression in wild house mice. Proc. Natl. Acad. Sci. U. S. A. 97: 3324- 3329. Metcalf, D., Furth, J. & Buffett, R.F. 1959. Pathogenesis of mouse leukemia caused by Friend Virus. Cancer Res. 19: 52-58. Mirand, E. a., Buffett, R.F. & Grace, J.T. 1966. Mode of transmission of Friend virus disease. Exp. Biol. Med. 121: 970-976. Morris, J.S., Ruff, J.S., Potts, W.K. & Carrier, D.R. n.d. Social dominance and locomotor 66 economy in male house mice. Proc. R. Soc. B. Myers, L. & Hasenkrug, K.J. 2009. Retroviral immunology: lessons from a mouse model. Immunol. Res. 43: 160-166. Nelson, A.C., Cauceglia, J.W., Merkley, S.D., Youngson, N. a, Oler, A.J., Nelson, R.J., et al. 2013a. Reintroducing domesticated wild mice to sociality induces adaptive transgenerational effects on MUP expression. Proc. Natl. Acad. Sci. U. S. A. 110: 19848-53. Nelson, A.C., Colson, K.E., Harmon, S. & Potts, W.K. 2013b. Rapid adaptation to mammalian sociality via sexually selected traits. BMC Evol. Biol. 13: 81. O'Brien, S.J., Troyer, J.L., Brown, M.A., Johnson, W.E., Antunes, A., Roelke, M.E., et al. 2012. Emerging viruses in the Felidae : Shifting paradigms. Viruses 4: 236-257. Padian, N.S., Shiboski, S.C. & Jewell, N.P. 1991. Female-to-male transmission of human immunodeficiency virus. JAMA 266: 1664-7. Perkins, S.E., Ferrari, M.F. & Hudson, P.J. 2008. The effects of social structure and sexbiased transmission on macroparasite infection. Parasitology 135: 1561-1569. Portis, J.L., McAtee, F.J. & Hayes, S.F. 1987. Horizontal transmission of murine retroviruses. J. Virol. 61: 1037-44. R Core Team. 2016. R: a language and environment for statistical computing. Vienna, Austria. Robinson, D.P. & Klein, S.L. 2012. Sex differences in immune responses to viruses. In: Ecoimmunology (G. E. Demas & R. J. Nelson, eds), pp. 497-529. Oxford University press, New York, NY. Rosenkranz, P., Aumeier, P. & Ziegelmann, B. 2010. Biology and control of Varroa destructor. J. Invertebr. Pathol. 103. S96-S119. Ruff, J.S., Hugentobler, S.A., Suchy, A.K., Sosa, M.M., Tanner, R.E., Hite, M.E., et al. 2015a. Compared to sucrose, previous consumption of fructose and glucose monosaccharides reduces survival and fitness of female Mice. J. Nutr. 145: 434- 441. Ruff, J.S., Saffarini, R.B., Ramoz, L.L., Morrison, L.C., Baker, S., Laverty, S.M., et al. 2015b. Fitness assays reveal incomplete functional redundancy of the hoxa1 and hoxb1 paralogs of mice. Genetics 201: 727-736. Ruff, J.S., Suchy, A.K., Hugentobler, S. a, Sosa, M.M., Schwartz, B.L., Morrison, L.C., et al. 2013. Human-relevant levels of added sugar consumption increase female 67 mortality and lower male fitness in mice. Nat. Commun. 4: 2245. Nature Publishing Group. Ruscetti, S.K. 1999. Deregulation of erythropoiesis by the Friend spleen focus-forming virus. Int. J. Biochem. {&} cell Biol. 31: 1089-1109. Sapolsky, R.M. 2004. Social status and health in humans and other animals. Annu. Rev. Anthropol. 33: 393-418. Singleton, G.R. & Krebs, C.J. 2007. The secret world of wild mice. In: The mouse in biomedical research, pp. 25-51 Stephenson, J.F., Young, K.A., Fox, J., Jokela, J., Cable, J., Perkins, S.E., et al. 2017. Host heterogeneity affects both parasite transmission to and fitness on subsequent hosts. doi: 10.1098/rstb.2016.0093. Thaler, M.M., Park, C.K., Landers, D. V, Wara, D.W., Houghton, M., VerremanWauters, G., et al. 1991. Vertical transmission of hepatitis C virus. Lancet 338: 17- 18. Úbeda, F. & Jansen, V.A. 2016. The evolution of sex-specific virulence in infectious diseases. Nat. Commun. 7: 13849. Walker, J.G., Plein, M., Morgan, E.R., Vesk, P.A. & Walker, J.G. 2017. Uncertain links in host - parasite networks : lessons for parasite transmission in a multi-host system. Philos. Trans. R. Soc. B 372. 1-14. Werren, J.H., Baldo, L. & Clark, M.E. 2008. Wolbachia: master manipulators of invertebrate biology. Nat. Rev Microbiol 6: 741-751. Whipps, C.M. & Kent, M.L. 2006. Phylogeography of the cosmopolitan marine parasite Kudoa thyrsites (Myxozoa: Myxosporea). J. Eukaryot. Microbiol. 53: 364-373. Zuk, M. 1996. Disease, endocrine-immune interactions, and sexual selection. Ecology 77: 1037-1042. Zuk, M. & McKean, K. a. 1996. Sex differences in parasite infections: Patterns and processes. Int. J. Parasitol. 26: 1009-1024. 68 Figure 3.1. Seminatural enclosure and population design. (A) Seminatural enclosure set up with six territories with four optimal territories (bins) and two suboptimal territories (small bins). (B) Study design showing three different population design groups: female and male index (left), male only index (center) and female only index (right). 69 Figure 3.2. Transmission dynamics of F-MuLV and SFFV in seminatural enclosures by contact (A-C) and index (B-D) mouse sex. (A) Male contact mice (squares) contracted F-MuLV at rates 1.63 times higher than contact females (circles). (B) Populations founded with only index males (filled triangles) had transmission rates 2.51 times higher across all contacts than those founded with only index females (open triangles), while rates did not differ in populations with only male index relative to those containing index animals of both sexes (diamonds). (C) No significant difference was seen between contraction in males and females in SFFV. However, male populations had 2.31 times higher transmission than females. Interestingly, populations founded with both male and female index animals have reduced transmission efficiency than those founded by only index males and are only marginally significantly different from those founded solely by female index mice. Data points represent populations (n = 7), lines represent population means, and error bars standard error. * Indicates a P value < 0.05., **< 0.01, ***< 0.001. 70 Figure 3.3. Spleen mass (A), F-MuLV titers (B), and their correlation (C) in female and male mice from seminatural enclosures. (A) Index mice (open symbols) of both sexes had spleens 6.92 times the mass of contact (closed symbols) animals, a log scale is used to enhance visualization. (B) Males (squares) had higher F-MuLV titers, by ~ 1 order of magnitude, than females (circles), but posthoc tests only confirmed this sexually dimorphic pattern between contact mice. Additionally, index mice of both sexes had titers ~ 2 orders of magnitude higher than contacts. (C) Though spleen mass and F-MuLV titers were positively correlated in contact males, no such relationship was observed in females and the relationship between spleen mass and FMuLV titers significantly differed between the sexes. Data points represent individual mice, in A and B lines indicate means and error is standard error, while in panel C lines are regressions. from post-hoc Tukey multiple comparison tests. Dashed lines indicates that origin is below zero. * Indicates a P value < 0.05, ** < 0.01, *** < 0.001 71 Figure 3.4. Influence of male social dominance on F-MuLV (A) and SFFV (B) contraction, virulence (C), and titer (D) in seminatural enclosures. (A) The dominance status of contact males did not influence their likelihood of contracting FMuLV; (B) this was also the case for SFFV. (C) Likewise, virulence as measured by spleen mass was not influenced by social status. (D) However, F-MuLV titers were higher in nondominant relative to dominant males Data points represent populations (n = 7) in A, and individuals (n = 22) in B and C. lines represent means, and error bars standard error. Dashed line indicates that origin is below zero. 72 3.7 Supplementary Material Figure 3S.1. Gel showing the ~170bp SFFV fragment. Positive control (lane 2) shows the positive SFFV RNA fragment at ~170bp whereas the negative RNA controls (lanes 34) do not. The ladder is represented in lane 1 and has the 500bp, 200bp and 100bp bands labelled. 73 Figure 3S.2. Spleen mass (A) and F-MuLV titer (B) in female and male caged mice. (A) Females had spleens 3.46 times the mass of those of males twelve days postinfection; a log scale was used to enhance visualization. (B) A trend was detected in regards to females having higher F-MuLV titers than males. **Indicates a P value < 0.01). 74 Figure 3S.3. Influence of population design on virulence and titer(A) and F-MuLV titer (B). No significant influence of population design was seen on either spleen mass or F-MuLV titer of contact mice. For statistics see (Table 3S.5); no comparisons between groups were significantly different. Dashed line indicates that origin is below zero. 75 Table 3S.1. Summary of LM results concerning viral titers and spleen mass of caged animals. F-MuLV titers (log transformed) by spleen mass (log transformed) of both sexes Fixed effects Estimate Std. Error t value Pr(>|t|) Intercept (Female) 2.49 0.43 5.75 <0.0001*** Sex (Male) -0.30 0.88 1.07 0.296 Spleen mass (log transformed) 0.94 0.54 -0.55 0.209 Sex (Male) X Spleen mass (log transformed) 1.14 0.80 1.43 0.165 SFFV RNA titers (log transformed) by spleen mass (log transformed) of both sexes Fixed effects Estimate Std. Error t value Pr(>|t|) Intercept (Female) -0.35 0.40 -0.88 0.39 Spleen mass (log transformed) 1.13 0.46 2.45 0.02* Sex (Male) -0.005 0.73 -0.006 0.995 Sex (Male) X Spleen mass (log transformed) 0.19 0.67 0.29 0.77 SFFV RNA titers (log transformed) by F-MuLV titers (log transformed) of both sexes Fixed effects Estimate Std. Error t value Pr(>|t|) Intercept (Female) 4.91 2.10 2.34 0.03* SFFV titer (log transformed) -0.37 0.42 -0.88 0.39 Sex (male) -2.50 2.34 -1.07 0.30 SFFV Titer (log transformed) x Sex (Male) (log transformed) 0.35 0.48 0.72 0.48 * Indicates a P value < 0.05., ***< 0.001. 76 Table 3S.2. Summary of LMMs concerning the virulence and titer of F-MuLV of contact mice by population design. Virulence (Spleen mass) of contact mice by index sex LMM (groups: 7; observations = 51) Random effects Variance Std. Deviation Population 0.020 0.142 Fixed effects Estimate Std. Error t value Pr(>|t|) Intercept 0.717 0.121 5.92 0.001** Sex (Female) -0.123 0.103 -1.19 0.241 Index sex (Female) -0.060 0.199 -0.30 0.771 Index sex (Male) 0.152 0.164 0.93 0.405 F-MuLV titers (log transformed) of contact mice by index sex LMM (groups: 7; observations = 51) Random effects Variance Std. Deviation Population 0.061 0.247 Fixed effects Estimate Std. Error t value Pr(>|t|) Intercept -0.561 0.360 -1.56 0.162 Sex (Female) -1.409 0.387 -3.65 0.001 ** Index sex (Female) 0.084 0.602 0.14 0.892 Index sex (Male) 0.626 0.431 1.45 0.265 ** Indicates a P value < 0.01, *** < 0.001. 77 Table 3S.3. Summary of GLMM results for FVC transmission dynamics in seminatural enclosures. F-MuLV Transmission by contact and index sex GLMM with binomial distribution and logit link (groups: populations = 7; observations = 14) Random effects Variance Std. Deviation Population 0.00 0.00 Fixed effects Estimate Std. Error Z value Pr(>|z|) Intercept (Contact Female, Index Female) -2.12 0.56 -3.81 <0.0001*** Sex of contact mouse (Male) 1.16 0.49 2.37 0.018* Sex of index mice (Male only) 1.85 0.59 3.14 0.002** Sex of index mice (Both Female and Male) 2.35 0.61 3.88 <0.0001*** SFFV Transmission by contact and index sex GLMM with binomial distribution and logit link (groups: populations = 7; observations = 14) Random effects Variance Std. Deviation Population 0.00 0.00 Fixed effects Estimate Std. Error Z value Pr(>|z|) -1.94 0.55 -3.54 0.0004*** Intercept (Contact Female, Index Female) Sex of contact mouse (Male) 0.91 0.50 1.79 0.0727 Sex of index mice (Male only) 2.55 0.62 4.14 <0.0001*** Sex of index mice (Both Female and Male) 1.02 0.59 1.74 0.0823 F-MuLV Transmission to contact males by social status GLMM with binomial distribution and logit link (groups: populations = 7; observations = 14) Random effects Variance Std. Deviation Population 0.00 0.00 Fixed effects Estimate Std. Error Z value Pr(>|z|) Intercept (Nondominant contact males) 0.17 0.41 0.41 0.683 Social status (Dominant) -0.05 0.64 -0.08 0.938 SFFV Transmission to contact males by social status GLMM with binomial distribution and logit link (groups: populations = 7; observations = 14) Random effects Variance Std. Deviation Population 0.00 0.00 Fixed effects Estimate Std. Error Z value Pr(>|z|) Intercept (Nondominant contact males) -0.20 0.45 -0.45 0.655 Social status (Dominant) -0.02 0.65 -0.03 0.973 * Indicates a P value < 0.05., ** < 0.01, *** < 0.001. 78 Table 3S.4. Summary of LMM results concerning titer and virulence measures between index and contact animals. Virulence (Spleen mass) of contact and index mice by sex LMM (groups: populations = 7; observations = 75) Random effects Variance Std. Deviation Population 0.0137 0.117 Fixed effects Estimate Std. Error t value Pr(>|t|) Intercept 0.704 0.098 7.19 <0.0001*** Sex (Male) 0.032 0.130 0.25 0.807 Type (Index) 0.489 0.162 3.02 0.004 ** Interaction (Male:Index) 0.327 0.238 1.37 0.176 FMuLV titers (log transformed) of contact and index mice by sex LMM (groups: populations = 7; observations = 75) Random effects Variance Std. Deviation Population 0.353 0.594 Fixed effects Estimate Std. Error t value Pr(>|t|) Intercept -1.475 0.364 -4.06 0.001 ** Sex (Male) 1.095 0.419 2.61 0.011 * Type (Index) 2.310 0.531 4.35 <0.0001*** Interaction (Male:Index) -0.181 0.789 -0.23 0.819 * Indicates a P value < 0.05., ** < 0.01, *** < 0.001. 79 Table 3S.5. Summary table of correlations of titer and virulence in contact animals FMuLV titers (log transformed) by Virulence (Spleen mass) of contact females 2 LM; r = 0.048 , F1,27 = 1.326, n = 29 Fixed effects Estimate Std. Error t value Pr(>|t|) Intercept 1.077 0.385 2.80 0.009** Spleen mass (log transformed) -0.583 0.500 -1.17 0.253 FMuLV titers (log transformed) by Virulence (Spleen mass) of contact males 2 LM; r = 0.398 , F1,20 = 13.23, n = 22 Fixed effects Estimate Std. Error t value Pr(>|t|) Intercept 0.053 0.544 0.10 0.924 Spleen mass (log transformed) 2.472 0.680 3.63 0.002** FMuLV titers (log transformed) by Virulence (Spleen mass) of both contact sexes 2 LM; r =0.421 , F3,47 =11.4, n = 51 Fixed effects Estimate Std. Error t value Pr(>|t|) Intercept (Female) 1.077 0.397 2.71 0.009** Sex (Male) -0.583 0.657 -1.13 0.126 Spleen mass (log transformed) -1.024 0.516 -1.56 0.264 Sex (Female) X Spleen mass (log transformed) 3.055 0.832 3.67 <0.0001*** ** Indicates a P value < 0.01, *** < 0.001. 80 Table 3S.6. Summary table of GLMMs of male social status on F-MuLV titer and virulence F-MuLV Transmission to contact males by social status GLMM with binomial distribution and logit link (groups: populations = 7; observations = 14) Random effects Variance Std. Deviation Population 0.00 0.00 Fixed effects Estimate Std. Error Z value Pr(>|z|) Intercept (Nondominant contact males) 0.17 0.41 0.41 0.683 Social status (Dominant) -0.05 0.64 -0.08 0.938 Virulence (Spleen mass) of contact males by social status LMM (groups: populations = 7; observations 22) Random effects Variance Std. Deviation Population 0.00 0.01 Estimate Std. Error t value Pr(>|t|) Intercept (Nondominant contact males) 0.80 0.09 8.94 <0.0001*** Social status (Dominant) -0.17 0.14 -1.19 Fixed effects 0.250 FMuLV titers (log transformed) of contact males by social status LMM (groups: populations = 7; observations 22) Random effects Variance Std. Deviation Population 0.08 0.29 Estimate Std. Error t value Pr(>|t|) Intercept (Nondominant contact males) 1.22 0.38 3.21 0.007** Social status (Dominant) -1.48 0.57 -2.59 0.018* Fixed effects CHAPTER 4 MICROBIAL TRANSPLANT INCREASES SUSCEPTIBILITY TO A SYSTEMIC VIRAL INFECTION 4.1 Abstract Many physiological processes in the host are influenced by the microbial communities living within the gut. These influences range from digestion of fibrous foods to immunity influencing susceptibility to enteric pathogens. Though much work has shown a strong relationship with enteric pathogen susceptibility and the microbiome, little has been done on how changes in the microbiome influence systemic infection. We seek to differentiate between two hypotheses that may lead to changes in FVC virulence: (1) that because of trade-offs between resources available for systemic and enteric immunity of the host, microbial transfers of the gut microbiota will cause an increase in susceptibility to the systemic pathogen Friend Virus Complex (FVC); (2) there will be a shift in FVC susceptibility that corresponds with host resistance of the donor microbiome. Using established microbial transfer techniques, we reconstituted two different treatment groups with different microbiotas: native treatments consisted of hosts receiving gut microbiome stock from animals with their same MHC genotype, and novel treatments consisted of hosts receiving gut microbiome stock from animals of a different 82 MHC genotype. Novel treatment groups had higher titer and virulence than the native groups, suggesting that the animals in the novel treatment are more susceptible to FVC infection. Furthermore, there was a marginally significant reduction in overall lymphocyte percentage in the novel treatment group, suggesting animals receiving novel microbiomes are potentially immunocompromised; However, there was a significant increase in proportion of CD4+ T-cells These results suggest that caution should be taken when fecal transplants are done to combat enteric infections as the treatment may cause a patient to become more susceptible to other systemic pathogens and that more work needs to be done in order to fully understand the immune factors leading to this virulence increase. 4.2 Introduction Gut microbial communities are known to play a role in the maintenance of homeostasis and health in the host where specific microbial constituents have been shown to aid in digestion of fibrous foods, immunity, and in the ability to detoxify toxins (reviewed in Dethlefsen et al., 2007; Sekirov et al., 2010). Colonization immunity, or the ability of the microbiome associated with a host to confer resistance to enteric pathogens, has implications in human health; leading to a potential therapeutic capacity to combat enteric disease (Stecher et al., 2007; Sekirov et al., 2010; Buffie & Pamer, 2013). However, it is still unclear to what extent physiological processes can be influenced, positively or negatively, by microbiome transfer. Microbiomes residing within the host gut have a large influence on host immune function and changes in microbial composition of communities alters overall immune 83 function (Maynard et al., 2012). These dynamics are clearly seen in germ-free mouse systems where the absence of microbiota leads to severely immunocompromised hosts (Cebra, 1999; Khosravi et al., 2014). Microbiome transfers have shown shifts in immune function; even the ability to shift susceptibility to enteric pathogens (Willing et al., 2011; Kubinak et al., 2015). These immunity shifts have been found to be driven by specific immune cell subtypes and differential cytokine signaling that change when interacting with novel microbial constituents (Chen et al., 1999; Hensley-McBain et al., 2016). Relationships with particular constituents associated with microbial colonization (i.e., liposaccharide [LPS]) and increased inflammation have also be found, which may shift immunity away from antiviral immunity, allowing viruses to replicate more freely (Pullen et al., 1995; Jude et al., 2003). However, the microbiota may be interacting with systemic components of the immune system, which may, in turn be influencing interactions with systemic pathogens. The major histocompatibility complex (MHC) is the key interface involved in the differentiation of self from nonself and mediates much of the adaptive immune response. It is the most diverse region in the vertebrate genome with over 2,400 alleles for one locus in the human population alone (Bodmer et al., 1986; Wegner et al., 2003; Robinson et al., 2015). MHC genotypes are highly associated with resistance and susceptibility profiles to numerous pathogens (Jeffery & Bangham, 2000). Furthermore, the host MHC genotype has a strong influence on shaping the microbial communities in the gut, which has been demonstrated using MHC congenic mice (Mus musculus) (Kubinak et al., 2015). Collectively, the centrality of the MHC in adaptive immunity, its high degree of polymorphism, and its proven ability to sculpt microbial communities make MHC 84 congenic mouse lines an ideal model in which to characterize the influence of microbiota on systemic pathogen virulence. Friend Virus Complex (FVC) is a systemic murine retrovirus that is made up of two independent genomes that work synergistically to cause an erythro-proliferative disease: Friend Murine Leukemia Virus (F-MuLV) and Spleen Focus Forming (SFFV) (Friend, 1957; Ruscetti, 1999). Due to FVC being well characterized and its ability to be used as an acute acting virus, it is ideal to use in the short infection study completed here (Hasenkrug & Chesebro, 1997; Kubinak et al., 2012, 2013). Previous work with FVC has shown increases in virulence in the absence of host microbiota; however, it has not been assessed if different microbial communities influence FVC virulence (Mirand & Grace, 1963). Here we test the influence fecal microbial transplants (FMT) have on virulence of a systemic pathogen-FVC. We accomplish the FMT by transferring the gut microbial constituents of two different strains of MHC congenic mice. Two independent groups were created: the novel group received microbiota from the dissimilar MHC genotype whereas the native group received microbiota from the same MHC genotype. We seek to differentiate between two hypotheses: (1) that transferring gut microbiota between two different genotypes of hosts will increase the susceptibility to viral infection by FVC due to the tradeoff in immunity, or (2) there will be a shift in susceptibly that corresponds to the resistance of the host microbiome transferred. 85 4.3 Methods Mice (Mus musculus) were purchased from Jackson laboratories and bred under specific pathogen free conditions. Two different strains of MHC congenic Balb/c mice were used: BALB/cByJ (Balb/cdd), and C.B10-H2b (Balb/cbb). All animals used in the experiment were females between two and six months of age. Twenty-eight animals were initially placed on antibiotic treatment before infection and were grouped into four treatments (Figure 4.1), two control groups and two treatment groups. Due to severe weight loss during antibiotic treatment, seven animals were removed from the study before infection leading to 21 total animals being infected. All viral stocks were generated from a previously described biological clone of the NB-tropic strain of Friend Virus Complex (Kubinak et al., 2012). The viral stock used to infect the experimental animals was generated by passaging the biological clone through one round of highly susceptible DBA mice in order to produce a virus that was virulent enough to cause disease in both host genotypes used in the study. After infection, all spleens were collected, mixed in a 1:1 ratio with 1x PBS, homogenized, and centrifuged at 5,000 RPM for 5 min. Supernatants were collected, mixed with each other, then aliquoted and stored at -80°C. All viral stocks were kept frozen until use and only thawed once. All infections were intraperitoneal (IP) and were injected at a 20 µL volume. After infection, the virus was incubated in vivo for 12 days at which time animals were sacrificed. At sacrifice, spleens were harvested and portions were removed and weighed for use in flow cytometry. The remainder of each spleen was weighed then homogenized in a 1:1 ratio with 1x PBS. 86 To obtain estimates of the number of viral insertions into the host genome, quantitative PCR (qPCR) was conducted on F-MuLV using a previously described protocol (Kubinak et al., 2012). Due to a strong correlation and previous work showing similarities between F-MuLV and SFFV titers only F-MuLV was quantified here (Kubinak et al., 2012). Viral titers were standardized to total GAPDH estimates to obtain a viral genome by GAPDH ratio. Furthermore, the F-MuLV/GAPD ratios calculated previously were multiplied by spleen mass in order to obtain total viral genomes by spleen estimates. All qPCR reactions were conducted on a Biorad CFX96 system. All microbiota stocks were generated by sacrificing four animals of each MHC genotype and collecting microbiota from their intestines (duodenum to anus). The microbiome from each genotype was then combined in a 1:1 ratio with 1x PBS in order to create the stocks used for oral gavage. All microbiota stocks were spun at 5,000 RPM for 5 min., supernatants were aliquoted, flash frozen in liquid nitrogen and then kept at 80°C until use. Animals were treated daily for two weeks with 0.5mg/mL of each of the following antibiotics: neomycin sulfate, ampicillin, gentamicin, and erythromycin. Following antibiotic treatment mice were put into their respective native or novel groups. Native groups consisted of animals receiving a daily oral gavage (100 µL) of microbiota from a donor of their own genotype (i.e., Balb/cbb genotype animals received Balb/cbb microbiota). Novel groups consisted of animals that received a daily oral gavage (100 µL) of microbiota from the different host genotype (i.e., Balb/cbb genotype animals received Balb/cdd microbiota). Oral gavage occurred daily for one week to allow for sufficient reconstitution to occur. One day after the last oral gavage procedure, each 87 animal was infected with a standardized volume of stock FVC. Flow Cytometry was conducted at the flow cytometry core at the University of Utah using a FACs Canto cytometer running FACsDiva software. All samples were isolated from the spleen and run through a 0.22µ filter in order to obtain a single cell suspension. Cells were stained per BioLegend cell staining protocol using the following conjugated antibody stains: antimouse CD4-FITC, antimouse CD8-PercyPcy5.5, antimouse CD3-Pacific Blue, antimouse CD11a-PE/Cy7, antimouse CD19-PE. These conjugated antibodies were chosen to obtain estimates of standing and activated T-cell abundance along with standing and activated B-cell abundance. All conjugated antibody stains were ordered from BioLegend and used per manufacturer recommendations. Linear models were run with host genotype and microbiome treatment groups as fixed effects and either spleen mass, proviral load, or spleen corrected proviral load as the response variable. These models accounted for variation due to both host genotype and microbiome treatment group and their interactions. Post-hoc t-tests were completed when a significant interaction term was seen between genotype and treatment group. When interaction terms in the linear models were not significant those terms were removed from the model as in the titer and spleen corrected titer models. Models were conducted with 21 observations each and an intercept set at zero. All statistical analyses were conducted on log10 transformed data to improve normality and completed in R statistical modeling software (3.3.1) (R Core Team, 2016). 88 4.4 Results Novel microbiome caused a significant increase in virulence at the model intercept (0.2218 ± 0.0681) when compared to spleen size from animals with native microbiomes (LM; t = 3.255, p = 0.005, Table 4.1). There was also a large effect of host genotype with the more susceptible genotype (Balb/cdd) having significantly higher virulence than the resistant genotype (Balb/cbb) (LM; t = 14.44, p < 0.0001). Furthermore, the interaction between host genotype and treatment was also significant (LM; t = -2.177, p = 0.045); though post hoc tests indicate the effect is only significant in the Balb/cbb genotype (Post hoc Welch T test; Balb/cdd, t = -0.031, p = 0.976; Balb/cbb, t = -3.085, p = 0.0150) (Figure 4.2A-B). Mice with novel microbiomes had marginally significantly increased (0.4922 ± 0.2339) F-MuLV titers relative to those with native microbiomes (LM; t = 1.75, p = 0.098). Furthermore, the Balb/cdd genotype showed a significant increase of F-MuLV viral titer over the Balb/cbb genotype (LM; t = 4.44, p < 0.001). There was not a significant interaction between host genotype and microbiome treatment as both genotypes showed an increase in titer in novel treatment, so the interaction term was removed from the final model (LM; t = -1.190, p = 0.254) (Figure 4.2C-D). The total number of viral genomes per spleen also showed a marginally significant increase (0.6646 ± 0.2515) in hosts receiving novel microbiomes over native microbiomes (LM; t = 1.92, p = 0.071) (Figure 4.2E-F). Again, host genotype had a significant effect on spleen corrected titer with Balb/cdd genotypes having increased levels over Balb/bb genotypes (LM; t = 6.84, p < 0.0001). There was no significant interaction showing that both genotypes are acting similarly when novel microbiomes are 89 present, so the interaction term was not included in the final model (LM; t = -1.527, p = 0.1464). In novel mice, there was a marginally significant reduction on overall lymphocyte proportion (LM: t = -1.90; p = 0.074; Figure 4.3; Table 4.2); however, no other significant relationships between tested cell subsets were seen. However, activated CD3+ and CD4+ T-cell subsets showed marginally significant and significant (respectively) increases in proportion when novel microbiomes were seen (CD3+, LM: t = 1.982, p = 0.0649; CD4+, LM: t = 2.36, p = 0.031; Figure 4.3; Table 4.2). 4.5 Discussion The results presented here support the hypothesis that there may be a tradeoff in immunity between systemic and enteric immunity of the host, where the enteric immune response is pulling resources away from the systemic response, as we did not see a shift in susceptibility corresponding with resistance to FVC as would be expected in hypothesis 2. Comparisons between mice receiving either native or novel microbiome show an increase in susceptibility, though not significant with each measure, to FVC when a novel microbiome is seen; however, the virulence is driven by the significant increase in the Balb/cbb genotype. Furthermore, a marginally significant reduction in overall lymphocyte proportion was seen in hosts receiving novel microbiomes, potentially leaving them less able to fight a systemic pathogen; this result is likely driven by a sharp reduction in lymphocytes when Balb/cbb hosts receive Balb/cdd microbiota. Although lymphocyte proportion was reduced by a marginal amount, there were increases in CD3+ and CD4+ T-cell subsets, again in the mice receiving a novel 90 microbiome. The patterns seen in the lymphocyte, but not the downstream T-cells, suggests that there are other important subsets playing a role and need to be determined in future work. It is thought that cross talk between host and microbiome is important in the microbiomes ability to influence disease outcomes of pathogens (Buffie & Pamer, 2013). Multiple studies have demonstrated the interconnectivity of the host and microbiome; for example, how the Major histocompatibility complex (MHC) shapes the microbial community in the gut and how the microbial constituents of the gut influence cytokine signaling and immune response outside of the gut (Kubinak et al., 2015). However, little work has been completed that has shown an influence of differential communities of microbiota on virulence of systemic pathogens. Mirand and colleagues in the 1960s used germ free animals to understand the effect the microbiome has on FVC virulence (Mirand & Grace, 1963); though they showed an increase in virulence in germ-free animals, this may be due to the severely immunocompromised nature of germ-free animals as their immune systems never have the opportunity to mature (Marietta et al., 2015). Using antibiotic treatment allows us to determine the effect of gut microbiota with fully immunologically competent animals. Furthermore, other immune responses such as T helper, IL-23, IL-17 and development of lymphoid tissue have been correlated with microbiota shifts (Cebra et al., 1998; Mazmanian et al., 2005; Zaph et al., 2008). More specifically, T-cell effector cells, such as Th17+, Th22 and CD4+ cell, fluctuate in number when rhesus macaques are infected with simian immunodeficiency virus (SIV) after the microbiota is perturbed (HensleyMcBain et al., 2016). Here we do not see a strong influence on the basic systemic 91 adaptive immune response (i.e. CD4+, CD8+ and B-cells). However, the drastic decrease in total lymphocyte proportion suggests there is another cell type that is influencing the effect. These data suggest more detailed immune cell phenotyping is warranted and perhaps from more tissues than just spleen. Due to human health related pathogens, such as causative agents for diarrhea, the majority of studies involving microbiota have been done in the context of enteric pathogens (Kolling et al., 2012). As a result, specific microbial constituents have been found that confer resistance to enteric pathogens, such as Salmonella enterica or Clostridium sp. (reviewed in Kamada et al., 2013). However, these same relationships are not known for systemic pathogens. Moreover, the work presented here indicates that microbiota transferred between individuals influences the virulence of a systemic pathogenic infection. Furthermore, the finding that there is an increase in susceptibility when microbiota is transferred between MHC congenic animals, provides further evidence that MHC may be influential in shaping the protective nature of the microbial community in the gut. The increase in fitness and virulence of FVC may be due to multiple physiological responses such as immune resources being taken away from the systemic compartment to fight the novel microbiome in the gut or it may also be a change in the target cell number present when reconstitution occurs. The latter hypothesis seems less likely as data suggest that animals treated with antibiotics have reduced myeloid precursor cells suggesting a reduction in target cells for FVC to infect (Hill et al., 2012; Khosravi et al., 2014). It has become clear in the last decade that the resident microbiome is acting like another organ of the host and that perturbations to that community have rippling effects 92 throughout host physiology, especially on immunity (Rohmer et al., 2011). By treating animals with antibiotics then transferring intact microbiomes, we simulate a more ecologically relevant situation, as these conditions occur naturally in humans or domestic animals when they are prescribed antibiotics for pathogenic bacterial infections or during microbial transplant treatment routines (Sullivan et al., 2001; Noverr & Huffnagle, 2004). Furthermore, fecal transplants are becoming more accepted and commonplace, especially due to the success medical professionals have had in treating Clostridium difficile infections with fecal transplants from healthy donors (Bakken et al., 2011; Vrieze et al., 2013). Though more work is needed to determine the generality of this pattern across systemic pathogens; the findings presented here suggest that caution is warranted during fecal transfers as treating the enteric disease could make the host more susceptible to a systemic viral infection. 4.6 References Bakken, J.S., Borody, T., Brandt, L.J., Brill, J. V., Demarco, D.C., Franzos, M.A., et al. 2011. Treating clostridium difficile infection with fecal microbiota transplantation. Clin. Gastroenterol. Hepatol. 9: 1044-1049. Elsevier Inc. Bodmer, W.F., Trowsdale, J., Young, J. & Bodmer, J. 1986. Gene clusters and the evolution of the major histocompatibility system. Philos. Trans. R. Soc. Lond. B. Biol. Sci. 312: 303-315. Buffie, C.G. & Pamer, E.G. 2013. Microbiota-mediated colonization resistance against intestinal pathogens. Nat Rev Immunol 13: 790-801. Nature Publishing Group. Cebra, J.J. 1999. Influences of microbiota on intestinal immune system development. Am. J. Clin. Nutr. 69: 1046-1051. Cebra, J.J., Periwal, S.B., Lee, G., Lee, F.A.N. & Shroff, K.E. 1998. Development and maintenance of the gut-associated lymphoid tissue (GALT): The roles of enteric bacteria and viruses. Dev. Immunol. 6: 13-18. 93 Chen, T., Isomaki, P., Rimpilainen, M. & Toivanen, P. 1999. Human cytokine responses induced by Gram-positive cell walls of normal intestinal microbiota. Clin. Exp. Immunol. 118: 261-267. Dethlefsen, L., McFall-Ngai, M. & Relman, D.A. 2007. An ecological and evolutionary perspective on human-microbe mutualism and disease. Nature 449: 811-818. Friend, C. 1957. Cell-free transmission in adult swiss mice of a disease having the character of a leukemia. J. Exp. Med. 105. 307-318. Hasenkrug, K.J. & Chesebro, B. 1997. Immunity to retroviral infection : The Friend virus model. Proc. Natl. Acad. Sci. U. S. A. 94: 7811-7816. Hensley-McBain, T., Zevin, A.S., Manuzak, J., Smith, E., Gile, J., Miller, C., et al. 2016. The effects of fecal microbial transplantation on microbiome and immunity in SIVinfected macaques. J. Virol. 90: JVI.00099-16. Hill, D.A., Siracusa, M.C., Abt, M.C., Kim, B.S., Kobuley, D., Kubo, M., et al. 2012. Commensal bacteria - derived signals regulate basophil hematopoiesis and allergic inflammation. Nat. Med. 18: 538-546. Nature Publishing Group. Jeffery, K.J. & Bangham, C.R. 2000. Do infectious diseases drive MHC diversity? Microbes Infect. 2: 1335-41. Jude, B. a, Pobezinskaya, Y., Bishop, J., Parke, S., Medzhitov, R.M., Chervonsky, A. V, et al. 2003. Subversion of the innate immune system by a retrovirus. Nat. Immunol. 4: 573-578. Kamada, N., Chen, G.Y., Inohara, N. & Núñez, G. 2013. Control of pathogens and pathobionts by the gut microbiota. Nat. Immunol. 14: 685-90. Khosravi, A., Yanez, A., Price, J.G., Chow, A., Merad, M., Goodridge, H.S., et al. 2014. Gut microbiota promote hematopoiesis to control bacterial infection. Cell Host Microbe 15: 374-381. Kolling, G., Wu, M. & Guerrant, R.L. 2012. Enteric pathogens through life stages. Front. Cell. Infect. Microbiol. 2: 114. Kubinak, J.L., Ruff, J.S., Cornwall, D.H., Middlebrook, E.A., Hasenkrug, K.J. & Potts, W.K. 2013. Experimental viral evolution reveals major histocompatibility complex polymorphisms as the primary host factors controlling pathogen adaptation and virulence. Genes Immun. 14: 365-72. Nature Publishing Group. Kubinak, J.L., Ruff, J.S., Whitney, C., Slev, P.R. & Potts, W.K. 2012. Experimental viral evolution to specific host MHC genotypes reveals fitness and virulence trade-offs in alternative MHC types. Proc. Natl. Acad. Sci. U. S. A. 109: 3422-3427. 94 Kubinak, J.L., Stephens, W.Z., Soto, R., Petersen, C., Chiaro, T., Gogokhia, L., et al. 2015. MHC variation sculpts individualized microbial communities that control susceptibility to enteric infection. Nat. Commun. 6: 8642. Marietta, E., Rishi, A. & Taneja, V. 2015. Immunogenetic control of the intestinal microbiota. Immunology 145: 313-322. Maynard, C.L., Elson, C.O., Hatton, R.D. & Weaver, C.T. 2012. Reciprocal interactions of the intestinal microbiota and immune system. Nature 489: 231-41. Mazmanian, S.K., Liu, C.H., Tzianabos, A.O. & Kasper, D.L. 2005. An immunomodulatory molecule of symbiotic bacteria directs maturation of the host immune system. Cell 122: 107-118. Mirand, E.A. & Grace, J.T. 1963. Responses of germ-free mice to Friend virus. Nature 200: 92-93. Noverr, M.C. & Huffnagle, G.B. 2004. Does the microbiota regulate immune responses outside the gut? Trends Microbiol. 12: 562-568. Pullen, L.C., Park, S.H., Miller, S.D., Dal Canto, M.C. & Kim, B.S. 1995. Treatment with bacterial LPS renders genetically resistant C57BL/6 mice susceptible to Theiler's virus-induced demyelinating disease. J. Immunolgy 155: 4497-4503. R Core Team. 2016. R: A language and environment for statistical computing. R Foundation for Statistical Computing, Vienna, Austria. Robinson, J., Halliwell, J.A., Hayhurst, J.D., Flicek, P., Parham, P. & Marsh, S.G.E. 2015. The IPD and IMGT/HLA database: Allele variant databases. Nucleic Acids Res. 43: D423-D431. Rohmer, L., Hocquet, D. & Miller, S.I. 2011. Are pathogenic bacteria just looking for food? Metabolism and microbial pathogenesis. Trends Microbiol. 19: 341-348. Elsevier Ltd. Ruscetti, S.K. 1999. Deregulation of erythropoiesis by the Friend spleen focus-forming virus. Int. J. Biochem. Cell Biol. 31: 1089-109. Sekirov, I., Russell, S. & Antunes, L. 2010. Gut microbiota in health and disease. Physiol. Rev. 90: 859-904. Stecher, B., Robbiani, R., Walker, A.W., Westendorf, A.M., Barthel, M., Kremer, M., et al. 2007. Salmonella enterica serovar typhimurium exploits inflammation to compete with the intestinal microbiota. PLoS Biol. 5: 2177-2189. Sullivan, Å., Edlund, C. & Nord, C.E. 2001. Effect of antimicrobial agents on the 95 ecological balance of human microflora. Lancet Infect. Dis. 1: 101-114. Vrieze, A., De Groot, P.F., Kootte, R.S., Knaapen, M., Van Nood, E. & Nieuwdorp, M. 2013. Fecal transplant: A safe and sustainable clinical therapy for restoring intestinal microbial balance in human disease? Best Pract. Res. Clin. Gastroenterol. 27: 127- 137. Elsevier Ltd. Wegner, K.M., Reusch, T.B.H. & Kalbe, M. 2003. Multiple parasites are driving major histocompatibility complex polymorphism in the wild. J. Evol. Biol. 16: 224-32. Willing, B.P., Vacharaksa, A., Croxen, M., Thanachayanont, T. & Finlay, B.B. 2011. Altering host resistance to infections through microbial transplantation. PLoS One 6: e26988. Zaph, C., Du, Y., Saenz, S.A., Nair, M.G., Perrigoue, J.G., Taylor, B.C., et al. 2008. Commensal-dependent expression of IL-25 regulates the IL-23 - IL-17 axis in the intestine. J. Exp. Med. 205: 2191-2198. 96 Figure 4.1. Microbiota transfer experimental design. Microbiota stocks were harvested from either Balb/cbb or Balb/cdd hosts. The microbiota stocks were transferred into hosts with either the same host genotype (Native) or into host of a different host genotype (Novel). 97 Figure 4.2. Virulence and fitness of FVC in hosts receiving either native or novel microbiomes. (A) A significant increase in virulence was seen in novel (open symbols) over the native (filled symbols) treatment, however, there is also a significant interaction suggesting the two genotypes are responding differently. Due to the significant interaction in virulence post hoc t-tests were performed which showed the effect of the LM is driven by the Balb/cbb genotype (square symbols). (B) Proviral load showed a marginally significant increase in hosts receiving novel microbiomes. (C) Furthermore, spleen corrected titer also showed a marginally significant increase in the novel group with both groups showing a similar pattern. Round symbols represent Balb/cdd genotypes. 98 Figure 4.3. Immune cell subset proportions of total cells or total lymphocytes. (A) Total lymphocyte proportion of total cells counted, showing a reduction in proportion in novel treatment groups when compared to native treatment groups. (B) CD3+ (Tlymphocytes) proportion of total lymphocyte population. There was a marginally significant increase in CD3+ T-cells in novel treatment groups over native treatment groups. However, this increase is driven by (C) CD4+ T-cells having a significant increase in novel treatment groups. Symbols denote genotype with filled symbols indicating native treatemtn group and open symbols indicating novel treatments. Data are represented by mean ± SEM. 99 Table 4.1. Linear models comparing treatment groups (native and novel), and host genotype. Virulence (Spleen mass) of treatment groups LM (F3,16 = 125.9; R2 = 0.9594; p-value = 2.44e-11) Fixed Effects Estimate Std. Error t value Pr(>|t|) Intercept (Native Microbiota) -0.6956 0.0503 -13.823 2.59e-10 *** Host (Balb/cdd) 1.0899 0.0755 14.438 1.35e-10 *** Treatment (Novel Microbiota) 0.2218 0.0681 3.255 0.00497 * Interaction (Balb/cdd:Novel) -0.2214 0.1017 -2.177 0.04480 * F-MuLV titers (log transformed) of treatment groups. LM (F3,16 = 8.551; R2 = 0.6159; p-value = 0.00128) Fixed Effects Estimate Std. Error t value Pr(>|t|) Intercept (Native Microbiota) -2.6409 0.2593 -10.186 1.18e-08 *** Host (Balb/cdd) 1.3328 0.2999 4.444 0.00037 ** Treatment (Novel Microbiota) 0.5247 0.2999 1.750 0.0982 . Spleen Corrected F-MuLV titers (log transformed) of treatment groups. LM (F3,17 = 25.35; R2 = 0.7193; p-value < 0.0001) Fixed Effects Estimate Std. Error t value Pr(>|t|) Intercept (Native Microbiota) -3.2825 0.2910 -11.280 2.58e-09 *** Host (Balb/cdd) 2.3009 0.3366 6.836 2.90e-05 *** Treatment (Novel Microbiota) 0.6473 0.3366 1.923 0.0714 . * Indicates a P value < 0.05., ** < 0.01, *** < 0.001. 100 Table 4.2. Linear models comparing treatment groups (native and novel), and host genotype of Total immune cell subsets. Total Lymphocyte proportion by treatment groups LM (F2,17 = 4.127; R2 = 0.2476; p-value = 0.035) Fixed Effects Estimate Std. Error t value Pr(>|t|) Intercept (Native Microbiota) 0.3694 0.0476 7.754 5.57e-07 *** Host (Balb/cdd) -0.1176 0.0551 -2.135 0.047 * Treatment (Novel Microbiota) -0.1047 0.0551 -2.901 0.074 . Total Activated CD3+ proportion by treatment group LM (F2,17 = 1.93; R2 = 0.089; p-value = 0.176) Fixed Effects Estimate Std. Error t value Pr(>|t|) Intercept (Native Microbiota) 0.128 0.0240 5.341 5.4e-05 *** Host (Balb/cdd) -0.005 0.0278 -0.163 0.872 Treatment (Novel Microbiota) 0.054 0.0278 1.960 0.067 . Total Activated CD4+ proportion by treatment group LM (F2,17 = 10.59; R2 = 0.5023; p-value = 0.001) Fixed Effects Estimate Std. Error t value Pr(>|t|) Intercept (Native Microbiota) 0.086 0.0259 3.312 0.0041 ** Host (Balb/c ) 0.118 0.0301 3.928 0.0011 ** Treatment (Novel Microbiota) 0.071 0.0301 2.356 0.0307 * dd * Indicates a P value < 0.05., ** < 0.01, *** < 0.001. APPENDIX A SERIAL INFECTION OF DIVERSE HOST (MUS) GENOTYPES RAPIDLY IMPEDES FITNESS AND VIRULENCE 102 103 104 105 106 107 108 109 110 APPENDIX B SEXUAL SELECTION CONSTRAINS BODY MASS OF FEMALE BUT NOT MALE MICE 112 113 114 115 116 APPENDIX C MHC HETEROZYGOTE DISADVANTAGE DURING INFECTION OF FRIEND VIRUS COMPLEX 118 C.1 Introduction Animals heterozygous at the major histocompatibility complex (MHC) are thought to be more protected against pathogens as they have the potential to recognize twice the pathogenic peptide as homozygotes (Kubinak et al., 2012a). Furthermore, it has been shown that vertebrates increase the MHC heterozygosity of their offspring through disassortative mating preferences, suggesting heterozygosity is beneficial in natural populations (Penn & Potts, 1999; Mays & Hill, 2004). However, heterozygosity is not universally protective against pathogens, as there are MHC alleles that have dominate susceptibility profiles; this has been shown in salmonella and Theiler's virus where some heterozygote combinations are more susceptible to the pathogens than either of the homozygotes, though when both pathogens are coinfected together the heterozygote genotypes seem to be protective relative to the homozygotes (McClelland et al., 2003). Here we use Friend Virus Complex (FVC) to test whether MHC heterozygote animals are more resistant to the pathogen than MHC homozygote animals through serial passage, we would expect that if there was an advantage to being an MHC heterozygote, heterozygote animals lower virulence and viral titer in the MHC homozygote animals. Furthermore, if MHC heterozygotes are more difficult for FVC to adapt to, then we would hypothesize that homozygote passage would increase in fitness and virulence more rapidly, as has been shown previously, whereas heterozygote passages will take a higher number of passage rounds to increase to the same level of fitness and virulence as the homozygotes. 119 C.2 Methods Ten animals of each MHC genotype were infected in order to understand how virulent the virus is during the initial infection: Balb/cdd, Balb/cbb, Balb/ckk, Balb/cdk, and Balb/cbd. The second and third rounds consisted of five animals of each genotype due to constraints in animal number. FVC was serial passaged through each genotype three times to further test whether the virus has a more difficult time adapting to heterozygote animals than homozygote animals. Animals were sacrificed 12 days postinfection at which time spleen mass (virulence) was measured. After measurement, the spleens were mixed in a 1:1 ratio with 1x Phosphate Buffered Saline (PBS), homogenized and spun at 5000 RPM. Supernatants were collect and frozen at -80°C until thawed to infect the next round of animals. DNA was extracted from spleen homogenates and qPCR was done in order to determine the number of proviral insertions occurred which is a measure of viral titer (fitness). C.3 Results and Discussion After the first 12 day infection there was no heterozygote advantage seen as each of the homozygote genotypes tested had significantly reduced virulence than their heterozygote counterparts; however, heterozygotes did not show a significant decrease in viral fitness (LMM; Virulence, t = -2.42, p = 0.019; Fitness, t = -3.59, p = 0.19; Fig C1AB). Furthermore, this disadvantage continued into the second round of serial passage, with virulence still showing an increase in heterozygotes over homozygotes (LMM; Round 2, t = -2.34, p = 0.029; Fig C2A); though the effect was lost in the third round where there was no significant difference in virulence between the two treatment groups 120 (LMM; t = -1.11, p = 0.382; Fig C2B). Here we show that there is an overall disadvantage to being an MHC heterozygote when infected with FVC, though this disadvantage disappears through passage of the virus. The lack of increase through serial passage in homozygote animals has not been seen before as previously each time FVC has been serially passaged through these MHC homozygote genotypes there has been a dramatic increase in both fitness and virulence, which we do not see here (Kubinak et al., 2012b, 2013, 2015). Regardless of the lack of increase through passage there was still an initial disadvantage in heterozygote animals; however, more work needs to be completed to determine if the heterozygote disadvantage seen in the first two rounds of infection is accurate. C.4 References Kubinak, J., Nelson, A., Ruff, J. & Potts, W. 2012a. Trade-offs limiting MHC heterozygosity. In: Ecoimmunology (G. E. Demas & R. J. Nelson, eds), pp. 225- 258. Oxford University press. Kubinak, J.L., Cornwall, D.H., Hasenkrug, K.J., Adler, F.R. & Potts, W.K. 2015. Serial infection of diverse host (Mus) genotypes rapidly impedes pathogen fitness and virulence. Proc. Biol. Sci. 282. 1-9. Kubinak, J.L., Ruff, J.S., Cornwall, D.H., Middlebrook, E.A., Hasenkrug, K.J. & Potts, W.K. 2013. Experimental viral evolution reveals major histocompatibility complex polymorphisms as the primary host factors controlling pathogen adaptation and virulence. Genes Immun. 14: 365-72. Nature Publishing Group. Kubinak, J.L., Ruff, J.S., Whitney, C., Slev, P.R. & Potts, W.K. 2012b. Experimental viral evolution to specific host MHC genotypes reveals fitness and virulence tradeoffs in alternative MHC types. Proc. Natl. Acad. Sci. U. S. A. 109: 3422-3427. Mays, H.L. & Hill, G.E. 2004. Choosing mates: Good genes versus genes that are a good fit. Trends Ecol. Evol. 19: 554-559. McClelland, E.E., Penn, D.J. & Potts, W.K. 2003. Major Histocompatibility Complex heterozygote superiority during coinfection. Infect. Immun. 71: 2079-2086. 121 Penn, D.J. & Potts, W.K. 1999. The evolution of mating preferences and major histocompatibility complex genes. Am. Nat. 153: 145-164. 122 Figure C.1. Spleen mass and proviral estimates by treatment group in round 1 passages. (A) Heterozygotes had significantly higher spleen masses than homozygotes in the initial round of infection; (B) whereas there was no difference detected in viral titer. Data represented by mean ± SEM. 123 Figure C.2. Virulence estimates in the second and third round of infections. (A) The heterozygote disadvantage persisted into the second round as heterozygotes had significantly larger spleens than homozygotes. (B) In the third round there was no detectable difference in spleen mass. Data represented by mean ± SE.