Drosophila and the Rhabdovirus Sigma

MISSING IMAGE

Material Information

Title:
Drosophila and the Rhabdovirus Sigma a Model System for the Evolution of Virulence following Interspecific Host Shifts
Physical Description:
1 online resource (166 p.)
Language:
english
Creator:
Matos, Luis Fernando
Publisher:
University of Florida
Place of Publication:
Gainesville, Fla.
Publication Date:

Thesis/Dissertation Information

Degree:
Doctorate ( Ph.D.)
Degree Grantor:
University of Florida
Degree Disciplines:
Entomology and Nematology
Committee Chair:
Maruniak, James E
Committee Members:
Frank, J. Howard
Alto, Barry W
Wayne, Marta L

Subjects

Subjects / Keywords:
drosophila -- evolution -- host -- melanogaster -- rhabdovirus -- shift -- sigma -- simulans -- virulence
Entomology and Nematology -- Dissertations, Academic -- UF
Genre:
Entomology and Nematology thesis, Ph.D.
bibliography   ( marcgt )
theses   ( marcgt )
government publication (state, provincial, terriorial, dependent)   ( marcgt )
born-digital   ( sobekcm )
Electronic Thesis or Dissertation

Notes

Abstract:
Understanding the evolution of virulence continues to be an importantarea of research as we strive to prevent and cure diseases of humans andlivestock. Of particular interest are the events that occur after a pathogeninvades a novel host (host shift). Most of the currently accepted theory on theevolution of virulence following a host shift is based on data developed usingsingle celled hosts in serial passage experiments. Based on these data, it isexpected that after a host shift a pathogen will be highly virulent to itsnovel host. Studying these parameters is difficult because natural host shiftshappen in the absence of observation and because the standing genetic variationpresent in the pathogen population immediately prior to the shift is unknown.Additionally, the host shifts of interest involve complex multicellular hosts.To address these issues I developed a tractable model system using Drosophila spp. and the sigmarhabdovirus which, respectively, are good models for dipteran disease vectorsand pathogenic RNA viruses. Using this system, the genetic variation in wildsigma virus was determined using next generation SOLiDTM sequencing,a shift onto a novel host was induced and tracked in the laboratory and theresulting evolved virus was returned to its native host. The sigma rhabdovirushas lower genetic variation than is typically expected of RNA viruses and apart of the variation is due to ADAR activity. The sigma was more infective onthe novel host than on the native host and, as predicted, it was more virulenton the novel host. Contrary to prediction, I found no evidence of attenuationwhen the evolved virus was returned to its native host; the opposite was trueearly in the infection. These results indicate that the evolution of virulencefollowing a host shift is more complex than expected and indicate that this newmodel system can be used successfully to learn about the evolution ofvirulence.
General Note:
In the series University of Florida Digital Collections.
General Note:
Includes vita.
Bibliography:
Includes bibliographical references.
Source of Description:
Description based on online resource; title from PDF title page.
Source of Description:
This bibliographic record is available under the Creative Commons CC0 public domain dedication. The University of Florida Libraries, as creator of this bibliographic record, has waived all rights to it worldwide under copyright law, including all related and neighboring rights, to the extent allowed by law.
Statement of Responsibility:
by Luis Fernando Matos.
Thesis:
Thesis (Ph.D.)--University of Florida, 2012.
Local:
Adviser: Maruniak, James E.
Electronic Access:
RESTRICTED TO UF STUDENTS, STAFF, FACULTY, AND ON-CAMPUS USE UNTIL 2013-06-30

Record Information

Source Institution:
UFRGP
Rights Management:
Applicable rights reserved.
Classification:
lcc - LD1780 2012
System ID:
UFE0045024:00001


This item is only available as the following downloads:


Full Text

PAGE 1

DROSOPHILA AND T HE RH ABDOVIRUS SIGMA: A MODEL SYSTEM FOR THE EVOLUTION OF VIRULENCE FOLLOWING INTERSPECIFIC HOST SHIFTS By LUIS FERNANDO MATOS A DISSERTATION PRESENTED TO THE GRADUATE SCHOOL OF THE UNIVERSITY OF FLORIDA IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF DOCTOR OF PHILOSOPHY UNIVERSITY OF FLORIDA 2012 1

PAGE 2

2012 Luis Fernando Matos 2

PAGE 3

To my fa mily 3

PAGE 4

ACK NOWLEDGMENTS I thank the following peopl e and entities who have had a positive impact on me during my doctoral program. Dr. James Maruni ak, provided me with the opportunity of continuing my graduate studies and for that I am forever indebted to him. I thank my other committee members: Dr Howard Frank and Dr. Barry Alto for valuable comments to the research proposal and this doc ument. Dr. Marta Wayne has been simply amazing. I have benefited from her mentorship and friendship more than I can tell. She was and is an excellent example of how to be a good scientist and a good person. Finally, I thank my amazing wife because wit hout her sustaining me through the trials and triumphs of these past few years this di ssertation would not have been possible. Thank you! 4

PAGE 5

TABL E OF CONTENTS page ACKNOWLEDG MENTS..................................................................................................4LIST OF TABLES............................................................................................................8LIST OF FI GURES ........................................................................................................10LIST OF ABBR EVIATIONS...........................................................................................12ABSTRACT ...................................................................................................................13CHAPTER 1 INTRODUC TION....................................................................................................15Backgroun d.............................................................................................................15Evolution of Virulenc e.............................................................................................16Coincidental Hy pothesis...................................................................................17Short-Sighted Hy pothesis.................................................................................17Trade-Off Hy pothesis .......................................................................................18Experimental organisms.........................................................................................21Drosophila Spp.................................................................................................21Sigma Rha bdovirus ..........................................................................................23Signific ance............................................................................................................272 LOW GENETIC VARIATION BOTH WITHIN AND BETWEEN HOSTS FOR SIGMA VIRUS (RHABDOVIRIDAE) IN DROSOPHILA ..........................................32Backgroun d.............................................................................................................32Materials and Methods............................................................................................36Fly Collection and Propagatio n.........................................................................36Ribonucleic Acid (R NA) Extrac tion...................................................................36Reverse Transcription and Poly merase Chain Reacti on..................................37SOLiDTM Reads Mapping and Alignm ent.........................................................38Adenosine Deaminases Acting on RNA and Polymorphism Analysis..............38Result s....................................................................................................................39Whole-Genome S equencin g............................................................................39Adenosine Deaminases Acting on RNA (ADAR)-Driven S ubstitutions.............40Discussio n..............................................................................................................433 EVOLUTION OF VIRULENCE IN THE SIGMA VIRUS FOLLOWING A HOST SHIFT IN DRO SOPHILA........................................................................................61Backgroun d.............................................................................................................61Materials and Methods............................................................................................65 5

PAGE 6

Fly Li nes ...........................................................................................................65Inoculum and In jections....................................................................................65Rearing Conditions, Artificial Selection for Infection, and Relaxed Selection...66Fecundity and Ha tchability...............................................................................67Ribonucleic Acid (R NA) Purifi cation.................................................................68Reverse Tran scripti on......................................................................................68Quantitative Polymerase Chain R eaction .........................................................69Statistical Analyse s..........................................................................................69Result s....................................................................................................................70Artificial Host-Shift Success and Selection for Transmissi on............................70Infection under Rela xed Select ion....................................................................71Virus Ti ters.......................................................................................................72Host Fit ness.....................................................................................................73Discussio n..............................................................................................................74Artificial Host-Shift and Se lection for Tr ansmissi on..........................................74Transmission under Rela xed Select ion............................................................76Virus Ti ters.......................................................................................................77Virulenc e..........................................................................................................774 SEQUENCING AND ANALYSIS OF VIRAL GENOMES PASSAGED IN DROSOPHILA LI NEAGES.....................................................................................91Backgroun d.............................................................................................................91Materials and Methods............................................................................................92Ribonucleic Acid (R NA) Purifi cation.................................................................92Reverse Tran scripti on......................................................................................93Viral Genome Sequenci ng................................................................................93Statisti cs...........................................................................................................94Result s....................................................................................................................94Discussio n..............................................................................................................965 TESTING ATTENUATION BY CROSS-INFECTING NAVE NATIVE AND NOVEL HOSTS WITH EXPERIMEN TALLY-EVOLVED VIRUS...........................122Backgroun d...........................................................................................................122Materials and Methods..........................................................................................123Nave Host s....................................................................................................123Inocula and Injections .....................................................................................124Rearing Conditions and Artificial Selection for Infection.................................125Fecundity and Hatchability.............................................................................125Ribonucleic Acid (RNA ) Purifica tion...............................................................126Reverse Transcr iption ....................................................................................126Statisti cs.........................................................................................................127Results ..................................................................................................................127Cross-Infection Success and Selection for Transmission...............................127Components of Host Fitness as a Proxy for Vi rulence ...................................129Fecundity as a Proxy for Virul ence.................................................................129 6

PAGE 7

Hatchabilit y as a Proxy for Virul ence..............................................................131Discussio n............................................................................................................1326 CONCLUSION S...................................................................................................142Summary of Findings ............................................................................................142Adenosine-Dependent RNA Deaminases.............................................................143Transmissibility and Virulence Following a Host Shift...........................................144Sequencing Evolv ed Virus ....................................................................................146Testing Att enuation ...............................................................................................146APPENDIX A MOLECULAR BIOLOGY PROTOCOL S...............................................................149RNA Isolation from One Fl y..................................................................................149Reverse Transcr iption ...........................................................................................150The Polymerase Chain Reacti on..........................................................................151Quantitative Polymerase Chain Reaction.............................................................151B FLY HANDLING PROTOCOL S............................................................................153Drosophila Rearing...............................................................................................153Artificial Injection of Female Drosophila ................................................................153Inoculum Pr oduction .......................................................................................154Artificial In jection............................................................................................154Fecundity and Hatchability....................................................................................155LIST OF REFE RENCES.............................................................................................157BIOGRAPHICAL SKETCH ..........................................................................................166 7

PAGE 8

LIST OF TABLES Table page 2-1 Descriptive statistics parameters char acterizing the number of reads per site at 6433 sites for each virus samp le....................................................................45 2-2 Independent nearest neigh bor preferences for adenos ine deaminases acting on RNA (ADAR). Based on Leh mann and Bass ( 2000)*....................................46 2-3 Adenosine deaminases acting on RNA (A DAR) triplet preferences. Based on Lehmann and Bass (2000).................................................................................46 2-4 Mutations at 1747 sites harboring c onsensus A nucleotide within proteincoding regions, excluding the 32 site s that harbor high-frequency single nucleotide polym orphism s..................................................................................47 2-5 Adenosine deaminases acting on ribonucleic acid (ADAR) preferentially changed strong ADAR sties................................................................................48 2-6 Descriptive statistics parameters char acterizing the number of reads per site at protein-coding sites in the 1st codon position for each sample, separated by whether the site is Aharboring [and further subdivid ed into weak or strong adenosine deaminases acting on RNA (A DAR preference] or not.....................49 2-7 Descriptive statistics parameters char acterizing the number of reads per site at protein-coding sites in the 2nd c odon position for each sample, separated by whether the site is Aharboring [and further subdivi ded into weak or strong adenosine deaminases acting on RNA (A DAR) preference] or not....................50 2-8 Descriptive statistics parameters char acterizing the number of reads per site at protein-coding sites in the 3rd codon position for each sample, separated by whether the site is Aharboring (and further subdivided into weak or strong ADAR preference) or not....................................................................................51 2-9 P-values from Fishers exact test and mean numbers of different classes of mutations per site [adenosine deami nases acting on RNA (ADAR)-driven A to G hypermutations and non-ADAR-driven A to T or A to C mutations] at strong and weak AD AR site s..............................................................................52 2-10 Summary of non-synonymous substitutions (different from the consensus) that occur in each sample, organized by gene, codon number and base position within each codon (codon posit ion).......................................................53 3-1 Sigma virus titers (mean standard error) in females of Drosophila melanogaster and D. simulans at six and 16 generations after artificial infection by injecti on...........................................................................................79 8

PAGE 9

3-2 Mean sigma virus titers across spec ies, line and generati on (data presented in Figure 3-4)......................................................................................................79 3-3 Fecundity data (eggs/female/day) for sigma virus-infe cted and uninfected Drosophila melanogaster (native host) and D. simulans (novel host)................80 3-4 Hatchability data (% of eggs hatch ing) for sigma virus-infected and uninfected Drosophila melanogaster (native host) and D. simulans (novel host)...................................................................................................................81 4-1 Sequences for 19 primer pairs using to produce overlapping amplicons from viral complementary DNA (BRUSINI et al. 2012 in re view)...................................99 4-2 Sigma virus strains evolved in Drosophila simulans (DSimSV-E) or D. melanogaster (DMelSV-E) were compared to the ancestral virus (DMelSVA)......................................................................................................................100 4-3 Summary of synonymous and non-synonymous substitutions in ancestral virus evolved in Drosophila melanogaster (DMelSV-A) and in strains evolved in D. simulans (DSimSV-E) and D. melanogaster ............................................111 4-4 Summary of statistical analysis compar ing each of the six genes from twelve viral variants. One was the ancestral virus and eleven were passaged in flies for sixteen host generations ..............................................................................113 4-5 Consensus sequence for ancestral virus collected from wild-caught Drosophila melanogaster ..................................................................................114 5-1 The virus lines used for cross infection inoculum production exhibited high virus tite rs.........................................................................................................135 5-2 Confidence intervals for per cent infection in Drosophila simulans (DSim) and D. melanogaster (DMel) adults infected with ei ther evolved (DSimSV-E) or ancestral (DMelSVA) virus..............................................................................135 5-3 Confidence intervals for fecundity standardized to the uninfected control mean in Drosophila simulans (DSim) and D. melanogaster (DMel) females infected with either evolved (DSimSV-E) or ancestral (DMelSV-A) virus..........136 5-4 Confidence intervals for hatchability standardized to the uninfected control mean in Drosophila simulans (DSim) and D. melanogaster (DMel) females infected with either evolved (DSimSV-E) or ancestral (DMelSV-A) virus..........136 9

PAGE 10

LIST OF FIGURES Figure page 2-1 Correlation of read number for our te chnical replicate indicates that the two samples behaved similarly in the sequencing process ( P < 0.001 and R = 0.921). ................................................................................................................552-2 Shown are two possible transitions ( ) and transversions () mutation events (that together form a s ubstitution ma trix)............................................................562-3 Flow chart describing how the 1747 candidate sites for adenosine deaminases acting on RNA (ADAR) activity were identifie d...............................572-4 Minority reads were a minor component of our total read count. The number of minority nucleotide r eads accounted for < 5% of the reads on any one base position al ong the genom e.........................................................................582-5 A-G site changes occurred disproportionally more often in a ll flies indicating adenosine deaminases acting on RNA (ADAR) ac tivity.....................................592-6 Total number of substitutions (that occur with frequencies of at least 5% in one of the samples) and type of substitution in each of t he flies te sted..............603-1 The novel host was easier to in fect, but harder to stabi lize................................823-2 Maximum transmission efficiency (100%) was harder to achieve on the novel host.....................................................................................................................833-5 Heterogeneity in titer over ti me between lines fo r virus load. .............................863-6 Virus titer is significantly corre lated with transmission efficiency for both hosts, but the sign of the co rrelation changes with time .....................................873-7 Fecundity was not negatively impacted by infection with sigma virus.................883-8 Hatchability was negatively impacted by infection with sigma virus in both hosts but at differ ent generati ons.......................................................................893-9 Hatchability standardized to the relev ant uninfected mean for fly lines with high (large symbols) and low (sma ll symbols) virus titers in Drosophila melanogaster and D. simulans ...........................................................................904-1 Maximum likelihood phyl ogenetic relationship among ancestral sigma virus from Drosophila melanogaster (DMelSV-A) and sigma virus evolved in naive native hosts (DMelSV-E) and in novel D. simulans host (DSimSV-E) lineages.. ..........................................................................................................121 10

PAGE 11

5-1 Infectivity of virus passaged in the novel host, Drosophila simulans (DSimSVE), was not attenuated to the native host, D melanogaster .............................1375-2 Transmission efficiency of virus evolved in Drosophila simulans (DSimSV-E) was not attenuated to the native host, D melanogaster ...................................1385-3 The ability to produce increasing infection levels in successive generations was not attenuated when virus evolved in Drosophila simulans (DSimSV-E) and re-introduced to its native host, D melanogaster when compared to the performance of the unevolved ancestral virus (DMelSV-A ) in the native host..1395-4 Contrary to prediction, the virus passaged in Drosophila simulans (DSimSVE, white circles) appears to be mo re virulent in the native host ( D melanogaster ) than the ancestral virus (DMe lSV-A, black ci rcles)...................1405-5 Hatchability was indistinguishable for Drosophila melanogaster receiving unevolved ancestral virus (black circles) or virus evolved in D. simulans (white circ les)....................................................................................................141 11

PAGE 12

LIST OF ABBREVIAT IONS DMELSV The wild type Drosophila melanogaster sigma rhabdovirus purified from the offspring of wild-caught flies DMELSV-A Ancestral sigma virus evolved in Drosophila melanogaster that was injected into nave native and novel host DSIMSV-E Sigma rhabdovirus that was evolved in D. simulans ADAR Adenosine deaminases acting on RNA PCR Polymerase chain reaction 12

PAGE 13

Abstract of Dissertation Pr esented to the Graduate School of the University of Fl orida in Partial Fulf illment of the Requirements for t he Degree of Doctor of Philosophy DROSOPHILA AND THE RH ABDOVIRUS SIGMA: A MODEL SYSTEM FOR THE EVOLUTION OF VIRULENCE FOLLOWI NG INTERSPECIFIC HOST-SHIFTS By Luis Matos December 2012 Chair: James Maruniak Major: Entomology and Nematology Understanding the evolution of virulence continues to be an important area of research as we strive to prevent and cu re diseases of humans and livestock. Of particular interest are the ev ents that occur after a pathogen invades a novel host (host shift). Most of the currently accepted theory on the evoluti on of virulence following a host shift is based on data developed using single celled hosts in serial passage experiments. Based on these data, it is ex pected that after a host shift a pathogen will be highly virulent to its novel host. Studying these parameters is difficult because natural host shifts happen in the absence of observation and because the standing genetic variation present in the pathogen populat ion immediately prior to the shift is unknown. Additionally, the host shifts of inte rest involve complex multicellular hosts. To address these issues I developed a tractable model system using Drosophila spp. and the sigma rhabdovirus which, respective ly, are good models for dipteran disease vectors and pathogenic RNA viruses. Using th is system, the genetic variation in wild sigma virus was determined us ing next generation SOLiDTM sequencing, a shift onto a novel host was induced and tracked in the laboratory and the resulting evolved virus was returned to its native host. The sigma rhabdovirus has lower genetic variation than 13

PAGE 14

14 is typically expected of RNA viruses and a part of the variation is due to ADAR activity. The sigma was more infective on the nov el host than on the native host and, as predicted, it was more virulent on the nov el host. Contrary to prediction, I found no evidence of attenuation when the evolved vi rus was returned to its native host; the opposite was true early in the infection. Thes e results indicate that the evolution of virulence following a host shift is more complex than expected and indicate that this new model system can be used successfully to l earn about the evolution of virulence.

PAGE 15

CHA PTER 1 INTRODUCTION Background Virulence, simply defined, is a pathogens ability to cause disease in its host ( KNIPE and HOWLEY 2001 ; LITTLE et al. 2010 ). Avirulent or mildly virulent organisms cause no or little disease, respectively, and can persist in the population through many host generations with zero or minima l advers e affects to the host ( BEACH et al. 2009 ). Conversely highly virulent organisms us ually damage their hosts severely, thus inflicting fitness losses and even death of the host ( KNIPE and HOWLEY 2001 ). The latter situation can occur when a pathogen host-shifts from its native host onto a novel host species. The human immunodef iciency virus (HIV) ( RAMBAUT et al. 2004 ) and severe acute respiratory syndrome (SARS) ( MULLER and MCGEER 2007 ) are just two examples of pathogens shifting hosts from non-human species into humans. In recent years several other pathogens have shifted into humans hosts, leading to severe disease in some cases ( FRASER et al. 2009 ; MULLER and MCGEER 2007 ; TSUKAMOTO et al. 2007 ; VAN DER MEULEN et al. 2005 ). Unfortunately, predictions of how severely these pathogens would affect human populat ions were not always correct. The recent case of swine flu (H1N1 influenza) is a prime exampl e of how inaccurate these predictions can be. The predicted deadly effect of t he emergi ng pathogen on the human population was, thankfully, grossly overestimated as the numbers of influenza cases were similar to previous years ( COLLIGNON 2010 ; ISAACS 2010 ; KELLY 2010 ). This error led to increased anxiety among the worlds popula tion and produced ser ious financial losses to industry and governments worldwide as t hey prepared to combat an epidemic that materialized far below the predicted case levels ( FLYNN 2010 ). Because of the poor 15

PAGE 16

predictions, vaccines were overproduced and thus wasted. Holland alone was left with nearly 20 million unused doses ( DOLGIN 2010 ). To prevent these issues we need a better understanding of the evolution of virulenc e that occurs following an interspecific host shift. This knowledge will assist the scientific community as it predicts the epidemiolo gy of emerging pathogens. Here I will review the literature on 1) the evolution of virulence, 2) introduce the model syst em, 3) discuss the background and significance of the project, 4) and describe the exper imental design that will be used to test hypotheses about the evolutio n of virulence after an interspecific host shift. Evolution of Virulence Virulence is an old concept that originated from Latin virulentus and the first recorded use of virulent occurred in t he 1400s A.D to describe something bad or poisonous (Council of Sci. Eds., 2011). This definition persisted for a long time and was even used by Charles Darwin (The virulence of this poison) to describe the potency of a spiders venom ( DARWIN 1909-14 ). More recently the te rm virulence has been more directly associated wit h the level of disease that a pathogen is capabl e of inflicting on its host ( KNIPE and HOWLEY 2001 ; LITTLE et al. 2010 ). Thus, modern studies on the evolution of virulence seek to understand what factors contribute to changes (increases or decreases) in a pathogens virulence level. To this end, three major hypotheses have been formulated to explain the evolution of vi rulence. These hypotheses argue that the evolution of virulence is either coincidental ( LEVIN and EDEN 1990 ), the result of shortsightednes s on the part of the pathogen (LEVIN and BULL 1994 ) or a tradeoff ( ANDERSON and MAY 1982 ; EWALD 1983 ). 16

PAGE 17

Coinciden tal Hypothesis The coincidental hypothesis of virulenc e evolution postulates that a factors contribution to a pathogens level of virulenc e within a particular host evolved for some reason other than effecting virulence in that host ( LEVIN 1996 ; LEVIN and EDEN 1990 ). Levin (1996) uses Gould and Lewontins ( 1979) beautiful analogy of the spandrels of San Marco to illustrate this point. That is that the spandrels were no t constructed to frame the frescoes, though t hey do frame them wonderfully Instead, the spandrels are a necessary structural feature without which the church would not stand ( GOULD and LEWONTIN 1979 ). Therefore, the la ck of the spandrels woul d negate the successful construction of the church and would eliminate the need for fr escoes. In the same way, this hypothesis argues that virulence is a byproduct of some other necessary process that occurs while the pathogen colonize s the host. For example, the neurotoxins produced by Clostridium tetani and C. botulinum both soil bacteria, did not evolve for the purpose of being highly toxi c in humans, yet they are. Short-Sighted Hypothesis The short-sighted hypothesis for the evolution of virulence proposes that changes in pathogen virulence at the population level are a result of random mutation followed by selection occurring at the level of the individual ( LEVIN and BULL 1994 ). To support their hypothesis Levin and Bull exami ne three diseases [bacterial meningitis, poliomyelitis, and the acquired immune defici ency syndrome (AID S)] and use facts from the ecology of the respective disease-causing agents to support this hypothesis. The authors argue that the ev olutionary strategy for these three pathogens is short-sighted because the within-host dynamics do not im prove each pathogens ability to infect and successfully exploit the next host ( LEVIN and BULL 1994 ). In particular, the authors cite 17

PAGE 18

the case of polio where the pathogen infects the neurons of the hos t despite the fact that this does not increase transmission. Thus, the pathogens evolution is shortsighted by favoring short term intra-host success at the cost of long term inter-host transmission. In fact, this short-sightedness may yield a pathogen population that is well adapted wit hin a particular individual host t hat is potentially maladapted to invade another host when or before the present host expires. Interestingly, the only empirical evidence for the short-sighted hypothesis was reported in honeybee colonies infected with a social parasite ( MORITZ et al. 2008 ). Moritz et al. (2008) showed that, in fact, the parasitic bee species became extremely well adapted wit hin a single colony (host) at the cost of being able to infe ct other colonies (hosts). Trade-Off Hypothesis The trade-off hypothesis was deve loped in the early 1980s (ANDERSON and MAY 1982 ; EWALD 1983 ) and holds that for a parasite to in crease its ability to persist on a host it must give up something, i.e. a reducti on in virulence. Howeve r, for the parasite to succeed under the assumptions of this hy pothesis, the parasite must retain some minimal lev el of virulence and transmissibility ( ANDERSON and MAY 1982 ; EWALD 1983 ). Additionally this theory argues that the hosts natural death rate, its death rate due to infection, and the rate of recovery from infection are inextr icably linked. Because of this connection, the tradeoff for a decrease in vi rulence is a concomitant decrease in transmission such that as virulence appr oaches zero so too does transmission ( ALIZON et al. 2009 ). Therefore, under this hypothesis t he parasite cannot ever become avirulent or it will cease to be transmitted to new hosts and will go extinct. This flies in the face of the previously (pre-Anderson and Ma y, 1983) commonl y accepted avirulence hypothesis, which stated as a parasite and host co-evolve the parasite will become 18

PAGE 19

progressiv ely less virulent until it is completely avirulent ( ALIZON et al. 2009 ). Therefore, the development of the trade-of f hypothesis was a major turnin g point in the study of the evolution of virulence because it cemented the idea that the interaction between the host and the parasite is a fluid one that is a ffected not only by the parasite but also by the host parameters. This hypot hesis is appealing from an epidemiology management standpoint because it suggests that some measure of control might be had over parasite virulence by managing the parameter s that contribute to parasite virulence ( ALIZON et al. 2009 ). The idea here is that i f a host/pathogen system is managed correctly, one might be able to manipulate a given parameter such that the parasite woul d trade-off high virulence for some other parameter, say tr ansmissibility. In a human dis ease outbreak, for example, this would lead to many people (a large percentage of the population) being infected with a weekly virulent pathogen (few or no deaths) rather than fewer people (a small percentage of the popul ation) infected with a highly vi rulent pathogen (many deaths). The trade-off hypothesis proposes that a del icate balance exists between virulence, transmission, and host recovery ( ALIZON 2008 ; ANDERSON and MAY 1982 ). Exactly how the trade-offs in each parameter w ork to regulate the evolution of virulence is not yet fully understood and more research is needed to produce more data on tradeoff curves ( ALIZON et al. 2009 ; BOLKER et al. 2010 ). Alizon (2008) argues that when discussi ng trade-offs and parasite evolution the effects of host recovery-transmission tradeoff must be taken into a ccount in addition to or even above virulence-transmi ssion trade-offs. This is because the dynamics of the virus population within the host will affect host recovery and, in turn, produce a 19

PAGE 20

paralleling change in the rate at which the parasite is tr ansmitted from the recovering host to a nave host ( ALIZON 2008 ). This change in transmission rate is of particular importance in human diseases where either vaccination or antimicrobial drugs alter within-host virus dynamics and typically incr ease both the rapidity with whic h hosts recover and the total number of hosts recoveri ng. Additionally, it is possible that host recovery that is aided by vaccines and antimic robials could actually drive the evolution of more virulent strains capable of overcoming these countermeasures ( ANDRE et al. 2006 ; GENTON et al. 2002 ; MACKINNON et al. 2008 ). In the case of ma laria, for example, a trial vaccine selected for alternative alle les that were not targeted by the vaccine ( GENTON et al. 2002 ). Whether theses alternate alleles were more virulent is not known; however, children who were previously vacci nated against one allele were more likely to become infected with the alternate allele ( GENTON et al. 2002 ). This indicates that although v accination may reduce recovery time by altering the immune status of the host, this gain in recovery is traded-off wit h a concomitant increas e in transmission of the alternate allele in the population. Therefore, alt hough the example with malaria supports Alizons (2008) argument that reco very is an important parameter, it also clearly shows that unlike his model, recovery is only another piece in the puzzle that is the trade-off hypothesis of virulence evolution. To a certain extent, the three hypotheses that have been proposed to explain the evol ution of virulence are not mutually exclusive. Additionally, no one hypothesis encompasses all of the diverse biological systems that have been studied to date. However, most of the cases where we see the evolution of virulence seem to follow the tradeoff hypothesis. 20

PAGE 21

Experimental organis ms Understanding the evolution of virulence often require s the use of tractable model systems. Currently, one model system has been developed for the study of hostshifts by pathogens. The fungal causat ive agent of anther-smut disease ( Microbotryum violaceum ) was shown to host-shift from its native host Silene alba (white campion) to Silene vulgaris (bladder campion) and the ecology and genetics of this shift have been examined ( ANTONOVICS et al. 2002 ). This plant-fungus system, however, is not of immediate relevance to human diseases as host-shifts from a plant to an animal vector likely are extremely rare. Here w e propose to develop Drosophila melanogaster D. simulans and the rhabdovirus sigma (DMelSV, endemic in D. melanogaster ) as a model system for the evolution of virulence and host-shifting. Drosophila Spp. Insect study systems, particu larly dipteran systems, are pr ime candidates for use as model systems in both the evolution of virulence and the mechanisms of host-shifting. There are several reasons for this: 1) many emerging diseases have dipteran vectors and 2) insects offer the possibi lity of using large numbers of individuals per experiment without requiring the cost and specialized facilities associated with maintaining vertebrates ( SCULLY and BIDOCHKA 2006 ). In addition to these main reasons, using Drosophila spp. as hosts offers se veral unique f eatures that make it a good system in which to study virus host-shifts. Drosophila is a dipteran and dipterans such as mosquitoes, sand flies, and blackflies are all vectors of disease-causing RNA viruses ( HOGENHOUT et al. 2003 ). Using the Drosophila as the host in a model system to study the e volution of host-shifts eliminates the need to work and special fac ilities necessary to keep pathogen-carrying 21

PAGE 22

insects. However, dipteran vectors are extremely similar to Drosophila spp. in morphology, physiology, and genet ics. The vectors for which Drosophila is a good model include mosquitoes, which transmit m any viral pathogens (Dengue fever, all of the encephalitis agents, Yellow fever, and many others) and sand, tse tse or black flies, which transmit disease-causing protozoans and worms (Leishmaniasis, sleepin g sickness, and river blindness ( SCHNEIDER 2000 ). Further, although many genetic tools have been developed in non-model organis m like mosquitoes (deletions and nonrecombining chromosomes), these are st ill no match for the tools available in Drosophila ( SCHNEIDER 2000 ). Establishing Drosophila as the host in this model system for the evolution of virulence likely will, as before ( ROBERTS 2006 ), spur on and facilitate work in other organisms. Additionally, it is a tractable a nd widely used system that will allow current users, should they choose, to expand their resear ch into this exciting field with minimal effort. This could allow the evol ution of virulence to be studied by a larger number of people to produce more new dat a to inform theory on this subject. Finding a good model system for host-shifts (one where the two hosts are closely related, but where the hostshift has not actually happen ed yet) is challenging; the D. melanogaster / D. simulans/DMelSV system provides such an opportunity Drosophila simulans can be reared easily in the lab under the same conditions as D. melanogaster and has the same short generation time and wo rldwide distribution as its better known sibling. Additionally, genomic re sources are now available for D. simulans with 6+ genomes completed ( www.dpgp.org ), and tiling and expression microarrays are under development (McIntyre, Wayne, and Nuzh din per. comm.), opening the door to host genomics as well as viral genomics in the future. Another benefit of the Drosophila host 22

PAGE 23

is that D. m elanogaster and D. simulans do not hybridize successfully despite their close phylogenetic relationship. Hybrids have only been possible in the lab and, even then, they are sterile ( STURTEVANT 1920 ) which explains why we dont find hybrids in the wild. Finally, D. m elanogaster is infected by a rhabdovirus (DMelSV). DMelSV is endemic to wild D. melanogaster populations and is not found in D. simulans, the sibling species to D. melanogaster even though the two species are sympatric over most of their range ( FLEURIET 1982 ; FLEURIET 1988 ). Therefore, this system is ideal because it allo ws us to artificially shift the virus at will. Sigma Rhabdovirus The sigma rhabdovirus is an excellent pathogen to be used in a system modeling the evolution of virulence because it infects D. melanogaster thus affording us all of the benefits of that model organism. Sigma is a member of the Mononegavirales order. This virus order contains the causative agent s of human diseases such as rabies, hemorrhagic septicemia, hematopoietic nec rosis and several other economically important diseases of livestock ( HOGENHOUT et al. 2003 ). There are many other RNA viruses of human interest with dipteran vector s, some of which already infect humans and others which do not yet infect us: La Cr osse encephalitis, Rift Valley fever virus (Bunyaviridae), Chikungunya virus O'nyong'nyong virus, Eastern, Western, and Venezuelan equine encephalitis, Ross River virus, Semliki Forest virus (Togaviridae), Yellow fever virus, and Japanes e encephalitis (Flaviviridae) ( FAUQUET CM 2005 ). We will use the D. m elanogaster / D. simulans /Sigma system to model the evolution of virulence as the RNA virus host-shifts between dipteran hosts. This virus has a negative single-stranded RNA genome and belongs to the family Rhabdoviridae. As with many me mbers of this family, the DMelSV genome includes six 23

PAGE 24

genes in the order 3-N-P PP3-M-G-L-5 ( CARPENTER 2008 ; TENINGES et al. 1993 ). These genes code for the nucleocapsid protei n (N), the polymerase-associate d protein (P), the matrix protein (M), the glycoprotein (G), the polymerase (L) and a reverse transcriptase (PP3) ( CARPENTER 2008 ; TENINGES et al. 1993 ). Interestingly, the PP3 gene is found in only a few other Rhabdovir us species and shares greater homology with retroviruses and retrotransposons, suggesting that DMelSV ac quired this gene through recombination with one of these sources ( CARPENTER 2008 ). Further examination of the viral genome revealed that virus populations are genetically similar. Ca rpenter et al. (2007) reported very low genetic diversity for virus isolates from Europe and North America. The authors suggest that this is the case because eit her Drosophila melanogaster acquired the virus within the last 200 years or a single virus variant swept through the host populations ( CARPENTER et al. 2007 ). The latter is likely as this is a vertically transmitted and the inherit ance frequency is higher than would be predicted by st rict Mendelian inheritance ( VODOVAR et al. 2004 ). Following the sweep the virus has subsequently accumulated mutations that account for the minor differences among the widely distributed populations. Some of these new mutations in the DMelSV genome lik ely arose by adenosine deaminases that act on RNA (ADAR). ADAR s are enzymes that target dsRNA and change adenine (A) bases to Inosine (I). This change affects translation when I is read as guanine (G) and thus paired with anti-codons that possess a cytosine (C) instead of a thymine (T) at that particular position ( MOERDYK-SCHAUWECKER et al. 2009 ). Additionally the I bases can pair with C bas es during RNA dependent RNA replication. Subsequently, the C bases pair with G bases resulting in the introduction of a mutation 24

PAGE 25

(A to G) in the genomes of RNA viruses ( MOERDYK-SCHAUWECKER et al. 2009 ). Carpenter et al (2009) sequenced ~6kb from viral genomes from wild flies and flies kept in the lab for 10-20 years. The number of singleton mutations was determined and an analys is of the distances between the singletons was done. In DMelSV, more candidate bases (A without upstream G) were mutated to Gs than would be expected by random change alone ( CARPENTER et al. 2009 ). Additionally, these mutations occur clustered in specific regions of the genome ( MOERDYK-SCHAUWECKER et al. 2009 ). Clustering of A-G mutations is another indication of ADAR acti vity. These data confirm that ADAR activity is occurring in the DMelSV genome and is, at least in part, responsible for the mutations accumulated after the recent sweep ( CARPENTER et al. 2009 ; CARPENTER et al. 2007 ). The sigma rhabdovirus is vertically trans mitted transovarially from mother to offspring and produces CO2 sensitivity (exposed flies are permanently paralyzed) in infected individuals ( FLEURIET 1988 ; L'HERITIER 1958 ). Infections occurring early in the insects development lead to infection of the germ cells and consequently a stable Sigma infection in D. melanogas ter individuals of both sexes ( BRUN and PLUS 1980 ). Whereas stabilized males produc e non-stabiliz ed offspring at modest frequencies, nearly 100% of offspring from stabilized females and over 80% of offspring from nonstabilized females are infected ( FLEURIET 1988 ). Sigma is widespread in the flys body, without any apparent tissue preference by the virus in both naturally and artificially infected flies ( BRUN and PLUS 1980 ). Flies can be artificially infected by injection with infected hemolymph in saline to produce a stable infection and CO2 sensitivity ( L'HERITIER and HUGON DE SCOEUX 1947 ). The mean incubation time necessary for express ion of CO2 sensitivity is inversely related to the number of infectious units (I.U.) 25

PAGE 26

deliv ered by injection and ranges from 5-20 days (1000 to <10 I.U.) ( BRUN and PLUS 1980 ). The ease of artificial infection (b y injection) coupled with stable vertical transmission post-injection and Sigmas natural virulenc e to D. melanogaster are additional traits that favor using this system for host-shift studies. Finally, DMelSV is virulent in D. melanogaster. A vertically transmitted pathogen can only survive in the host population when its virulence is low and vertical transmission is high; typically horiz ontal transmission is also high ( LIPSITCH et al. 1995 ). Howev er, it is unlikely that Sigma is acquir ed horizontally. There is no evidence that the CO2 sensitivity imparted to infected adult f lies by Sigma is relevant in natural populations because adult flies likely never exper ience high concentrations of the gas in the field. Infected flies show decreased egg to adult survival ( FLEURIET 1981a ), reduced female ferti lity, and decreased overwintering ability in both sexes ( FLEURIET 1981b ). Moreover, the virulenc e of DMelSV was em pirically demonstrated by an increase in infection frequency under relaxed selection pressure followed by a decrease in infection frequency after selection pressure was restored ( YAMPOLSKY et al. 1999 ). These data follow the ecological theory which predicts t hat incidenc e of infection will decrease in a system where the pathogen is uni parentally transmitted and decr eases the fitness of its host ( FINE 1975 ). This being the case, natural selection against the infected host will cause the parasites frequency to decline ( FINE 1975 ). DMelSV fits these predictions because it is only effectively transmitted vert i cally by the female [stabilized males cannot sire stabilized offspri ng on un-stabilized females ( FLEURIET 1988 ) and according to Yampolsky et al. ( 1999 ) DMelSV infection declines to zero when the flies are under strong selection pressure. Theref ore, DMelSV is virulent in D. m elanogaster and we 26

PAGE 27

exp ect that DMelSV will be virulent in D. simulans because the two species are genetically and, perhaps more important, mo rphologically and physiologically very similar. This and the other charac teristics mentioned above make the D. melanogaster / D. simulans/DMelSV system a good model in which to study host-shift evolution. The D. melanogaster / D. simulans /DMelSV model system will allow us to study several of the important traits of virus host-shifts which can facilitate the emergence of diseases, namely: how do virus fitness and virulence change between the native and novel host, and what role do virus and host evolution play in successful colonization of a novel host. Significance The genetics of the evolution of virulence remain an active area of investigation, particularly with respect to vi ral pathogens. Here we examine the evolution of virulence of the Drosophila melanogaster rhabdovirus virus (DMelSV) a fter it was artificially shifted onto a novel host, Drosophila simulans A stable host-shift is at least a two stage process: 1) an initial stage in which the novel host receives the parasite, and 2) a subsequent stage during which the parasite adapt s to the new host so that it can be maintained in the new hosts population. Th is latter stage is relatively understudied ( WOLFE et al. 2007 ), yet it is key to understandi ng the emergence of those most devastating diseases such as HIV that are passed efficiently from human to human ( MAY et al. 2001 ). These diseases are able to persist on novel host because they adapt. We define an adaptation as an individual nucleotide substitution that increases fitness. Adaptations are expected to be rare events for two reasons: first, because most mutations are unconditionally deleterious; and second, because most organisms are thought to be at equilibriu m with their environment, i.e. already well adapted, thus 27

PAGE 28

limiting the opportunity for fu rther adaptation. Open questions about adaptation include: 1) the frequency with which adap tations arise; 2) whether adaptations tend to come from extant genetic variation in the population, or from new mu tations; 3) the effect size of adaptive substitutions, and whether this decr eases over the course of adaptation to a new environment; and 4) how frequently organi sms tend to use the same solution to a given environmental challenge, i.e. convergent and/or parallel adaptation, and whether or not this evolutionary repeatability is dependent on the complexity of the genome ( ORR 2005 ). The genetics of adaptation are especially intriguing in hostparasite systems. Abundant empirical evidence dem onstrates t hat most new mutations are deleterious, rather than adaptive. However, due to the compactness of the genome of RNA viruses like DMelSV (~12kb), their la rge population sizes, and their extraordinarily high mutation rate, it is possible that a relatively high fr action of mutations which begin as deleterious experience fitness reversals during their lif espan in the population due to compensatory mutation at other sites, such that they eventually fix as adaptations ( COWPERTHWAITE et al. 2006 ). This casts into question whether most adaptations in RNA viruses are conferred by existing genetic vari ation, or by new mutations. Much of what we know about adaptation in response to host-shifts comes from serial pass age experiments (SPEs), where infe ction is caused by the experimenter ( i.e. by injection), and typically no initial genetic va riation is present. In such cases, clearly adaptation is conferred by new mutations. However, it is unclear whether this would be the case were genetic variation present. Ac cordingly, our system will evaluate the 28

PAGE 29

relative contributions of standing variation versus new variation to adaptation by including m ultiple, naturally occurring genetic vari ants in the original inoculum of virus. One of the major conclusions from SPE work is that virulence tends to increase with successive passages. This has important imp lications for emerging disease, and is often interpreted to mean that virulence w ill increase in a new host. This leads to two questions. First, will adaptation in a new host behave similarly to SPE, such that virulence actually will increase? And second, does adaptation to a new host actually require an increase in virulence? One possibility is that due to co-infection by multiple genetic variants, competition for growth rate occurs in the new host, t hus selecting for adaptations conferring rapid replication and thus rapid cons umption of host resources. This theory implicitly assumes that growth rate and virulence are, if not interchangeable, tightly correlated. However, virulence may arise from host responses such as stress instead of, or in addition to, consumption of resources by replicating viru ses. The pattern of change in virulence and its relationship to viral load has rare ly been studied outside of the SPE context. Additionally, virulence with SPEs may increase bec ause it is not costly: viral survival is guaranteed by the experiment er; therefore it is selectively irrelevant if viral precedence comes at such a cost to the host as to preclude transmissibility. Thus, increase in virulence may be an artifact of the SPE protocol ( EBERT 1998 ). A second major conclusion from SPE studie s of the evolution of virulence is that increasing virulence in the new host is accompanied by attenuation to the original host. Attenuation is defined as both a decrease in virulence and a decreased ability to replicate and/or infect the ancestral host. Why should the new adaptations come at the 29

PAGE 30

cost of the old ones? While thes e dat a have been interpreted in light of the generalist/specialist hypothesis, a more general answer to this question is loss of selective constraint: when the virus encoun tered the ancestral host repeatedly, fitness recovery on the ancestor was quick, more consistent with increases in frequency of alleles still present in the population than with back-mutati on. Though budget constrains us from exploring this particular wrinkle in virulence evolution, we will assess the generality of attenuation by introducing the D. simulans evolved virus back into D. melanogaster after 16 fly generations ( >> 16 viral generations). Another question of great interest is whet her there are multiple adaptive solutions to the same evolutionary problem. This has most famously been phrased as whether or not the outcome of evolution would be the same if the tape of life were played over again ( GOULD 1990 ). Some recent work has suggested that even in complex eukaryotes, replicate tapes might hav e surprisingly similar outcomes ( VERMEIJ 2006 ). Convergent and/or parallel evolut ion (evolution arriving at similar solutions to the same challenge) has been widely found in experim ental evolution systems involving microbes ( MOYA et al. 2004 ). Does this conclusion depend on whether adaptations arise from new mutations, or from standing variation? Bu ll et al. (1997) demonstrated a remarkable rate of convergence among RNA viruses in the context of host-pathogen coevolution due solely to new mutation ( BULL et al. 1997 ). Similarly, convergence and parallelism were demonstrated in E. coli desc ended from a single genetic variant, but those experiments were in the context of adaptation to a non-evolving, chemostat environment ( WOODS et al. 2006 ). However, while mutational identity has been demonstrated in RNA viruses ( WICHMAN et al. 1999b ), this result was not demonstrated 30

PAGE 31

31 in E. coli ; rather, different mutations occurred in the same genes at rates greater than in a set of randomly chosen comparison genes ( WOODS et al. 2006 ). The D. melanogaster / D. sim ulans /DMelSV system models two key features of emerging diseases: DMelSV is an RNA virus, t hus likely to evolve rapidly, and the hostshift is between closely related species ( HOLMES and RAMBAUT 2004 ). Additionally, controlled experiments in a m odel system will allow us to qu antify the virulence of the virus in the native and the novel hosts before, during, and after multiple, identic al hostshifts. Therefore, this system is an excellent model in whic h to test several hypotheses on the evolution of virulence: 1) standing genetic variation is the source material for the evolution of virulence; 2) the virus is more infective and more virulent on the novel host; 3) virus evolution on the novel host adversely affects re-infection and replication on the native host.

PAGE 32

CHA PTER 2 LOW GENETIC VARIATION BOTH WITHIN AND BETWEEN HOSTS FOR SIGMA VIRUS (RHABDOVIRIDAE) IN DROSOPHILA Background The sigma virus is a member of t he Mononegavirales, the virus order that contains the causative agents of diseases such as rabies, hemorrhagic septicemia, hematopoietic necrosis and other economically important diseases of humans and livestock ( HOGENHOUT et al. 2003 ). Additionally, there are many other RNA viruses with dipteran vectors that are also human pathogens ( FAUQUET CM 2005 ). DMelSV is vertically transmitted in the dipteran Drosophila melanogaster and causes a characteristic CO2 sensitivity such that when infected flies are exposed to the gas, they die or are paralyzed. We use the DMelSVD. melanogaster model to examine the standing genetic variation that is present in wild virus populations. Rhabdoviruses have negative single-s tranded RNA genomes. DMelSV encodes six genes (3-NP-PP3-M-G-L-5) ( CARPENTER 2008 ; TENINGES et al. 1993 ). These genes code for the nucleocapsid protei n (N), the polymerase-associat ed protein (P), the matrix protein (M), the glycoprotein (G), the polymerase (L) and a reverse transcriptase (PP3) ( CARPENTER 2008 ; TENINGES et al. 1993 ). Interestingly, the PP3 gene is found in only a few other rhabdov irus species, and shares greater homology with retroviruses and retrotransposons, suggesting that thes e viruses acquired this gene through recombination with one or both of these sources ( CARPENTER 2008 ). Viruses wit h RNA genomes are notorious for their high mutation rates ( DOMINGO and HOLLAND 1997 ). However, it is well known that negative strand RNA viruses may have lower mutation rates than positive strand viruses ( DOMINGO and HOLLAND 1997 ). 32

PAGE 33

Although mutation rate has not been directly es timated in DMelSV, substitution rate has ( 4.6 x 105 substitutions/site/year, CARPENTER et al. 2007 ). For neutral sites, substitution rate is equal to mutati on rate ( KIMURA 1968 ). The substitution ra te estimated for DMelSV is likely to be close to the neutral mutation ra te, but if so, is lower than average even for a negative strand virus. Additional evidence that the mutation rate for DMelSV may be low comes from Brusini et al. 2012 (American Naturalist, in review), where DMelSVinfected D. melanogaster lines were selected for either high or low virus titers for 12 host generations. Although a significant respons e to selection was found, very few mutations were fixed in the virus; indeed, in some case s, no mutations were fixed between pairs of lines with large divergence in titers. The low mutation rate may be due to DMelSVs lifestyle as a vertically transmi tted parasite: some vertically transmitted, apparently symbiotic viruses in plants al so have extremely low mutation rates ( ROOSSINCK 2010 ). However, some genetic variation does exist in the virus, on decada l time scales and across space; and in previous studies of DMelSV variation, a subset of the genomic variation in DMelSV was attributed to host-encoded adenosine deaminases acting on RNA (ADAR) ( CARPENTER et al. 2009 ), rather than mutation during viral replication. Adenosine deaminases action on RNA are enzymes that target dsRNA. The presence of dsRNA is often a sign of vira l infection (for ex ample, replication intermediates for single-stranded RNA viruse s are double stranded), and it is thought that ADARs may have evolved as an antiviral response ( KEEGAN et al. 2001 ). Specifically ADARs change adenosine (A) bases to inosine (I). This change affects translation, because I then pairs with a cytidine (C) instead of a thymidine (T) at that 33

PAGE 34

particular position on the anticod on (KEEGAN et al. 2001). Additionally, the I bases can pair with C bases during RNA-dependent RNA replication of the + strand genomic intermediate that is generated during rhabdovirus replication ( ROSE and WHITT 2001 ). Subsequently, the C pairs with G, resulting in the introduction of a mutation (A to G) in the viral genomes (KEEGAN et al. 2001). Tw o enzymes (ADAR1 and ADAR2) are capable of this type of mutation ( KIRCHER and KELSO 2010 ). ADAR1 catalysis the A-to-I deamination in double st randed RNA targets, while ADAR 2 cataly ses the A-to-I mutations in pre-mRNA ( GEORGE and SAMUEL 2011 ). The likelihood that a given A will be deaminated is conditioned on it s 3 or 5 neighbors. Deam ination is more likely to happen when the 5 neighbor is ei ther an A or U (27 and 32% deamination, respectively U A > C = G; Table 2-2-1 and Table 2-2) (MUELLER et al. 2006 ). This preference spans many host taxa [Rhabditid nematodes to humans ( BASS 2002 )], including D. melanogas ter ( LANDER and WATERMAN 1988 ). In Drosophila there is a single ADAR gene, dADAR. dADAR deletion mutants exhibited locomotor issues and tremors t hat worsened with age, indicating that dADAR activity also affects nervous system functi on and stability, as well as potentially having antiviral activity ( PETSCHEK et al. 1996 ). Although dADAR was not specifically implicated, ADAR activity also has been found in the DMelSV infecting D. m elanogaster ( CARPENTER et al. 2009 ) Carpenter et al (2009) s equenced ~6kb from viral genomes from wild flies and flies kept in the lab for 10-20 years. In the D. melanogaster sigma virus, more candidate bases (A without upstr eam G) were mutated to Gs than would be expected by chance alone ( CARPENTER et al. 2009 ). Additionally, these A-to-G mutations were clustered in specific regi ons of the genome, as is typical of ADAR 34

PAGE 35

activity ( CARPENTER et al. 2009 ; MOERDYK-SCHAUWECKER et al. 2009 ). These data confirm that ADAR activity is occurring in the DMelSV genome and is, at least in part, responsible for the mutations accumulated after the rece nt sweep (CARPENT ER et al. 2009; CARPENTER et al. 2007). Host-medi ated dADAR activity can lead to hypermutation in some rhabdoviruses ( CARPENTER et al. 2009 ; OHARA et al. 1984 ). Hypermutation, in turn, can contri bute to decreases in pathogen virulence ( MEYERS et al. 2003 ). Thus, the upregulation of ADAR activity in DMelSV-infected D. m elanogaster could be an antivirulence mechanism. Here we use SOLiDTM deep-sequencing from multip le Drosophila individuals within a single host population to comprehen sively evaluate the patterns of sigma sequence variation within a single populati on and within hosts, and to determine how much, if any, of the observed variati on is attributable to ADAR. The SOLiDTM sequencing platform is ideal fo r this project because the short nature of the DMelSV genome will produce extremely deep coverage at every site. Therefore, once the standard quality controls are implemented, accurate, hi gh quality, deep coverage of viruses within a single host can be obtained ( HARISMENDY et al. 2009 ). Given high coverage, we can exp ect the probability of false negatives (failing to identify variants) to be vanishingly low, while the probability of false positives will only approach 0.1, making the probability of either e rror type extremely rare ( BENAGLIO and RIVOLTA 2010 ; KUIKEN et al. 2006 ; LUSTIG et al. 2000 ; ONDOV et al. 2008 ). We will first identify mutated sites, then consider the surrounding sequence to i dentify the role of ADAR mutation within and between flies (Table 21 and Table 2-2) ( LEHMANN and BASS 2000 ). 35

PAGE 36

Materials and Methods Fl y Collection and Propagation Flies were collected from a single loca tion in Athens, GA in August 2007 using banana baits. Flies were allowed to oviposit on prepared Drosophila food for 24 hrs, after which DMelSV infection was dete rmined by exposing the flies to CO2 because flies infected with this virus becom e paralyzed upon exposure ( BRUN and PLUS 1980 ; L'HERITIER and HUGON DE SCOEUX 1947 ). The offspring of each infected fly was kept under standard reading conditions (24C and 16:8 light: dark). The result was six isofemale lines infe cted with DMelSV. Ribonucleic Acid (RNA) Extraction Ribonucleic acid was extracted from each of the six infe cted females using TRIzol (www.invitrogen.com) according to the standard manufacturers protocol and stored at -80C. Briefl y, the fly was submerged in TRIzol in a 1.5ml Eppendorf tube and homogenized with a Kontes Teflon homogen izer. Chloroform was added to the homogenate and the tube was vortex ed and incubated at room te mperature for five min and then centrifuged at 12,000g and 2C for 10 min. The upper aqueous phase that formed was removed and the RNA was prec ipitated by the addition of isopropanol followed by incubation at -20C for 30 min. The precipitated RNA was centrifuged again with the conditions listed above. The resulting pellet was washing twice in cold 70% ethanol, dried briefly at room temperat ure and suspended in TE (10mM Tris, 1mM EDTA, pH 8.0). Purified RNA was quantified using a NanoDrop and the samples were stored at -80C. For each fly, 500ng of RNA were used per reaction. 36

PAGE 37

Reverse T ranscription and Polymerase Chain Reaction A portion of the viral genome (6370 bp; partial N, G, M, X, P and partial L genes) was amplified from each fly in overlapping pieces using the Superscript III one-step RTPCR system with platinum Taq high fidelity and following the manufacturers standard protocol ( www.invitrogen.com ). Briefly, for each primer pair, 500ng of fly RNA were mixed with the reaction mix, sense and antisense primers, the su perscript enzyme mix, 1 l of RNAse inhibit or (www.promega.com) and water. The reactions were incubated at 45C for 30 min and then 94C for 2 min. This was followed by 20 cycles of 94C for 15 seconds, 55C for 30 seconds and 68C for 1 mi n. A final extension step for 10 min at 68C was run. Each of the 4 segments, which ranged from 1073 to 2544 bases, was amplified twice indep endently for 20 cycles to reduce pr opagation of PCR error, while producing enough product to sequence (thus a to tal of 4 genomic regions x 2 replicates = 8 fragments per sample). The resulting PCR products were electroph oresed, the bands were cut out of the agarose gel, and the DNA was pu rified from the resulting plugs using Qiagen QiaQuick Gel Extraction gel purification kits followi ng manufacturer's prot ocols (www.qiagen.com). The resulting purified fragm ents were quantified after ex traction using a NanoDrop and only products of 4ng/ l or greater were used for sequencing. The eight purified fragments representing the virus from each fly were combined into individual samples to yield a minimum of 2 g of DNA per sample. Libraries were made from each sample, following the standard manufacturers protocol for SOLiDTM sequencing. Briefly, the 2g of DNA for each sample was sheared and size-selected. The sequencing adaptors we re ligated and amplified. The dsDNA library was mixed with the bead emulsion acco rding to the manufacturers instructions. 37

PAGE 38

The beads were deposited and 48bp reads were obt ained. All 8 samples (six flies, with two flies repeated as internal controls ) were bar-coded, pooled, and run on a single region of a SOLiDTM 5500xl (Applied Biosystems, Foster City, CA) plate. Two of the runs (two replicates of the same fly) fail ed; therefore, only 6 samples (representing five individual flies) were incl uded in subsequent analyses. SOLiDTM Reads Mapping and Alignment Read from SOLiDTM sequencing were mapped to t he reference genome of the sigma AP30 (GenBank accession number NC_0 13135) using default settings of MAQ 0.6.6 ( http://maq.sourceforge.net/index.shtml ). Once reads were mapped, the total number of reads per site for each sample and overall were determi ned. Out of 6570 sequenced sites, there were 137 sites that were covered by 100 reads or less in at least one of the samples, and thus, these sites were excluded from further consideration (of these, only 45 sites were excluded because of under-representation of a single sample), for a total of 6433 sites used in further analysis. The parameters of t he reads density per site distri butions for each sample are given in Table 2-1. The median values r anged from 9391 (in BC1 sample) to 99574 (in BC7 sample). There were a total of 6084 and 349 sites located either in protein-coding (also referred to as coding) or non-coding regions, respectively. Adenosine Deaminases Acting on RNA and Polymorphism Analysis Sites with ADAR activity were determined using the consensus nucleotides in each individual sample, taking into account 5 adjacent nucleotides (Table 2-2). The sites have A-G SNPs were classified into strong and weak ADAR sites based on their 5 neighbors. Out of 6084 coding site s, there were 1779 ADAR sites, and 4305 38

PAGE 39

non-ADAR sites. Among ADAR sites, there were 493 sites in the weak category, and 1286 in the strong category. Results Whole-Ge nome Sequencing One of the unique samples and one of t he replicate sequencing samples failed, leaving us with five unique sequence samp les (BC1, BC2, BC3, BC5 and BC8). BC7 was the technical replicate of BC2. Although the absolute num ber of reads per site was significantly larger in BC7 than in BC2 (average of 137235 1756 versus 65846 962; paired t -test, P < 0.001), there was a strong correlation between the number of reads per site at individual sites in these two replicas ( R = 0.921; Figure 2-1), supporting consistency in the results derived from these technical replicates. Table 2-1 shows the summary statistics of the number of reads per site among all samples. The mean and the median numbers of reads per site varied from 22,283 ( 377) and 9391 in BC1 sample (which also had the smallest total number of reads) to 165,650 ( 2971) and 86496 in the BC8 sample (w hich had the largest number of total reads). Overall coverage (computed as per the Lander/Waterman equation, where coverage = L (N / G), where G is the haploi d genome length, L is t he read length, N is the number of reads) varied from slightly over 1,000,000X in BC1 to over 7,700,000X in BC8 ( Table 2-3, LANDER and WATERMAN 1988 ). However, their coverage ranged from over 2,200X to almost 18,000X among t he samples when only the reads with nonconsensus variants were considered (Table 2-1). Our initial analysis examined the over all sequencing coverage across the genome. The nucleotide composition vari ed between pr otein-coding and non-coding regions, with the latter category having more A/T sites than the former. However, 39

PAGE 40

because the majority of ADAR sites were located in the coding regions, we focused on the subset of 6084 coding si tes and characterized different classes of observed mutation events (Figure 2-2). The results of this analysis are presented in Tables 2-4 through 2-10. Subsequently we characteriz ed the number of read that fell within the coding regions and were either in non-A sites, weak or strong ADAR sites for every codon posit ion (Tables 2-6, 2-7 and 2-8). Adenosine Deaminases Acting on RNA (ADAR)-Driven Substitutions Strong ADAR sites are an adenosine (A) with preferred 5 and 3 neighbors ( Table 2-1, LEHMANN and BASS 2000 ): the combination of nei ghbor ing bases affects the propensity for ADAR activity at a particular A base ( Table 2-2, LEHMANN and BASS 2000 ). There were 6433 total sites with sufficient read coverage to be included in our analys is ( i.e. that passed quality control tests) of which 6084 sites fell within the sequenced, protein-coding regions of the portion of the DMelSV genome. A further subset of 1779 sites within the coding regi on harbored the consensus A nucleotide and were thus identified as potential ADAR site s. We then considered the respective 5 nucleotide neighbor and assigned the A-harboring sites into strong and weak ADARpreferences groups (see Table 2-4 and Figure 2-3 for details). There were 1286 and 493 sites with strong and weak ADAR prefer ence, respectively. Notably, among the 1779 sites that harbored a consensus A nucleot ide, the vast majority (1747 sites, 98.2%) of these sites harbor ed non-consensus sequence variants at relatively low frequencies (<5%, See Figure 2-4); while the rema ining fraction (32 sites, 1.8%) of sites harbored minor variants with frequencies as high as 48.5% (in BC7 sample). Thus, we separated the sites into two sub-categorie s those with high-frequency SNPs (singlenucleotide polymorphisms) present in at least 5% of a ll reads (total of 32 sites), and 40

PAGE 41

sites without high-frequency SNPs (1747 si tes). Table 2-4 depicts the summary statistics (means and medians) for the obser ved reads harboring t he minority nucleotide at 1747 sites by individual samples. The dire ction of the mutation is assumed starting from the consensus nucleotide (A in all cases at these sites; for example, A-to-T, A-toC, etc.) and the 5 ADAR preference is also taken into account. As evident from Table 24 and Figure 2-5, the number of A-to-G reads exceeds the number of A-to-T and A-to-C reads in all samples and in all ADAR site categories (Table 2-4). It should be noted that although the absolute number of reads per site (Figure 2-5) cannot be directly compared across samples (because of di fferences in overall coverage between samples), the trends in A-to-G, A-to-T and A-to-C changes are essentially the same among samples. Further, minor differences can be observed between samples, with BC1 and BC5 having similar number of A-to-G changes in both weak and strong site categories, while BC3 and both technical replicates (BC2 and BC7) show slightly more A-to-G changes at weak ADAR sites. The opposite trend is observed in BC8. Interestingly, slightly higher numbers of both A-to-T and A-to-C changes are observed in all samples at sites characterized as strong ADAR preference sites. The reason for this is not clear; it might be biologically relevan t, or attributable to stochastic differences in the number of reads due to the 3-fold differ ence in the number of sites in each category (485 strong versus 1262 weak sites). The SNP sites harboring non-A-to-G changes (in other words, those with A-to-T or A-to-C changes) were combined into a single mutation category in 2x2 contingency tables to test whether the sites with st rong ADAR preference harbor more A-to-G mutations than the other two possible changes. A Fishers exact test confirms that three 41

PAGE 42

of the samples, BC8, BC2 and BC7, have significantly more mutations in ADARpreferred than in nonADAR-preferred site categories ( P = 0.0019, P = 0.0473 and P = 0.0184, respectively; Table 2-5). In BC 1, BC3 and BC5 the tw o preferred and nonpreferred categories are statis tically indistinguishable ( P = 0.3583, P = 0.0575 and P = 0.1990, respectively). It is reassuring that the technical replicat es, BC2 and BC7, have similar mutational trends, with smaller numbe r of A-to-G mutations harbored at strong ADAR sites compared to weak sites. On the other hand, BC 8 sample shows more A-toG hypermutations at the strong ADAR sites than at the weak ADAR si tes. Therefore, in these rare SNP sites there was betw een-host heterogeneity with regard to the mutational pattern. This pattern of ADAR-driven mutations is much more prominent at a subset of 32 sites that harbor high-frequency SNPs (withinfly variants at 5% or more). For these sites, the strong preference ADAR category sites harbor significantly more A-to-G hypermutation changes than the weak category sites in all but one sample (Fishers test: BC1, P = 0.1613, all other P < 0.0005; Table 2-7). As with the larger subset of sites, these differences cannot be solely attr ibuted to the differences in the absolute number of reads per site, as these values do not differ significantly between weak and strong site categories (ANOVA, P > 0.05 for all samples). Additionally, when the locations of the SNP-harboring A-consensus sites were considered, we noted a cluster of 7 ADAR sites (out of total 15 SNP sites) in the glycoprotein (G ) gene (Table 2-8), as might be expected if the ADAR mechanism were responsible for these mutational changes ( CARPENTER et al. 2009 ; LEHMANN and BASS 2000 ). 42

PAGE 43

The functional consequences of hi gh-frequency SNPs across all 94 SNPharboring s ites are explored in Table 2-8. On average, half of all SNPs in each sample lead to non-synonymous (amino acid changing) substitution, with a similar fraction of nonsynonymous changes occurring at ADAR site s (Figure 2-6). The overall numbers of sites harboring high-frequency SNPs are quite similar between samples (ranging from 54 to 60 between all but BC8 samples). Interestingly, the majority of sites with nonsynonymous substitutions (31 out of 42 sites) are shared between samples, with only 11 sites harboring different mutati ons in different samples, s uggesting that either some sites are evolving under rather strong sele ctive pressures leading to (convergent) amino acid changes, and/or flies may have been co-inf ected by multiple viral haplotypes that, in turn, had already been selected to r easonably high frequency in the viral population prior to infection. Discussion We set out to elucidate the role of ADAR on DMelSV genetics and role of the genetic diversity of DMelSV populati ons within individual wild-caught D. melanogaster females using SOLiDTM deep-sequencing. As expected, the SOLiDTM platform combined with the small natur e of the viral genome resulted in extremely deep sequencing coverage of the roughly 6500 bp region of the genome we explored. Although we found significant gen etic variation, this variat ion was not on the scale that might be expected from most RNA viruses ( DOMINGO and HOLLAND 1997 ). We also found signif icant ADAR activity, as well as di fferences in ADAR intensity between hosts, which suggests that, at least in wild virus (within wild flies), ADARs play an important role in shaping the genome of the DMelSV r habdovirus, potentially at least as important as replication-driven mutation. 43

PAGE 44

44 Error is always one of the main c oncerns with any deep-sequencing project, particularly one such as ours, where the sequencing platform providing extremely high coverage is combined with a very short templa te (6.5kb). To help correct for possible error, we included an internal control where the same sample was sequenced twice during the same sequencing run. Although the numbers of reads were different for the two runs, the numbers of reads per site per individual were highly correlated. This indicates that, on average, every position along the target length was proportionally represented in our two techni cal replicates. Thus, the differences in read number are likely resulting from slight differences in sample processing after the original sample was halved that led to an effectively larger sa mple in the BC7 replicate. Additionally, the two technical replicates had similar mutation trends; with the smalle r number of A-to-G mutations harbored at strong ADAR sites compar ed to weak ADAR site s. Therefore, the technical replicates were very similar despite the different initial read coverages. Among sequenced sites, we identified 1779 A-harboring sites on the consensus sequence that were likely candidates for ADAR activity. Of these, 1747 (98.2%) were identified as sites harboring rare variants wh ere A bases were mutated. Although many of these variants were A-to -G mutations, strong ADAR si tes were not significantly preferred over weak sites in all samples, which suggest that (i) ADAR activity was not responsible for the production of some A-to -G mutations in these minor variants; and that (ii) biological heterogeneity in ADAR-d riven activity exists among hosts.

PAGE 45

Table 2-1. Descriptive statistics parame ters characterizing the num ber of reads per site at 6433 sites for each vi rus sample. Sample Mean SE Mean Minimum Median Maximum Total Reads Minorityvariants Coverage BC1 22,283 377 105 9,391 257,565 143,349,074 313,668 1,047,299 BC2 65,846 962 208 36,078 590,978 423,588,427 947,782 3,094,710 BC3 96,756 1,848 213 33,455 1,353, 053 622,430,429 1,355,807 4,547,437 BC5 82,762 1,180 242 41,362 776,760 532,408,152 1,427,714 3,889,740 BC7 137,235 1,756 332 99,573 1,283, 161 882,832,544 1,959,110 6,449,918 BC8 165,650 2,971 641 86,496 2,389, 086 1,065,627,070 2,451,521 7,785,403 Shown are mean (with st andard error, SE), median, minimu m and maximum. BC2 and BC7 are technical replicates of the same fly. 137 sites which had at least one sample with the total number of reads below 100 were excluded (see Methods). 45

PAGE 46

Table 2-2. Independent nearest neighbor preferences for adenosine deamina ses acting on RNA (ADAR). Based on Lehmann and Bass (2000)*. *Lehmann KA, Bass BL 2000. Double-strand ed RNA adenosine deaminases ADAR1 and ADAR2 have overlapping specific ities. Biochemistry 39: 12875-12884 Site classification 5 neighbor Total % deaminated ADAR-preferred A 42 27 ADAR-preferred U 32 32 ADAR-non-preferred G 35 9 ADAR-non-preferred C 36 13 3 neighbor Total % deaminated ADAR-non-preferred A 42 11 ADAR-preferred U 34 27 ADAR-preferred G 37 29 ADAR-non-preferred C 32 13 Table 2-3. Adenos ine deaminases acting on RNA (ADAR) triplet preferences. Based on Lehmann and Bass (2000). Triplet % deaminated UA U 47 AA G 44 UA G 42 AA U 30 CA G 23 AA C 23 UA C 21 UA A 17 CA U 17 GA U 14 GA G 13 AA A 11 CA A 9 GA A 6 GA C 5 CA C 4 *Lehmann KA, Bass BL 2000. Double-stranded RNA adenosine deaminases ADAR1 and ADAR2 have overlapping specific ities. Biochemistry 39: 12875-12884 46

PAGE 47

Table 2-4. Mutations at 1747 sites harborin g consensus A nucleotide with in proteincoding regions, excluding the 32 site s that harbor high-frequency single nucleotide polymorphisms. Sample Mutation 5 ADAR preference N Mean SE Mean Median Weak 485 27.23.1 11.0 BC1 A-G Strong 1,262 27.21.8 10.0 Weak 485 5.50.8 1.0 BC1 A-T Strong 1,262 9.20.6 3.0 Weak 485 5.60.5 2.0 BC1 A-C Strong 1,262 8.70.9 2.0 Weak 485 89.78.8 37.0 BC2 A-G Strong 1,262 86.05.7 31.0 Weak 485 17.71.9 5.0 BC2 A-T Strong 1,262 25.21.9 7.0 Weak 485 14.71.3 6.0 BC2 A-C Strong 1,262 28.32.9 7.0 Weak 485 133.015.2 41.0 BC3 A-G Strong 1,262 112.87.7 33.0 Weak 485 23.93.1 6.0 BC3 A-T Strong 1,262 32.32.4 9.0 Weak 485 27.52.5 9.0 BC3 A-C Strong 1,262 34.43.8 8.0 Weak 485 113.310.5 50.0 BC5 A-G Strong 1,262 113.37.6 46.0 Weak 485 32.23.6 11.0 BC5 A-T Strong 1,262 39.03.3 14.0 Weak 485 28.92.3 13.0 BC5 A-C Strong 1,262 43.2 4.0 13.0 Weak 485 190.6 17.4 100.0 BC7 A-G Strong 1,262 159.8 9.0 76.0 Weak 485 39.5 4.5 14.0 BC7 A-T Strong 1,262 51.9 3.8 17.0 Weak 485 37.6 3.4 15.0 BC7 A-C Strong 1,262 55.1 5.5 16.0 47

PAGE 48

48 Table 2-4 Continued. Sample Mutation 5 ADAR preference N Mean SE Mean Median Weak 485 185.919.7 85.0 BC8 A-G Strong 1,262 214.122.5 72.5 Weak 485 33.03.0 12.0 BC8 A-T Strong 1,262 73.76.4 18.0 Weak 485 36.73.4 13.0 BC8 A-C Strong 1,262 67.59.7 14.0 Weak 485 11.11.1 1.0 BC1 A-TC Strong 1,262 17.91.2 6.0 Weak 485 32.42.8 13.0 BC2 A-TC Strong 1,262 53.53.8 17.5 Weak 485 51.45.1 17.0 BC3 A-TC Strong 1,262 66.75.0 18.0 Weak 485 61.05.1 26.0 BC5 A-TC Strong 1,262 82.25.7 32.0 Weak 485 77.16.9 30.0 BC7 A-TC Strong 1,262 107.07.7 37.0 Weak 485 69.85.5 28.0 BC8 A-TC Strong 1,262 141.213.2 38.0 Single nucleotide polymorphisms (SNPs) m ean numbers (with standard error, SE) of different categories of mutations (A to G, A to T, A to C and cumulative A to T or C) at (over 5% of alternative nucleotide). Sites are categorized as having either weak or strong adenosine deaminases acti ng on RNA (ADAR) preferences. Table 2-5. Adenosine deaminases acting on ri bonucleic acid (ADAR) preferentially changed strong ADAR sties. P-values from Fishers exact test and mean numbers of different classes of mutations per site (ADAR-driven A to G hypermutati ons and non-ADAR-driven A to T or A to C mutations) at strong and weak ADAR sites. Based on 1747 sites harboring consensus A nucleotide within protein-coding regions [excluding 32 sites that harbor high-frequency SNPs (over 5% of alternative nucleotide)]. Mean # of A-to-G Mean # of A-to-T-or A-to-C Line Weak Strong Weak St rong Fishers test (p) BC1 27 27 11 18 NS (0.3583) BC3 133 113 51 67 NS (0.0575) BC5 113 113 61 82 NS (0.1990) BC8 186 214 70 141 P = 0.0019 BC2 90 86 32 53 P = 0.0473 BC7 191 170 77 107 P = 0.0184

PAGE 49

Table 2-6. Descriptive statistics parameters characterizing the number of reads per site at protein-coding sites in the 1st codon position for each sample, separated by whether the site is A-harboring [and fu rther subdivided into weak or strong adenosine deaminases acting on RNA (ADAR preference] or not. Sample ADAR preference N Mean SE Mean Minimum Median Maximum Non-A 1,412 22,9498301059,432 257,565 Weak 184 23,9842,4742409,447 201,206 BC1 Strong 432 22,6451,3501469,818 191,636 Non-A 1,412 100,5714,09121333,881 1,353,053 Weak 184 102,73912,06759434,172 1,039,318 BC3 Strong 432 94,4196,47542633,858 951,589 Non-A 1,412 85,5572,58472742,149 776,760 Weak 184 87,5017,62578346,341 618,836 BC5 Strong 432 84,7884,27824245,353 575,819 Non-A 1,412 171,0506,61594989,231 2,389,086 Weak 184 183,49719,5602,08491,067 1,931,032 BC8 Strong 432 163,90910,1742,16592,499 1,774,885 Non-A 1,412 68,1282,11021636,476 590,978 Weak 184 67,0395,86548036,524 454,967 BC2 Strong 432 67,5653,62823337,213 423,946 Non-A 1,412 142,3303,898332102,740 1,283,161 Weak 184 141,19111,0981,58998,422 953,685 BC7 Strong 432 138,2126,227396100,121 917,676 Shown are mean (with standard error, SE), median, mini mum and maximum. BC2 and BC7 are technical replicates of the same fly. 49

PAGE 50

Table 2-7. Descriptive statistics parameters characterizing the number of reads per site at protein-coding sites in the 2nd c odon posit ion for each sample, separated by whether the site is Aharboring [and further subdivided into weak or strong adenosine deaminases acting on RNA (A DAR) preference] or not. Sample ADAR preference N Mean SE Mean Minimum Median Maximum Non-A 1,382 22,809 816 121 9,495 254,707 Weak 233 25,557 2,255 246 9,705 191,446 BC1 Strong 411 21,868 1,453 115 9,335 192,412 Non-A 1,382 99,675 4, 053 372 33,976 1,335,779 Weak 233 109,571 10,596 590 35,597 1,076,402 BC3 Strong 411 91,977 7,032 242 32,324 927,653 Non-A 1,382 85,337 2,559 692 44,627 770,288 Weak 233 92,355 6,812 1,706 41,464 600,593 BC5 Strong 411 81,956 4,575 309 41,058 581,909 Non-A 1,382 168,626 6, 504 979 89,840 2,360,591 Weak 233 194,946 17,214 2,269 99,048 1,880,237 BC8 Strong 411 162,423 11,358 1,095 87,804 1,733,237 Non-A 1,382 67,727 2,103 269 36,502 582,717 Weak 233 74,719 5,523 559 38,341 482,167 BC2 Strong 411 64,101 3,658 219 35,615 432,324 Non-A 1,382 140,730 3, 854 540 100,526 1,262,557 Weak 233 156,616 10,104 2,260 106,374 1,068,402 BC7 Strong 411 132,285 6,613 376 98,319 939,369 Shown are mean (with standard error, SE), median, mini mum and maximum. BC2 and BC7 are technical replicates of the same fly. 50

PAGE 51

51 Table 2-8. Descriptive statistics parameters characterizing the number of reads per site at protein-coding sites in the 3rd codon position for each sample, separated by whether the site is Aharboring (and further subdivided into weak or strong ADAR preference) or not. Sample ADAR preference N Mean SE Mean Minimum Median Maximum Non-A 1,511 22,582 774 114 9,419 237,843 Weak 76 23,110 3,682 113 10,898 220,251 BC1 Strong 443 24,021 1,537 275 9,823 252,351 Non-A 1,511 97,257 3,820 317 33,428 1,239,851 Weak 76 104,221 18,736 435 39,166 1,143,098 BC3 Strong 443 104,927 7,338 840 34,823 1,321,982 Non-A 1,511 84,472 2,445 623 43,215 727,102 Weak 76 87,874 12,070 243 43,643 676,046 BC5 Strong 443 87,974 4,608 2,451 42,835 765,518 Non-A 1,511 166,210 6,033 641 88,317 2,227,523 Weak 76 176,578 32,233 953 103,082 2,099,507 BC8 Strong 443 182,970 12,279 2, 740 95,069 2,330,211 Non-A 1,511 66,691 1,999 208 36,062 541,349 Weak 76 72,447 10,088 219 37,928 500,600 BC2 Strong 443 70,373 3,701 748 38,070 575,834 Non-A 1,511 138,177 3,580 470 99,490 1,084,533 Weak 76 149,555 18,917 427 105,172 1,059,282 BC7 Strong 443 147,188 6,955 3,171 105,992 1,242,504 Shown are mean (with standard error, SE), median, mini mum and maximum. BC2 and BC7 are technical replicates of the same fly.

PAGE 52

Table 2-9. P-values from Fishers exact test and mean numbers of different cl asses of mutations per site [adenosine deaminas es acting on RNA (ADAR)-driven A to G hypermutations and non-ADAR-driven A to T or A to C mutations] at strong and weak ADAR sites. Mean # of A-to-G Mean # of A-to-T-or-C Line Strong Weak Strong Weak Fishers test (p) BC1 24 22225151NS (0.1683) BC3 51 781 155 332 P < 0.0001 BC5 157 836 108 335 P = 0.0002 BC8 159 1204 235 505 P < 0.0001 BC2 48 614 58 220 P < 0.0001 BC7 99 923 198 458 P < 0.0001 Based on reads at 32 sites that harbor hi gh-frequency SNPs (over 5% of alternative nucleotide). 52

PAGE 53

Table 2-10. Summary of non-synonymous substitutions (diffe rent from the consensus) th at occur in each sample, organized by gene, codon number and base position within each codon (codon position). Gene Codon # Codon position Consens. base Consens. codon Consens. AA BC1 BC2 BC3 BC5 BC7 BC8 ADAR N 104 2 T GTA V E E E E E G N 114 1 T TCG S P P P P P A N 114 2 C TCG S Stop Stop Stop Stop Stop W N 407 2 G CGC R H H H H H N 412 2 C GCT A V P 689 2 T ATA I T T T T R P 734 3 C ACC T T P 825 1 T TAG STOP E E E X 855 1 T TTA L V X 901 1 A AAT N D D D D D D N/D X 903 1 G GTC V L L L L X 1027 2 T ATG M K K K K K R X 1120 2 C ACC T N N N N N S M 1262 1 T TGG W R R R R R G M 1262 2 G TGG W L L L L L M 1262 3 G TGG W C C C M 1263 1 C CAT H Y Y Y Y M 1263 2 A CAT H R H/R M 1264 1 G GCC A P P M 1264 2 C GCC A G A/G G 1414 3 C GAC D E E E E G 1578 2 A GAG E G G G G G G E/G G 1579 1 A AGC S G G G G G 1580 2 G GGG G E A A A A S/G G 1588 2 A CAT H P P P P P R H/R G 1604 2 A GAT D V V V V V G D/G G 1679 1 A AAA K E E E E K/E G 1757 2 T CTC L P P P P P G 1764 1 A ATT I F F F F F V I/V 53

PAGE 54

Table 2-10. Continued. Gene Codon # Codon position Consens. base Consens. codon Consens. AA BC1 BC2 BC3 BC5 BC7 BC8 ADAR G 1765 1 G GTT V F F F F G 1827 2 T CTG L Q Q Q Q Q R G 1835 2 T TTC F S G 1883 2 C TCC S F F F F F C G 1901 1 T TGA STOP G G G G G G 1901 3 A TGA STOP C C C L 1970 1 C CAT H Y Y Y Y D L 2005 3 G TGG W C C C C C L 2009 3 T TTT F L L 2058 3 C ATC I M L 2059 1 A ACC T P P P P P A T/A L 2059 2 C ACC T S S S S S S L 2060 1 T TTG L V Total Mutations 60 60 54 59 56 45 NonSynonymo us 32 34 28 32 24 26 Total ADA R786751 34232 2 Non-syn ADAR 9 54

PAGE 55

Figure 2-1. Correlation of read number fo r our technical replicate indicates that the two samples behaved similarly in the sequencing process (P < 0.001 and R = 0.921). 55

PAGE 56

56 Figure 2-2. Shown are tw o possible transitions ( ) and transversions () mutati on events (that together form a substitution matrix).

PAGE 57

Figure 2-3. Flow chart describing how the 1747 candidate sites for adenosine deamin ases acting on RNA (ADAR) activity were identified. 57

PAGE 58

Figure 2-4. Minority reads were a minor component of our total read count. The number of minority nucleot ide reads accounted fo r < 5% of the reads on any one base position along the genome. 58

PAGE 59

A C EF D B A C EF D BReads per site Reads per site Reads per siteAdar preference strength Substitution category Figure 2-5. A-G site changes occurred disp roportionally more often in all flies indicating adenosine deaminases acting on RNA (ADAR) activity. Sample-specific histograms summ arizing mean numbe rs of minorityharboring reads for each mutation type (e.g. A-to-G) computed per site at 1747 A-consensus protein-coding sites. A = BC1, B = BC3, C = BC5, D = BC8, E = BC2 and F = BC7. 59

PAGE 60

BC1BC2BC3BC5BC7BC8Substitutions 0 10 20 30 40 50 60 70 total non-synonymous total adar non-synonymous ADAR Figure 2-6. Total number of substitutions (that occur with frequencies of at least 5% in one of the samples) and type of substitution in each of the flies tested. 60

PAGE 61

CHA PTER 3 EVOLUTION OF VIRULENCE IN THE SIGMA VIRUS FOLLOWING A HOST SHIFT IN DROSOPHILA Background Animal parasites, particularly viruses, regularly shift from their native wild animal hosts to pets, livestock or other wild animals that associate with humans, thus creating the potential for zoonoses and the evolution of virulence ( BARBOZA et al. 2008 ; BULL 1994 ; MACKENZIE et al. 2001 ). However, it is expected that most instances of zoonosis go unnotic ed for several reasons: virulence may be low, occurrence of disease may be rare, and most shifted parasites fail to establish stable populations in a human host. Even those parasites that do establish in one individual may not have the capacit y for human-to-human transmission (COMBES 2005 ). Viruses that achieve human-to-hum an transmission can cause significant epidemics ( PIALOUX et al. 2007 ) or even pandemics: for example, human immunodeficiency virus (HIV) ( DENNEHY 2009 ; SALVAUDON et al. 2007 ) and, more recently, H1N1 influ enza ( FLYNN 2010 ; FRASER et al. 2009 ). It seems likely that viruses shift into humans on a regular basis; we cannot predict how each will affect the human population. Predicting the potential impa ct that viruses have on human populations requires a better understanding of the evolution of virulence following host shifts. Much of our understanding of the evolution of parasite virulence has been developed using serial passage experiments (SPE), in which pathogens are passaged through single-celled hosts for many generations ( EBERT 1998 ; LEVIN et al. 1999 ). Four generalizations hav e emerged from SPE studies concerning the evolution of parasite viru lence after a host shift: 1) virulence 61

PAGE 62

increases on the novel host, 2) att enuation in the former host occurs concomitantly with the incr ease in virulence on the novel host, 3) the same substitutions that cause increased virul ence in the new host are responsible for attenuation in the original host, and 4) t he changes generally occ ur most rapidly when the parasite in question is an RNA virus ( EBERT 1998 ). However, the degree to which thes e results can be extended to multicellular hosts is unclear. In the current study we focus on the first generalization: does virulence increase in the novel host following a host shift? Parasite virulence affects parasite fi tness. Maximal par asite fitness is mediated by an intricate tradeoff between virulence and transmission ( ANDERSON and MAY 1979 ; ANDERSON and MAY 1982 ). High virulence leads to high host mortality and thus decreased transmission. Therefore, in theory every parasite can reach its optimal fi tness by simu ltaneously exhibiting optimal transmission and virulence levels ( ALIZON 2008 ; ALIZON et al. 2009 ). However, some argue that the issue is more complex and requires the in clus ion of host recovery (pathogen clearance), drift, genetic bottlenecks as well as group and kin selection ( ALIZON 2008 ; ANDERSON and MAY 1982 ; FRANK 1992 ; FRANK 1996 ). Some argue that the tradeoff hypothesis alone is weak at best in explaining the evolution of virulence ( EBERT and BULL 2003 ). Others point out t hat it is hard to understand exactly how the trade-offs work for individual parameters affecting the evolution of virulence ( ANDERSON and MAY 1982 ; FRANK 1996 ); which suggests that more data on tradeoff curves are needed to help infor m each of the parameters in the model ( ALIZON et al. 2009 ; BOLKER et al. 2010 ). We expect that 62

PAGE 63

in a naturally evolving system, hig h virulence will typically come at the cost of low viral fitness and reduced transmission ( ANDERSON and MAY 1982 ). In SPE studies, however, virulence may increase be cause viral fitness is guaranteed by the experimenter and, consequently, viru lence does not incur a cost on the parasite within t he SPE system ( EBERT 1998 ). Here we use a dipteran model system to explore the evolution of virulence following an interspecific shift between mu lticellular hos ts. Dipteran vectors like mosquitoes, sand flies, and blackflies ca rry RNA viruses that are the causal agents of diseases that cause extensive human suffering ( HOGENHOUT et al. 2003 ). We use Drosophila melanogaster (nati ve host) and its sibling species, D. simulans (novel host), which are separated by more than one million years of evolution. Drosophila melanogaster and D. simulans are reproductively isolated; crosses of D. melanogaster to D. simulans produce large numbers of sterile female offspring, while the reciprocal cr oss is much more difficult to accomplish ( STURTEVANT 1920 ). The parasite used in the model is the rhabdovirus sigma (DMelSV), which is endemic to wild D. melanogaster populations and is not found in D. simulans, even though the two species are sympat ric over most of their range ( FLEURIET 1982 ; FLEURIET 1988 ). DMelSV is a negative single-stranded RNA virus (Mononegavirales) from the Rhabdoviridae, a virus fam ily that includes the causative agents of rabies, hemorrhagic septicemia, hematopoietic necrosis and several other diseases of humans and livestock ( FAUQUET CM 2005 ; HOGENHOUT et al. 2003 ). DMelSV imparts a characteristic CO2 sensitivity to infected 63

PAGE 64

indiv iduals that makes them easily identifiable ( FLEURIET 1988 ; L'HERITIER 1958 ). The virus is vertically transmitted by bot h parents; however, male transmission is both incomplete and ineffective over multiple generations ( FLEURIET 1996 ; WAYNE et al. 2011 ). Conversely, infected females pass the infection to most or all of the ir offspring; the latter produce 100% infe cted offspring and are termed stably infected ( BRUN and PLUS 1980 ). Pizzatto et al. ( 2010 ) demonstrated that it is easier to create a persistent infection fo llo wing a host shift with DMelSV in more closely related taxa than in more distant ones, similar to trends seen in vertebrate systems ( BASS et al. 1997 ; KIMURA 1968 ; NISHIKURA 2010 ). Collectively, the traits of the coevolving virus and dipteran host ma ke this an excellent system in which to test the generality that virul ence increases after a host shift. We used this system to test several questions pertaining to host shifts. First, is infecting a novel host more difficu lt than infecting a nave native host? Second, is virulence be higher on the on the novel host than on the native host, as was the case in SPEs? Third, does viru lence correlate with viral titer, or is virulence independent of titer and inst ead caused some other mechanism? To address these questions we artificially infected nave, highly inbred lines of D. melanogaster and D. simulans with the same viral inoculum. Having successfully infected the flies, we selected fly lines with complete transmission (100% infected offspring) over three g enerations and determined infect ion prevalence (a proxy for transmission efficiency) through 16 generations. We dete rmined host fitness (a proxy for virulence) and virus tite r using strand-specific quantitative PCR ( KOMURIAN-PRADEL et al. 2004 ) at regular intervals. 64

PAGE 65

Materials and Methods Fl y Lines We infected one effectively isogenic line of D. simulans (MD106; www.dpgp.org) and one highly inbred line of D. melanogaster [Diallel 75 ( YANG and NUZHDIN 2003 )] by injection with a common, pooled viral inoculum (see next section for description of inoculum ). MD106 is one of the effectively isogenic lines included in the Drosophila Population Genomics Pr oject (www.dpgp.org), and has a complete light shotgun sequence available. Diallel 75 is a highly inbred line (40 generations of fu ll sib inbreeding ( YANG and NUZHDIN 2003 )). Inoculum and Injections We infected one effectively isogenic line of D. sim ulans (MD106; www.dpgp.org ). The inoculum was prepared from 1000 DMelSV-infected flies (600 + 400 ) collected from twelve infected isofemale lines generated from infected females collected in Athens, GA (June 2008). The inoculum was prepared following Clark et al. (1979). Br iefly, the flies were homogenized in buffer (HB: 0.005 M Tri s-HCl, 0.25M su crose, pH 7.5). The homogenate was centrifuged at 1200g for 15 min at 2C. The resulting supernatant was centrifuged at 6000g for t en min at 2C. The resu lting supernatant then was filtered through a 0.45m filter and centrifuged at 19,500g for one hour. The resulting pellet was suspended in 2ml of HB, aliquoted and stored at -80C ( CLARK et al. 1979 ). The flies (1-2 days old females) w ere injected using a Sutter instruments XenoWorks Analog Microinjector equi pped with a 100 l syringe and Narishige GD-1 needle blanks that were pulled with a Kopf Model 720 vertical needle 65

PAGE 66

puller. Prior to injection, the microi njector tubing, needle holder, and the needle were back-filled with physiological grade mineral oil. Th e inoculum was thawed on ice, agitated, and the needl e back-filled with inoculum. The flies to be injected were anesthetized on ice and kept on ice until they were injected. The needle was inserted about the second abdominal tergite and 0.1 l of inoculum was dispensed. In total, 550 D. melanogaster and 550 D. simulans were injected. Additionally, 30 flies from each species were mock-injected with HB to control for any effects of the injection procedur es. These mock-injected flies produced our uninfected control lines. Rearing Conditions, Artifici al Selection for Infection, and Relaxed Selection All flies were reared at 24C wit h a 16:8 light:dark photoperiod on JazzMix Drosophila food following the manufacturers instructions and using 7 ml of food per fly vial. New vials were set up every 21 days with constant density and sex ratio (5 + 5 ); females were allowed to ov iposit for seven days. This schedule was followed for both species and was necessary because under our rearing conditions, D. simulans did poorly when kept on a 14 day schedule. Artificial selection was conducted to produce lines in which 100% of the offspring died (complete transmission). After injection, each fly (Generation 0, G0) was deposited on a prepared vial (food, KimWipe, yeast, cotton) along with a male and allowed to recuperate and ovi posit under standard rearing conditions (480 D. melanogaster and 485 D. simulans ). The offspring (G1) of the injected flies (G0) emerged and were transferred to individual vials of fresh food 20 days after the injection date. These G1 flie s were flipped onto new vials again 10 days later to prevent them from mixing with their offspring. When the G1 flies were 25 66

PAGE 67

days old ( days), they were split haphazar dly into two groups. One group was exposed to carbon dioxide for 5 min at room temperature and scored for CO2 sensitivity. For any vial that contained some gas-sensitive flies, their un-gassed siblings were then set up in fresh vials as 1 + 1 under the assumption that some of these flies were carrying the vi rus. Vials that had no gas-sensitive flies were discarded. This process of flipping, gassing, and 1 + 1 setup continued for three generations, or until 100% infecti on was achieved, whichever came first (Figure 1, 2). In total, 2,487 fly vials were gassed during this selection period. After three generations of selection, all 100% infection lines (13 lines for each species, see results section below) were maintained without selection at constant density (5 + 5 ) on a 21 day schedule under standard rearing conditions as described above. Fecundity and Hatchability Egg production and per cent hatchability were determined by allowing flies to oviposit on green tinted food (for ease of egg detection; tinted with McCormick & Co., Inc, Hunt Valley, MD) on gl ass slides over a 48 hour period. Approximately 60% of eac h standard glass slide was covered with 1ml of the colored diet, allowed to set, and stored at 4C. The slide were sprayed with a 4mg/ml yeast solution and provided to t he flies for oviposition. After a 12hr oviposition period, the slides were re moved from the vial and replaced with a fresh slide. The eggs were counted and the slide was incubated at room temperature for >24hrs. Subsequently, the eggs that failed to hatch on these same slides were counted to determine per cent hatchability. 67

PAGE 68

Ribonucleic Acid (RNA) Purification Ribonucleic acid was isolated from individual flies using TRIzol (www.inv itrogen.com) and followi ng the manufacturers directions. Briefly, each fly was homogenized in 200 l of TRIzol on ice using a Kontes Teflon homogenizer. Each sample was incubated at room temperature (RT) for 5 min. Each sample then received 40 l of chloroform and was vortexed for five seconds, incubated at RT for 2-3 min, then centrifuged at 17,000g for 10 min at 2C. The upper aqueous phase was collect ed and the RNA was precipitated by the addition of 100 l of isopropanol fo llowed by vigorous s haking and incubation at -20C for 10 min. The precipitate wa s pelleted by centrifugation at 17,000g for 10 min at 2C. The pellet was washed t wice with 1ml of cold (4C) ethanol, air dried for five min at RT, resuspended in 29 l of DEPC-treated water with 1 l of 40 M RNAse inhibitor (www.neb.com) and incubated at 55C for 10 min to aid resuspension. Samples were stored at -80 C until used. Reverse Transcription Ribonucleic acid was transcribed using the Promega AMV RT enzyme (www.promega.com) following t he manufacturers directions, except that the protocol was scaled down to use 50ng of template per reaction. Briefly, 50ng of template RNA and 12.5ng of a tagged reverse primer ( GCAGTATCGTGAGTTCGAGTGT CCGATGACCTGTCCGTAACT ; tag underlined and gene-specific portion in it alics) and water (as needed to bring total volume to 11 l) were incubated at 65C for 5 min and then plunged into an ice bath. Subsequently, 5 l of AMV 5X buffer, 1 l of RNAse inhibitor, 2.5 l of sodium pyrophosphate, 2.5 l of 10mM dNTP mix (www.promega.com), 0.1 l 68

PAGE 69

of AMV RT enzyme and water were added to the tube to attain a fi nal volume of 25 l. The sample was agitated, centri fuged briefly, inc ubated for one hour at 42C and stored at -20C. Quantitative Polymerase Chain Reaction Strand-specific qPCR was done to in sure that only negative strand RNA genomes were quantified (PURCELL et al. 2006 ). For each RT-product, a 50 l reaction was made with 25 l of TaqMan Gene expr ession master mix (www.appliedbiosystems.com, Cat. #4369016), 0.9 l of 50M tag primer (GCAGTATCGTGAGTTCGAGTGT), 0.9 l of 50 M gene specific reverse primer (GAGTCGCAGCTTTGGAGTTC), 1.25 l of a 10 M hybridization probe (CATGAGATGGAGGAACTTTCTCTCCCA), 3.3 l of RT product and 18.65 l of water. The reactions were vortexed and centrifuged and 15 l were plated in triplicate. A standard curve was made using a PCR-amp lified and purified fragment of the N-gene that was diluted from 107 to 101 copies/ l. The reactions for the standard curve were identical to the sample reactions. The samples were amplified and quantified on a StepOne Plus thermal cycler (www.appliedbiosystems.com) wit h the following conditions: 50C for 2 min, 95C for 10 min, then 40 cycles of 95C for 15 seconds and 60C for 1 min. Statistical Analyses Differences in the proportions of in jected flies from each species that became infected, and the proportions of infe cted flies that transmitted the virus to 100% of their offspring (complete trans mission), each were evaluated using G tests. 69

PAGE 70

The data for transmission efficiency under three generations of selection were not normal even when the data for each line were averaged (Shapiro-Wilk W test, P < 0.0001). As such, the data were analyzed with a Wilcoxon test followed by a non-parametric comparison fo r all pairs using the Wilcoxon each pair test. The data for trans mission efficiency under relaxed selection also were not normal (Shapiro-Wilk W test, P < 0.0001, generations 6-16) and were analyzed as above. Data for titer were also not normal (Shapiro-Wilk W test, P < 0.0001) and failed Levene's homogeneity of variance test (P =0.0492); as such the data were analyzed with a Welch's ANOVA. A dditionally, non-parametric Kendalls Tau correlation tests were used to compare in fection rates and titer, both to avoid assumptions of normality and to reduce the impact of outliers. The fecundity and hatchabilit y data for the infected flies were standardized to the mean of their uninfected count erparts to control for environmental differences between generations. Both standardized data sets were not normal (Shapiro-Wilk W test, P < 0.0001). The average relative fecundity and hatchability were calculated for each fly line at each generation to eliminate nonindependence effects and thes e data (also not normal, Shapiro-Wilk W test, P < 0.0001) were analyzed as above. All statistical analyses were conducted using JMP 10 ( www.jmp.com ). Results Artificial H ost-Shift Success an d Selection for Transmission We injected similar numbers of D. melanogaster (N = 480) and D. simulans (N = 485) females. Injected D. simulans females produced infected 70

PAGE 71

(CO2-sensitive) offspring ( N = 78, 16.1%) at a significantly higher rate than did D. melanogaster females ( N = 45, 9.3%; P = 0.0121; Figure 3-1). Although a larger proportion of the infected D. melanogaster lines (28.9%, or 13 lines) than infected D. simulans lines (16.7%, or 13 lines) achieved the maximum transmission efficiency (100% infected offspring), thes e differences were not significant ( P = 0.335, Figure 3-1). Therefor e, the two species responded similarly to artificial selection for maximum transmission effici ency. However, there was variation within species for transmission efficiency, as in both species, some lines reached maximum transmission while others did not. Overall transmission efficiency was higher in D. melanogaster than in D. simulans lines over the first three generat ions of artificial selection ( P < 0.0001, Figure 3-2). The two species responded diffe rently to selection across the three generations. While D. melanogaster lines exhibited significant increases in infection levels at each generation ( P < 0.0001), statistically significant increases in infection levels were obser ved only from G2 to G3 in D. simulans lines ( P < 0.0001, Figure 3-2 Infection under Relaxed Selection The 26 lines (13 of each species) that attained maximum transmission within three generations of selection cont inued to evolve under relaxed selection for 13 additional generations (for a total of sixteen generations post injection). From G4-G16, transmission efficiency was significantly higher in D. melanogaster than in D. simulans (P = 0.0003, Figure 3-3). As was the case during selection, there wa s genetic variation within species: some lines from each species remained at or near 100% tran smission efficiency, while others fell 71

PAGE 72

to near 0% transmission efficiency. Bo th species experienced decreases in transmission efficiency; however, a significant decline (l e vels significantly lower than G3) occurred in D. simulans at G9 ( P = 0.0105) while in D. melanogaster this significant decline occurred at G15 (P = 0.0167). Thus, transmission efficiency is erratic and decreases over time in both hosts, but decreases earlier and is overall lower in the nov el host than in the native host. Virus Titers Ribonucleic was purified from each of three female flies per infected line from generations six and sixtee n (i.e., G6 and G16), and viral titer was estimated by strand-specific quantitat ive PCR as described in the methods. The variance in viral titer was significantly greater in D. simulans, the novel host, than in D. melanogaster (Levenes test, P = 0.009), and the mean titer was also smaller (Figure 3.4). However, taking into acc ount the unequal variances, the virus titers in the two species were statistically indistinguishable (Welch's ANOVA, P = 0.3820, Figure 3-4). Additionally, there were no differences in copy number for either species between generations ( P = 0.7391). However, interesting trends are apparent from scrutinizing the data (Figures 3-3, 3-4, 35): there are more lines with low titer in D. simulans than in D. melanogaster at generation 6 and the titers in a number of D. melanogaster lines were markedly lower at G16 than at G6. Virus titer and transmission efficiency we re significantly correlated for both species at G6, though the correlation was negative for D. melanogaster ( = 0.5394, P = 0.0289), while it was positive for D. simulans ( = 0.4604, P = 0.0455; Figure 3-6). A significant positive co rrelation was observed between titer and 72

PAGE 73

transmission efficiency for D. m elanogaster at G16 ( = 0.5002, P = 0.0265), but not for D. simulans at G16 (, = 0.3750, P = 0.1042; Figure 3-6). Host Fitness Host fitness (fecundity and hatchability) was examined at G6, G11 and G16 (Table 3-3 and Table 3-4). The data were standardized to the uninfected means to correct for differences in environmental heterogeneity between generations (hereafter referred to as relative fecundity and relative hatchability). The data were not normal, as such, the line means across vials were used in the statistical analysis to eliminate nonindependence. The data were analyzed as described above using nonparametric tests. Relative fecundity was statistically indistinguishable for infected and uninfected D. simulans at every generation tested (Compared to uninfected control mean of 1.0; P 0.2563). Similarly, relative fecundity was statistically indistinguishable for infected and uninfected D. melanogaster at G6 and G11 ( P 0.3703). However, relative fecundity wa s significantly higher for infected G16 D. melanogaster than for the respective uninfected controls ( P = 0.0265). Additionally, relative fecundity at G 16 was significantly higher in infected D. melanogaster than in infected D. simulans at G16 ( P = 0.0009). Although the relative fecundity of infected D. simulans females was statistically indistinguishable across the three generations tested ( P 0.0513), it actually declined across the three generations (Figur e 3-7). Conversely, the fecundity of infected D. melanogaster females was statistically indistinguishable between G6 and G11 ( P = 0.1086) but rebounded at G16 ( P 0.0025). This suggests that the native host ( D. melanogaster ), or virus in the negative host, evolved to reduce 73

PAGE 74

virulence by G16 while the novel host (D. simula ns ) and/or its virus did not. Therefore, the virus is more virulent on the novel host. Relative hatchability was significantly lower for infected than for uninfected D. simulans at G6 ( P = 0.0313, Figure 3-8) but statistically indistinguishable ( P 0.0528, Figure 3-8) in subsequent generations of these flies. In D. melanogaster relative hatchability was significantly lowe r for infected than for uninfected flies at G11 ( P = 0.0307, Figure 3-8) but st atistically indistinguishable ( P 0.6761, Figure 3-8) in the other two generations of these flies. Relative hatchability at G6 was significantly higher in infected D. melanogaster than in infected D. simulans ( P = 0.0313, Figure 3-8) but there were not statistical differences between these two groups in subsequent generations ( P = 0.9783, Figure 3-8). Finally, relative fecundity was significantly higher in infected D. simulans at G6 than at G16 ( P = 0.0455, Figure 3-8) while there were not differences in these two generations for D. melanogaster (P = 0.935, Figure 3-8). Inte restingly, hatchability was associated with virus titer. The eggs from flies with high virus titers hatched at less frequently than the eggs from flies with low virus titers ( P = 0.0442, Figure 39).Thus, the virus is virulent in both spec ies but it is more virulent in the novel host. Discussion Artificial Host-Shift and Selection for Transmission The novel host ( D. simulans ) was easier to infect than the native host ( D. melanogaster ). The significant ease with which the DMelSV virus infected the novel host is certainly consistent with its ability to infect other Drosophila species ( L'HERITIER 1958 ; PIZZATTO et al. 2010 ). However, it seems to be m ore difficult to 74

PAGE 75

select for maximum trans mission efficiency in the novel host; and moreover, transmission efficiency declined earlier and to a greater extent in the novel hos t than the native host following relaxation of selection. These trends are analogous to disproportionately higher dead-end, or sink, infections (i.e. the newly infected host is incapable of transmitting the pathogen) in the novel host; and may have occurred because of lower within-host reproduction, reduced between host transmission or both ( BULLER et al. 2005 ; DENNEHY et al. 2006 ; MATROSOVICH et al. 1999 ; ROKYTA et al. 2005 ). Because DMelSV virus is vertically transmitted, virus-host interactions at the cellular level ar e critical for both withinand between-host transmission, and this intera ction likely acted as a barrier to successful colonization of some of the novel host lineages ( DENNEHY et al. 2007 ; PERLMAN and JAENIKE 2003 ). Howev er, another possibly comp lementary explanation for lower transmission efficiency in the novel host could be that DMelSV was more virulent in the novel host than the native host. Hatchability was significantly lower in infected than in uninfected D. simulans during the early stages of the infection. The failure of infected eggs to hatch would certainly have decreased transmission efficiency. However, virul ence also decreased with time in some (but not all) lines. The result was a si gnificantly different transmission trajectory for virus between native and novel host li nes. It appears that, as with influenza and other host-shifted parasites ( FRANK and SCHMID-HEMPEL 2008 ), the virulence of DMelSV attenuated wit h continual transmission in the novel host. 75

PAGE 76

Transmission under Relaxed Selection Under relaxed selection, transmissi on efficiency decreased faster in the novel host than in the native host. Unlike in fluenza, DMelSV is unipartite, so does not reassort, and there is no evidence that DMelSV recombines. As with other systems, it is likely that the number of virions transmitted between generations is low, leading to repeated bottlenecks ( LAWRENCE and MATOS 2005 ). These are simple barriers to transmission that ex ist in both hosts. However, this combination of challenges caused a si gnificant decrease in the transmission efficiency of the DMel SV earlier in the nov el host than in the native host (Figure 3-3). Given that the success of parasites results from the interaction of the pathogen's genetics with its host, this result is perhaps unsurprising as it confirms that the virus is well-adapted to its native host. Additionally, this suggests that the predicted mutational capacity of RNA viruses in general, and DMelSV in particular, is insufficient to fac ilitate quick adaptation to the novel host ( MATROSOVICH et al. 1999 ; SAMUEL 2011 ). Given that a genetically variable virus is expected to be more adaptable ( ANDRE and HOCHBERG 2005 ; MOYA et al. 2000 ), by extension a more genetically robust virus should be less successful in an adaptiv e walk. DMelSV was extremely su ccessful when aided by artificial selection, but experienced a drastic dec rease in transmission efficiency once selection was relaxed and natural forces took over. Others have noted that DMelSV seems to have a low mutation rate ( BRUSINI et al. 2012 in review ; CARPENTER et al. 2009 ). 76

PAGE 77

Virus Titers Virus titers were highest in the na tive host, while variance in titer was higher in the novel host (Figure 3-4). Ou r results are in agreement with previous work showing that infection on t he novel host can be inconsistent ( ANTONOVICS et al. 2002 ). Moreover, once selection as relaxed following G4, the native host lines maintained high virus titers two generati ons later, while only 58% of the novel host lines did so (Figure 3-5). This again supports the notion that the cellular barriers to within-host rep lication and between-host tran smission are stronger in the novel host ( PERLMAN and JAENIKE 2003 ). It is likely that suppressed virus replication within the novel host resulted in decreased within and between hos t transmission, and led to the eventual crash of the virus population in several of the novel host lines by G6 and in most of the novel host lines by G16. This is in accord with the positive correlation betw een titer and transmission efficiency that was observed here (Figure 36) and in other systems ( EWALD 1983 ; FRANK 1996 ; LENSKI and MAY 1994 ). The high titers (high replication rates) suggest that the vi rus is more fit on the native host than on the novel host and t hat the lines with high titers have virus that successfully adapted to the host. However, DMelSV is vertically transmitted. As such, this virus s hould suffer the classic tradeoff between virulence and transmission ( LIPSITCH et al. 1996 ; SHARON et al. 1999 ), but it does not, because higher titers within the hos t likely ensure increased transmission. Virulence The virulence of DMelSV was measured as a reduction in host fecundity and egg hatchability. Although the virus is virul ent in both species, as predicted, it 77

PAGE 78

is more virulent on the novel hos t, and it is more virulent in hosts with higher virus titers ( EBERT 1998 ). Relative fecundity rebounds in the native host, despite high titers, indicating a reduction in virulenc e on the native host. In contrast, titer and fecundity both seem to decrease slightly wit h time in the novel host, suggesting continued virulence. Hatchability was lowe r on the native host at G6 but any species differences disappear subsequently indicating that the virus is adapting to the novel host. Interestingly, the fly lines with high virus tite rs also exhibited decreased hatchability r egardless of species (Figure 3-9). However, hatchability was lower in the early generations of the novel host but rebounded by the end of the experiment. This shows that the virus adapted to both host but the adaptive walk took longer on the novel host. In c onclusion, our data are consistent with predictions from serial passage experiments indicating that a parasite will be more virulent on a novel host. 78

PAGE 79

Table 3-1. Sigma virus titers (me an standard error) i n females of Drosophila melanogaster and D. simulans at six and 16 generations after artificial infection by injection. G6 G16 D. melanogaster 1.3 x 106 2.0 x 1051.1 x 106 1.7 x 105 D. simulans 1.0 x 106 2.6 x 1051.0 x 106 3.1 x 105 Titer is viral genomes per microgram of fly RNA. Table 3-2. Mean sigma virus titers across species, line and generation (data presented in Figure 3-4). Generation Species Line Six Sixteen DMelSV-207 599,984 555 DMelSV-210 2,215,338 578 DMelSV-211 Drosophila melanogaster 1,563,223 551 DMelSV-216 1,508,642 1,789,897 DMelSV-235 976,914 843,461 DMelSV-238 474,685 749,393 DMelSV-302 964,328 2,668,256 DMelSV-332 5,732,529 1,750,351 DMelSV-339 2,747,062 1,511,670 DMelSV-343 805,400 1,711,248 DMelSV-349 1,012,250 2,474,306 DSimSV-36 1,440,993 1,988 Titer is viral genomes per microgram of fly RNA. DSimSV-52 1,506,997 3,544,808 DSimSV-60 2,621 1,039 DSimSV-61 3,120 173 DSimSV-68 1,823 851 DSimSV-70 580,547 118 DSimSV-91 1,979,920 5,180,157 DSimSV-138 125 694 DSimSV-147 426 344 DSimSV-151 3,101,075 18,143 DSimSV-155 2,367,946 1,774,932 Drosophila simulans DSimSV-177 804,883 1,767,926 79

PAGE 80

Table 3-3. Fecundity data (eggs/fema le/day) for sigma virus-infected and uninfected Drosophila melanogas ter (native host) and D. simulans (novel host). G6 G11 G16 Line Ave StDev Ave StDev Ave StDev 207 3.66 2.57 6.44 5.02 0.93 0.67 210 3.38 4.00 3.94 2.46 3.71 3.70 211 1.97 1.79 10.81 6.03 2.63 1.73 216 4.88 2.32 7.13 4.17 3.31 1.46 217 5.56 3.06 6.63 3.04 4.69 3.82 225 2.69 4.11 ----235 2.19 1.44 1.81 2.43 2.31 1.91 238 4.22 2.06 4.56 2.11 3.13 2.62 302 2.19 2.62 9.50 6.26 1.57 1.34 332 5.47 2.65 4.75 2.12 1.63 1.55 339 2.28 2.47 7.63 3.31 1.31 2.05 343 3.41 1.60 2.75 1.34 3.31 1.83 Infected 349 4.882.638.193.42 1.861.89 201c 3.25 2.22 6.88 3.06 0.88 1.30 204c 3.28 2.71 8.00 2.93 0.06 0.18 205c 5.19 2.89 5.56 3.23 2.17 1.04 211c 1.81 1.47 9.38 4.19 0.93 1.13 Drosophila melanogaster Uninfected 214c 2.912.287.194.81 0.641.07 31 8.53 2.55 1.00 0.65 6.88 2.53 36 8.31 6.51 4.25 2.19 5.19 2.52 52 5.03 2.54 3.50 3.05 1.75 1.16 60 6.16 2.61 2.88 2.34 4.81 1.67 61 6.28 2.42 2.44 1.35 3.69 1.58 68 4.72 3.03 5.63 3.70 2.88 1.06 70 0.78 0.59 0.75 1.25 1.86 1.03 91 3.53 1.91 7.44 4.38 3.31 2.09 138 7.97 3.38 5.38 2.85 3.93 1.06 147 15.94 26.11 1.44 0.86 4.57 1.30 151 4.47 1.22 5.38 2.88 2.25 0.71 155 6.25 3.71 4.00 3.60 4.21 1.38 Infected 177 5.881.9411.257.03 5.002.12 2c 6.56 3.56 3.56 5.25 5.19 4.54 3c 5.78 2.74 2.74 3.63 3.69 1.62 4c 2.13 1.99 1.99 4.63 5.88 2.62 5c 4.34 2.83 2.83 1.63 1.50 1.16 6c 3.03 1.92 1.92 2.44 3.71 1.75 8c 5.81 2.31 2.31 2.56 4.64 0.63 9c 6.25 2.01 2.01 5.38 6.38 2.83 12c 6.81 2.16 2.16 1.94 3.88 1.06 Drosophila simulans Uninfected 121c 3.912.562.564.75 3.562.80 80

PAGE 81

Table 3-4. Hatchabilit y data (% of eggs hatching) for sigma virus-infected and uninfected Drosophila melanogaster (native host) and D. simulans (novel host). G6 G11 G16 Line Ave StDev Ave StDev Ave StDev 207 81.0 22.5 91.5 11.2 100.0 0.0 210 55.2 39.8 85.0 14.0 78.1 25.6 211 76.7 23.5 76.1 28.9 79.7 25.0 216 77.0 19.9 79.8 18.8 68.3 18.5 217 71.7 23.6 82.6 15.6 69.7 30.9 225 70.4 25.7 ----235 87.8 12.7 73.5 15.3 64.0 29.7 238 66.2 29.1 71.1 28.8 86.7 16.3 302 76.3 29.7 68.7 16.9 93.5 11.6 332 64.9 35.0 71.7 16.8 72.9 28.4 339 71.8 17.4 62.6 24.1 80.1 29.7 343 70.4 27.6 91.0 10.7 54 33.4 Infected 349 52.832.691.511.2 1000.00 201c 71.8 33.2 84.4 14.3 86.1 22.2 204c 79.3 19.6 95.6 6.8 100.0 0.0 205c 74.9 20.0 87.3 15.9 87.8 10.7 211c 64.2 29.8 88.7 10.6 69.4 29.2 Drosophila melanogaster Uninfected 214c 70.928.386.718.2 66.711.8 31 60.5 22.1 94.4 16.7 56.1 30.4 36 50.1 28.8 92.0 13.5 63.0 12.0 52 77.0 17.9 69.3 21.2 76.9 23.5 60 72.3 20.6 71.6 21.3 63.9 36.4 61 70.3 22.2 65.2 25.9 84.1 19.8 68 55.5 33.5 79.4 21.3 63.9 34.9 70 72.3 25.0 66.7 31.2 72.5 30.9 91 55.8 25.2 69.4 23.6 78.5 19.9 138 71.7 22.4 76.2 18.9 64.5 38.2 147 62.4 30.5 86.4 23.4 72.5 34.3 151 58.3 24.7 95.4 10.1 60.7 31.7 155 60.1 24.9 70.8 18.3 62.0 27.7 Infected 177 52.024.257.623.6 54.940.1 2c 74.0 26.2 85.9 15.9 65.9 32.1 3c 78.5 18.9 85.5 14.1 52.5 32.0 4c 61.4 31.7 76.4 19.8 65.8 26.0 5c 75.9 29.6 95.5 10.1 60.5 29.3 6c 62.0 35.0 89.5 21.3 77.6 26.4 8c 88.6 17.5 79.9 21.3 71.5 31.7 9c 65.6 31.8 82.1 20.3 58.4 25.4 12c 67.6 34.0 86.9 19.0 90.6 18.6 Drosophila simulans Uninfected 121c 83.920.893.19.1 70.926.8 81

PAGE 82

InjectedInfectedCompleteNumber of lines 0 100 200 300 400 500 Figure 3-1. The novel host was easier to infect, but harder to stabilize. Total Drosophila melanogaster (grey bars) and D. simulans (white bars) females that were injected with sigm a virus inoculum (leftmost bars), subsequently produced infected offspring (middle bars), and ultimately produced lines with 100% infected offspring (maximum transmission efficiency, rightmost bars). Significantly more D. simulans became infected than D. melanogaster ( P = 0.0015); however, there was no significant difference in the proporti on of infected flies that produced stable lines between the two species ( P = 0.1758). 82

PAGE 83

Generation 0123Sigma-infected adults (%) 0 20 40 60 80 100 Figure 3-2. Maximum transmission effi ciency (100%) was harder to achieve on the novel host. Per cent of Drosophila melanogaster (native host, N = 45 lines, circles) and D. simulans (novel host, N = 78 lines, triangles) adults infected with DMelSV followi ng injection (G0) and during selection for transmission rate (G1 to G3). Transmission efficiency was higher in the native host than the novel host ( P < 0.0001). Significant increase occurred between every generation in the native host ( P < 0.0001) but only from G2 to G3 in the novel host ( P < 0.0001). Error bars are two standard errors among lines. 83

PAGE 84

Generation ( Drosophila melanogaster) 0123678910111213141516Infected adults (%) 0 20 40 60 80 100 Generation ( Drosophila simulans) 0123678910111213141516 Figure 3-3. Under relax ed selection transmission efficiency decreased more rapidly on the novel host. Once selecti on was relaxed (after G3), the native host, Drosophila melanogaster (left panel), maintained higher average transmission efficiency (thick line with circles) than the native host ( D. simulans thick line with triangles) ( P < 0.0003). Individual lines (thin do tted lines) were highly variable within each species. Significant decreases (from 100% at G3) occurred by G9 in D. simulans ( P < 0.0105) and by G15 in D. melanogaster ( P < 0.0167). Bars are two standard deviations (based on among line variance). 84

PAGE 85

Generation 616616Sigma virus genome copies per g of fly RNA 0 2e+6 4e+6 6e+6 8e+6 Figure 3-4. Virus titers were statisticall y indistinguishable in both species, but the means are lower and the variance is higher in females of Drosophila simulans (solid gray) than of D. melanogaster (hatched bars). The data failed Levenes homogeneity of variance test ( P = 0.009); differences in titer ( P = 0.3820) were tested using a Welch's analysis of variance. The summary of virus titers data are presented in Table 3-1. 85

PAGE 86

86 Dm G6Dm G16Ds G6Ds G16Sigma virus genome copi es per g of fly RNA 1e+2 1e+3 1e+4 1e+5 1e+6 1e+7 Figure 3-5. Heterogeneity in titer over ti me between lines for virus load. Titer in Drosophila melanogaster (Dm ) and D. simulans (Ds ) adults from unique lines was quantified at six and si xteen generations after artificial infection by injection. Raw dat a are presented in Table 3-2.

PAGE 87

Infected adult flies (%) 20 40 60 80 100 Sigma genome copies/g of Drosophila melanogaster RNA 01e+62e+63e+64e+65e+6Infected adult flies (%) 20 40 60 80 100 Sigma genome copies/g of Drosophila simulans RNA 01e+62e+63e+64e+65e+6 A D B C Figure 3-6. Virus titer is significantly correlated with transmission efficiency for bot h hosts, but the sign of the correlatio n changes with time. Kendalls tau correlation of DMelSV load and percent infected adults for Drosophila melanogaster (circles) and D. simulans (triangles) lines. A: D. melanogaster at G6, = -0.5394, P = 0.0289; B: D. simulans at G6, = 0.4604, P = 0.0455; C: D. melanogaster at G16, = 0.5002, P = 0.0265; D: D. simulans at G16, = 0.3750, P = 0.1042. 87

PAGE 88

Generation 61116Eggs/Female/Day Standardized to control mean 0.0 0.5 1.0 1.5 2.0 2.5 3.0 Figure 3-7 Fecundity was not negatively im pacted by infection with sigma virus. Fecundity standardized to the relev ant uninfected mean. Fecundity in the native host, Drosophila melanogaster (circles), rebounds in generation 16, while fecundity in the novel host, D. simulans (triangles), does not. Bars are two standard errors. Fecundity data are presented in Table 3-3. 88

PAGE 89

Generation 61116Hatchability Standardized to control mean 0.7 0.8 0.9 1.0 1.1 Figure 3-8. Hatchability was negatively impa cted by infection with sigma virus in both hosts but at different generati ons. Hatchability standardized to the relevant uninfected mean. Hatchability in the native host, Drosophila melanogaster (circles), is higher in generat ion 6 than in the novel host, D. simulans (triangles). Hatchability tends to increase with time in D. simulans but the trend is not significan t. Bars are two standard errors. Hatchability data are pres ented in Table 3-4. 89

PAGE 90

Generation D. simulans 61116 Generation D. melanogaster 61116Hatchability standardized to control mean 0.6 0.7 0.8 0.9 1.0 1.1 1.2 Figure 3-9. Hatchability standardized to the relevant uninfected mean for fly lines with high (large symbols) and low (s mall symbols) virus titers in Drosophila melanogaster and D. simulans On average, eggs from high lines hatch at significantly lower levels than eggs from low lines ( P = 0.0442), regardless of species. Bars are two standard errors. 90

PAGE 91

CHA PTER 4 SEQUENCING AND ANALYSIS OF VIRAL GENOMES PASSAGED IN DROSOPHILA LINEAGES Background Emerging infectious diseases succes sfully colonize novel hosts because they can adapt to those hosts ( LENDERBERGER ET AL. 1992 ). Therefore, understandi ng how parasites will evolve once they ent er a novel host is of particular importance, particularly because adaptability has been identified as one of six major factors that can contri bute to successful host shifts ( LENDERBERGER ET AL. 1992 ). Most of the pathogens that we know are s hifting successfully into novel hosts are RNA viruses ( HOLMES and RAMBAUT 2004 ); therefore, it is necessary to gain a bett er understanding of how these types of viruse s adapt to their novel host following a host shift. Some general patterns of RNA viru s evolution are known; RNA viruses tend to exhibit extensive co nvergent evolution with increased time spent in a novel host species ( BULL et al. 1997 ; CUEVAS et al. 2002 ). This occurs because benef icial mutations are usually fixed early and contribute the most to viral fitness in the new environment ( BULL et al. 1997 ; WAHL and KRAKAUER 2000 ). A dditionally, muta tion rate is constant and high, while genetic diversity is low during the course of adaptation and increases once fitness plateaus ( WICHMAN et al. 1999a ). Finally, mutations that are favor able in the nove l host usually decrease fitness in the original host ( BULL et al. 1997 ). Most of these observations were made in seri al passage experiment studies (SPEs) ( EBERT 1998 ); therefore, little is k nown about viral evolution in multicellular eukaryotes We will examine these general ities in the DM/D S/sigma system by sequencing the 91

PAGE 92

major genomi c variant from sigma virus st rains passaged naturally in the novel host D. simulans for sixteen generations. We have developed the sigma rhabdov irus and Drosophila spp. as a model system in which we can better understand viral adaptation following host shifts. The sigma rhabdovirus is a good model pathogen because like many other viruses of interest it has an RNA genome (ssRNA). An ancestral sigma rhabdovirus inoculum was generated from wild-caught D. melanogaster females collected in Athens, GA (USA). This virus was artificially shifted by injection onto novel hosts ( D. simulans ). Additionally, the virus was mock shifted onto its native host (D. melanogaster ) to control for any effe cts of procedures or the environment. Subsequently the virus wa s passaged naturally through sixteen host generations (Chapter 3). Here we test whether the shifted virus experienced convergence on the novel host. Materials and Methods Ribonucleic Acid (RNA) Purification Ribonucleic acid from each inoc ulum was isolated using TRIzol (www.invitrogen.com), following the manuf acturers directions, as described previously (Chapter 3). Briefly, a 50 l aliquot of inoculum was mixed with 200 l of TRIzol on ice. The mixture was vortexed briefly and incubated at room temperature (RT) for 5 min. Each samp le then received 40 l of chloroform and was vortexed for five seconds, incubated at RT for 2-3 min, then centrifuged at 12,000g for 10 min at 2C. The RNA in the aqueous phase was precipitated with isopropanol at -20C for 10 min. The precip itate was pelleted by centrifugation at 92

PAGE 93

12,000g for 10 min at 2C. The pellet was washed with cold (4C) ethanol, air dried for five min at RT, resuspended in 29 l of DEPC-treated water with 1 l of 40 M RNAse inhibitor (www.neb.com) and incubated at 55C for 10 min to aid resuspension. Samples were stored at -80 C until used. Reverse Transcription Reverse transcription of the DMelSV virus genome was performed using the SuperScript III First-Strand cDNA synthesis system ( www.invitrogen.com ) following the manufacturers recommendati ons. Briefly, 200ng of inoculum RNA were mixed with 1 l of primer (w hole1F: 5 TAGAAGCATCCTCGGCTTTC 3) and 1 l of 10mM dNTP mix and the reaction volume was brought to 10 l with DEPC-treated water. This reaction (part 1) was incubated at 60C for 5 min and then placed on ice. A cDNA synt hesis mix (2 l of 10X RT buffer, 4 l of 25mM MgCl2, 2 l of 0.1mM DTT, 1 l of 40U / l RNAseOUT and 1 l of 200U/ l SuperScript III enzyme, part 2) was prepared and added to part 1 and the mixture was incubated at 50C for 50 mi n, then at 85C for 5 min to stop the reaction and then chilled on ice. RNAse H (1 l) was added to the tube and the synthesis reaction was incubated at 37C for 20 min. The cDNA was stored at 20C. Viral Genome Sequencing The genome of the major viral variant of the ancestral virus (DMelSV-A) and of the virus evolved in D. simulans (DSimSV-E) were sequenced in 19 overlapping fragments that were produced using 19 primer pairs (Table 4-1). All amplicons were amplified using the fo llowing PCR parameters: 94C for 2 min followed by 40 cycles of 94C for 30 seconds, 55C for 60 seconds and 72C for 93

PAGE 94

60 seconds ; a final elongation step of 5 min at 72C was done after the cycling. The amplicons were electrophoresed to check for quality and sent for sequencing by Genewiz (www.genewiz.com) using their standard protocols. Each amplicon was sequence with the forward and reve rse primer. The sequenced amplicons were assembled and compared using the Geneious 5.6 software package (www.geneious.com). Final genomes were assembled to a 99% quality score. Further details on the materials and methods used here are presented in Appendix B. Statistics All statistical tests were implemented in MEGA 5.0 (TAMURA et al. 2011) unless otherwise noted in the text. The hypothesis of positive selection (HA: dN > dS) was tested for each of the six viral genes for all lines using the Z-test of selection. The bootstrap method (1000 replications) was used with a synnonsynonymous substitution setting using the Nei-Gojobori method (JukesCantor). Finally, the transition/transversion bias was determined using a NJ (neighbor joining) tree exam ining nucleotides using the Kimura 2-parameter model. A phylogenetic tree of the tw elve complete viral genomes was constructed using the maximum likeli hood method with a Jukes-Cantor model, and robustness of the nodes was evaluat ed using the bootstrap method (1000 replications). Results In total, twelve viral genomes we re sequenced: the ancestral genome DMelSV-A, five genomes evolved in D. simulans (DSimSV-E) and six genomes evolved in D. melanogaster (DMelSV-E). The sequence polymorphisms for all of 94

PAGE 95

the eleven derived lines sequenced are s hown in Table 4-2 and these data are summarized in Table 4-3. Three DSimSV-E lines had genomes carrying many differences when compared to th e ancestral genome (10.73 6.74 substitutions/Kb); predictably, only 1.82 1.38 per Kb were non-synonymous substitutions. Two DSimSV-E lines had low numbers of synonymous and nonsynonymous substitutions (0.153 0.308 and 0.0267 0.0923 substitutions/Kb, respectively) for the duration of the experiment. Finally, the DMelSV-E lines as a group had the lowest numbers of synonymous and non-synonymous substitutions (0.0.277 0.619 and 0.0267 0. 092 substitutions/Kb, respectively). For all the lines, synonymous changes outnumbered non-synonymous ones, as expected. Indeed, statistical analyses ( Z test as described in Materials and Methods) found no evidence s upporting positive selection ( P = 1.0 for each of the genes). This is not surprising given the relative scarcity of nonsynonymous changes. Finally, five of the six genes had a slight transition bias. Gene M was the exception, with a very lar ge bias (Table 4-4). The eleven lines sequenced carry 10 polymorphisms in the M gene: five G-to-A, three U-to-C and two C-to-U, all transitions. The full sequenc e of the DMelSV-A is presented as a reference in Table 4-5. In the phylogenetic analysis (Figure 41) the three DSim SV-E lines that had many polymorphisms grouped together on their own branch with 100% bootstrap support. The ancestral virus s hared a node with a branch containing all of other sequences (six DMelSV and two DS imSV) that are extr emely similar to 95

PAGE 96

the ancestor. Interesti ngly, the two DS imSV lines grouped on different branches within this cluster (95% bootstrap s upport for both of these branches). Discussion It is clear that the ancestral virus wa s well adapted to its native host, as the virus changed little during 16 generat ions on naive native host lines; however, beyond that simple conclusion, the data are perplexing. Attempting to correlate the sequence data to transmission efficiency, viral copy number, and virulence in the three lines harboring t he genetically different evolved virus did not yield any explanatory patterns (even if not significant). The three DSimSV-E virus lines that had very different genetics are neither more fit (higher titer or greater transmissibility, consistent wit h an adaptive change) nor less fit (lower titer or lower transmissibility). Additiona lly these viruses are not more or less virulent than the other two vi ruses that were very sim ilar to the ancestral virus genome. The presence of two major cl asses (many substitutions vs. few substitutions) of virus passaged in the nov el host suggests either a great deal of stochasticity such that different strategi es were employed or multiple challenges were met. One possibility is that upon infect ion, some virions (in the low variants) encountered by chance favorable cell popu lations and were able to immediately establish themselves as the major variant in that fly lineage. This scenario is equally likely for the low SNP and high SN P variants, assuming that the genetic variants present in the three distinct, fast evolving DSimSV-E lines were low variants in the inoculum but also we re viable and replication competent upon being injected into the novel host. Although sigma is less genetically diverse that one might expect for an RNA virus ( BRUSINI et al. 2012 in review ; CARPENTER et 96

PAGE 97

al. 2009 ), we showed in Chapter 2 that the wild population from w hich the ancestral inoculum was generated still ca rried a significant amount of genetic diversity. Moya et al. ( 2000 ) demonstrated that bottlenecks reduce viral fitness and, as seen in Chapt er 3, many of t he artificially generated infections failed because viral fitness was low (low transmissi on). However, rare variants that may be maladapted to their native host may have had a fitness advantage over the common variants and, as such, were more fit in the novel host (relative to the common variant) and persisted. This s eems unlikely, unless the model of adaptation underlying the Z test, which emphasizes the role of amino acid substitutions rather than synonymous c hanges, is not appropriate for virus evolution. Indeed, silent changes may be very important as they may lead to codon deoptimization and thus reduced viru s fitness at a level that was not detectable here (MUELLER et al. 2006 ). An alternative explanation which cannot be falsified with our current data i s that the mutation ra te has increased in these three lines. More directly evaluating t hese hypotheses would require sequencing viral variants collected in generations of host immediately following the host shift to determine whether these polymorphic va riants establish themselves soon after the host shift or evolve over time. Given that the two variants with few polymorphisms appeared to have equivalent fit ness (i.e. transmissi on and titer) to the three fast evolving strains, it seems more likely that stochastic events such as a bottleneck produced by the host shift fa cilitated the rapid es tablishment of a completely new major variant within t he novel host. Similarly, any mutational increase was not adaptive per se. In ei ther case, bottleneck or mutational 97

PAGE 98

speedup, evolution did not occur by small steps ( BURCH and CHAO 1999 ; FISHER 1930 ). 98

PAGE 99

99 Table 4-1. Sequences for 19 primer pairs using to produce overlapping amplicons from viral complementary DNA (BRUSINI et al. 2012 in review). Forward primer ID and sequence Re verse primer ID and sequence 44TAGAAGCATCCTCGGCTTTC 976+ CTTCCCTGTCTTTGCACCTC 679ACACACTTGGTCGGAAGGAG 1638+ CAAATTGCAACTGGCTCTCA 1323TGTCCACAAAGTATTTCCATGTC 2273+ TGAGTAGGCTGGTTCAAAACG 1964TTCTGCCATATGTCCTTTGC 2916+ TTGCATCGCTTCAGTTGTTC 2640AATAAGGTGCAGGGCTTTCC 3608+ GGTTCCTCATGCATCTGACA 3302TAAACCACTTGCTGCCATGA 4275+ GGGAAGACTTGTGTCATCCAT 3989AGTCGGCTGAATGATTGTCC 4894+ CATGGCAGATGACCTCACAA 4610TTCATCAGATTGCCCTCTCC 5482+ GACGACGGTGGAGAAAGAGA 5262AAGGTAATGTGTGCCATCCA 6164+ AGATCACCTTGCCAGAGGAA 5925GGTCCAGCATCCTGTTGTG 6852+ TCAATTGTCGCACTGGTTGT 6559TGTGTCCATAGAGAACGGATCA 7522+ CCATCCTAACTCTGTGACATTCC 7209TCTCCTCCGTGTTCAGTGTT 8173+ GTCATACGACCCCAAGAAGC 7885GTACGCCTTCGTGTTCACTG 8799+ TTCAAGCCTATTTTATTGTTCGTG 8535CCTCAATGAAGGACTCGATAGG 9444+ GCTGCAAATGACACTGGATG 9181TCGTCCATGTAATGTATCTTCACTG 10105+ CACCGTGCTTTACCATGAAA 9842GGCCCTCTTGGAAGTGCT 10762+ TGGACCAGGGTAATGAGAGC 10516GGGGTGCTCCACATCACTAT 11431+ TCACCGGGAAGAAGATATGG 11167GAGCTGTCACACCCGAGTTA 12028+ GATCCTCGCTCCATTTTCAA 11848TAATCCGATGGTTCCAAAGC 12498+ GACACCAAGACCGAGTTCGT

PAGE 100

Table 4-2. Sigma virus strains evolved in Dr osophila simulans (DSimSV-E) or D. melanogaster (DMelSV-E) were compared to the ancestral virus (DMelSV-A). New bases are given for each of the viral strains when a single nucleotide polymorphism was detected. The new amino acid is given wh enever the SNP was non-synonymous. Gene N N N N N N N N N N N N N N NC NC Codon Number 10 81 88 93 114 133 168 222 228 229 249 250 270 341 DMelSV-A codon GCU UUA AUCAACAAGUCUGGAAAUGAU GUG UCUGCCCCAACC A C SNP codon position 3 1 1 3 3 1 3 1 1 1 3 3 3 3 Consensus AA A L I N K S G N D V S A P T DSimSV-E-52 C A G G C U G G U New AA D DSimSV-E-91 New AA DSimSV-E-151 C A G G C U G G U New AA D DSimSV-E-155 C C G U A G G C U G U G U New AA V DSimSV-E-177 New AA DMelSV-E-216 New AA DMelSV-E-238 New AA DMelSV-E-302 New AA DMelSV-E-332 A A New AA N M DMelSV-E-343 New AA DMelSV-E-349 New AA 100

PAGE 101

Table 4-2. Continued. Gene P P P P P P NC NC NC X X X X X X Codon Number 479 481 482 550 636 695 773 780 882 893 905 914 DMelSV-A codon GGU GCG UAUGCACCCAAC G C G AUUAGAAGCAUCCCUACG SNP codon position 3 3 1 1 3 3 1 3 1 2 3 2 Consensus AA G A Y A P N I R S I P T DSimSV-E-52 C A C U U A U A G G G C C New AA H S V G T DSimSV-E-91 New AA DSimSV-E-151 C A C U U A U A G G G C C New AA H S V G T DSimSV-E-155 C A C U U A U G G G C C U New AA H S V G T M DSimSV-E-177 New AA DMelSV-E-216 New AA DMelSV-E-238 New AA DMelSV-E-302 New AA DMelSV-E-332 New AA DMelSV-E-343 New AA DMelSV-E-349 New AA 101

PAGE 102

Table 4-2. Continued. Gene X X X X X X X M M M M M NC G Codon Number 929 967 976 1014 1022 1024 1058 1120 1167 1207 1242 1292 1340 DMelSV-A codon CGC CGA ACUGUGUAU AUU ACA UUA ACU UAC GAGCAG C CAU SNP codon position 3 1 3 3 1 3 3 1 3 3 3 3 3 Consensus AA R R T V Y I T L T Y E Q H DSimSV-E-52 A A C C A A G New AA Q DSimSV-E-91 New AA DSimSV-E-151 A A C C A A G New AA Q DSimSV-E-155 U A A A C C C U A G New AA H Q DSimSV-E-177 New AA DMelSV-E-216 New AA DMelSV-E-238 U New AA DMelSV-E-302 C New AA DMelSV-E-332 New AA DMelSV-E-343 New AA DMelSV-E-349 New AA 102

PAGE 103

Table 4-2. Continued. Gene G G G G G G G G G G G G G G G Codon Number 1355 1358 1376 1384 1388 1390 1416 1421 1442 1449 1467 1510 1543 1570 1606 DMelSV-A codon CUG AUC GUA AGU GUGAAC GUUCUA GAU CCU GAC GAU AAG CGGCAU SNP codon position 1 3 3 3 3 3 1 1 3 3 3 3 3 3 3 Consensus AA L I V S V N V L D P D D K R H DSimSV-E-52 G C A U U C A U A A C New AA DSimSV-E-91 New AA DSimSV-E-151 G C A U U C A U A A C New AA DSimSV-E-155 U A G C A U A U C A U A A C New AA I DSimSV-E-177 C New AA DMelSV-E-216 New AA DMelSV-E-238 New AA DMelSV-E-302 New AA DMelSV-E-332 C New AA DMelSV-E-343 New AA DMelSV-E-349 C New AA 103

PAGE 104

Table 4-2. Continued. Gene G G G G G G G G G G G NC L L Codon Number 1654 1656 1728 1759 1769 1819 1852 1859 1859 1877 1895 1952 1997 DMelSV-A codon CUU GUC AAC GAGUCU GCA AUC UUU UUU AGGUCC C GAGACG SNP codon position 3 3 3 2 3 3 3 1 3 2 3 2 3 Consensus AA L V N E S A I F F R S E T DSimSV-E-52 C U U C G U C C A A U A New AA LLK DSimSV-E-91 New AA DSimSV-E-151 C U U C G U C C A A U A New AA LLK DSimSV-E-155 C U U G C U C C A A U G A New AA G LLK G DSimSV-E-177 New AA DMelSV-E-216 New AA DMelSV-E-238 New AA DMelSV-E-302 New AA DMelSV-E-332 New AA DMelSV-E-343 New AA DMelSV-E-349 New AA 104

PAGE 105

Table 4-2. Continued. Gene L L L L L L L L L L L L L L L Codon Number 2021 2025 2036 2039 2040 2042 2074 2087 2104 2118 2130 2132 2193 2198 2211 DMelSV-A codon GGA CUA AAA CUU GAA GCUCUGGAA ACC CCC UUC CCC AUA UUGUUG SNP codon position 3 1 3 3 3 3 1 3 3 3 3 2 1 1 1 Consensus AA G L K L E A L E T P F P I L L DSimSV-E-52 C U G G A U G U A U U C C C New AA LL DSimSV-E-91 New AA DSimSV-E-151 C U G G A U G U A U U C C C New AA LL DSimSV-E-155 C U G C G A U G U A U C C C New AA LL DSimSV-E-177 New AA DMelSV-E-216 New AA DMelSV-E-238 New AA DMelSV-E-302 New AA DMelSV-E-332 New AA DMelSV-E-343 New AA DMelSV-E-349 New AA 105

PAGE 106

Table 4-2. Continued. Gene L L L L L L L L L L L L L L L Codon Number 2237 2291 2334 2336 2377 2415 2436 2448 2455 2488 2507 2555 2601 2622 2626 DMelSV-A codon UUC GGG GUUCCU GAC CGGGUGAUC AGA CCGUUGCCGAAG AGA ACU SNP codon position 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 Consensus AA F G V P D R V I R P L P K R T DSimSV-E-52 U C A U A A A G U A A G C New AA DSimSV-E-91 A New AA DSimSV-E-151 U C A U A A A G U A A G C New AA DSimSV-E-155 U U A U A G U A A G C New AA DSimSV-E-177 New AA DMelSV-E-216 New AA DMelSV-E-238 New AA DMelSV-E-302 New AA DMelSV-E-332 New AA DMelSV-E-343 New AA DMelSV-E-349 New AA 106

PAGE 107

Table 4-2. Continued. Gene L L L L L L L L L L L L L L L Codon Number 2635 2654 2738 2750 2812 2835 2855 2871 2886 2923 2945 2951 2957 2988 2989 DMelSV-A codon GCC CAG ACG UCU GCA UUC UUU CUU GCU GGAGGAGCUUUA AUU AGA SNP codon position 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 Consensus AA A Q T S A F F L A G G A L I R DSimSV-E-52 A A A C C U G A G C G C G New AA DSimSV-E-91 New AA DSimSV-E-151 A A A C C U G A G C G C G New AA DSimSV-E-155 A A A C C G G G G C G C G New AA DSimSV-E-177 New AA DMelSV-E-216 New AA DMelSV-E-238 New AA DMelSV-E-302 C New AA DMelSV-E-332 New AA DMelSV-E-343 New AA DMelSV-E-349 New AA 107

PAGE 108

Table 4-2. Continued. Gene L L L L L L L L L L L L L L L Codon Number 3028 3068 3069 3089 3108 3113 3156 3164 3172 3202 3234 3296 3341 3359 3361 DMelSV-A codon ACA AAG GGUACA GCA GGGCGA GGUCCG CUC AUU UUC ACG ACC CUC SNP codon position 3 1 2 3 3 3 3 1 3 3 3 3 3 3 3 Consensus AA T K G T A G R G P L I F T T L DSimSV-E-52 U C A G G A C A A A A A New AA QD S DSimSV-E-91 C U New AA DSimSV-E-151 U C A G G A C A A A A A New AA QD S DSimSV-E-155 U C A G G A C A A A U A A New AA QD S DSimSV-E-177 C U U New AA DMelSV-E-216 C U U New AA DMelSV-E-238 C U New AA DMelSV-E-302 C U New AA DMelSV-E-332 C U U New AA DMelSV-E-343 C U U New AA DMelSV-E-349 C U U New AA 108

PAGE 109

Table 4-2. Continued. Gene L L L L L L L L L L L L L L L Codon Number 3392 3399 3426 3430 3465 3474 3525 3531 3655 3666 3755 3769 3771 3808 3820 DMelSV-A codon UCG CCU GCCUCU GCA CUGAUA AGC AAU AGC CAG UCA GUUAUA UUG SNP codon position 3 3 3 3 3 3 2 3 2 3 3 3 3 1 1 Consensus AA S P A S A L I S N S Q S V I L DSimSV-E-52 A C G C A C A U G G C New AA TR L DSimSV-E-91 C New AA DSimSV-E-151 A C G C A C A U G G C New AA TR L DSimSV-E-155 A C G C A C A G C New AA TR L DSimSV-E-177 G U New AA SH DMelSV-E-216 G New AA DMelSV-E-238 New AA DMelSV-E-302 C New AA DMelSV-E-332 G U New AA SH DMelSV-E-343 G New AA DMelSV-E-349 G U New AA SH 109

PAGE 110

110 Table 4-2. Continued. Gene L L L L L L NC Codon Number 3825 3872 3881 3927 3996 3997 DMelSV-A codon GGG UUG UAC CAA GAU AAU C SNP codon position 3 1 3 3 3 3 Consensus AA G L Y Q D N DSimSV-E-52 A U C C U New AA DSimSV-E-91 C New AA DSimSV-E-151 A U C C U New AA DSimSV-E-155 A U C C U New AA DSimSV-E-177 New AA DMelSV-E-216 New AA DMelSV-E-238 G New AA DMelSV-E-302 C New AA DMelSV-E-332 New AA DMelSV-E-343 New AA DMelSV-E-349 New AA

PAGE 111

Table 4-3. Summary of synonymous and nonsynonymous substitutions in anc estral virus evolved in Drosophila melanogaster (DMelSV-A) and in strains evolved in D. simulans (DSimSV-E) and D. melanogaster N= nucleoprotein, P = polymerase assisting phosphoprotein, X = unknown function, M = matrix protein, G =glycoprotein, L =pol ymerase, NC = non-coding sequence, SNPs = single nucleotide polymorphisms = synonymous, Kb = kilobase. Evolved Strain Gene Total SNPs Syn Non-Syn Total SNPs/Kb Non-Syn /Kb N 7 6 1 5.76 0.82 P 5 3 2 5.31 2.12 X 8 5 3 8.53 3.20 M 3 3 0 4.49 0 G 22 18 4 12.29 2.23 L 68 60 8 10.80 1.27 DSimSV-E52 NC 8 --25.81 -N 0 0 0 0 0 P 0 0 0 0 0 X 0 0 0 0 0 M 0 0 0 0 0 G 0 0 0 0 0 L 5 5 0 0.79 0 DSimSV-E91 NC 0 --0 -N 7 6 1 5.76 0.82 P 5 3 2 5.31 2.12 X 8 5 3 8.53 3.20 M 3 3 0 4.49 0 G 22 18 4 12.29 2.23 L 68 60 8 11.87 1.62 DSimSV-E151 NC 8 --25.81 -N 11 10 1 9.05 0.82 P 5 3 2 5.31 2.12 X 13 8 5 13.86 5.33 M 2 2 0 2.99 0 G 25 19 6 13.97 3.35 L 66 57 9 10.48 1.43 DSimSV-E155 NC 7 --22.58 -111

PAGE 112

Evolved Strain Gene Total SNPs Syn Non-Syn Total SNPs/Kb Non-Syn /Kb N 0 0 0 0 0 P 0 0 0 0 0 X 0 0 0 0 0 M 0 0 0 0 0 G 1 1 0 0.56 0 L 5 3 2 0.79 0.32 DSimSV-E177 NC 0 --0 -N 0 0 0 0 0 P 0 0 0 0 0 X 0 0 0 0 0 M 0 0 0 0 0 G 0 0 0 0 0 L 4 4 0 0.64 0 DMelSV-E216 NC 0 --0 -N 0 0 0 0 0 P 0 0 0 0 0 X 0 0 0 0 0 M 0 0 0 0 0 G 0 0 0 0 0 L 3 3 0 0.48 0 DMelSV-E238 NC 1 --3.23 -N 0 0 0 0 0 P 0 0 0 0 0 X 0 0 0 0 0 M 1 1 0 1.50 0 G 0 0 0 0 0 L 5 5 0 0.79 0 DMelSV-E302 NC 0 --0 0 N 2 0 2 1.65 1.65 P 0 0 0 0 0 X 0 0 0 0 0 M 0 0 0 0 0 G 1 1 0 0.56 0 L 5 3 2 0.79 0.32 DMelSV-E332 NC 0 --0 -112

PAGE 113

Table 4-3 Continued. Evolved Strain Gene Total SNPs Syn Non-Syn Total SNPs/Kb Non-Syn /Kb N 0 0 0 0 0 P 0 0 0 0 0 X 0 0 0 0 0 M 0 0 0 0 0 G 0 0 0 0 0 L 4 4 0 0.64 0 DMelSV-E343 NC 0 0 0 N 0 0 0 0 0 P 0 0 0 0 0 X 0 0 0 0 0 M 0 0 0 0 0 G 1 1 0 0.56 0 L 5 3 2 0.79 0.32 DMelSV-E349 NC 0 Table 4-4. Summary of statistical analysis co mparing each of the six genes from twelve viral variants. One was the ancestral virus and eleven were passaged in flies for sixteen host generations. dN > dS R (trans/trans) Gene P Z Statistic G 1 -4.246 5.56 L 1 -7.804 2.84 M 1 -2.107 10558160 N 1 -2.493 13.08 P 1 -1.514 5.02 X 1 -1.739 3.35 113

PAGE 114

Table 4-5. Consens us sequence for ancestr al virus collected from wild-caught Drosophila melanogaster 1-50 GACACCAAGACCGAGUUCGUGGUCCGAACCCCUGAAUUUACAAGUCCGGU 51-100 AGAGUAUCCGUUCACCUGGUUUACAAACAAUAAGACAAAGCCACUAUUCA 101-150 AGAUCUCUGUGAUGGCUGACUGUAGCCUGGAAACGGCCAGAGUCGCAGCU 151-200 UUGGAGUUCCUUAUGGGAG AGAAAGUUCCUCCAUCUCAUGUGAUCGAUUA 201-250 CCUUUACGAAUUUUGUAAGACUAUGACUCAGGAAUUAGAUACCAACUGGG 251-300 AGAGUUACGGACAGGUCAUCGGUAAG AAGGGUGAUACAGUUACUCCUAUU 301-350 AAUCUCCUCCACGUCAUGAUAUCUCAAGACACGCGCAAGUACACACCAAA 351-400 AAAUCCGCGCUUACUGACCGAUGAGGUUGAUAUCUAUUUGGUGGGUAACC 401-450 UCCUUUGCUCAUAUAGGUAUAAC AAGACACAUGAGAAAAUGCAAACGGCU 451-500 UACGGCACCAAGGUCGCUUCGAUCCUCGCUCCAUUUUCAACGCAUGAUAC 501-550 ACAAAACAUCCGAAUCUCU ACAUUCUUGACUAACUCCAAAUCACUUGUAG 551-600 ACCAUCCAAACUUUGAGAUAAUGGCA UCUGCUAUUGAUAUGUUUUUAGAA 601-650 AGAUUCCCCUCGCAUGG AAUGGGAAAGCUGCGCUUUGGAACCAUCGGAUU 651-700 AAGGUACCAAGGGUGCUCUGGUCUU GUGGACUUAACCUACCUUAAACAAA 701-750 UACUUAAAAAGUCCGGAUUAGCAGAC GCAAUAAAAUGGCUAUUUGCAGGG 751-800 UGCCUAGUGUCUGAAGUAUUCCAGAU GAUGAUCAAUCAUCAGGAUGAAAU 801-850 UGAUGUGAAACACUCUUACUUCCCAUACAUGAUCGGAUUCAAGAUAAGCA 851-900 ACAAGAGUCCUUUCUCUGCCGGGAGU AACCCCUCAGUCCACACCCUUGUC 901-950 CACCUGAUCGGUUCCUUGCUUGGAUCACCAAGGUCAAUCAAUGCGAUCAU 951-1000 GAUUGAAUAUGGUGUCAUAAGUGACAUUGUGGUGAAUGCAGCAAUAGUUU 1001-1050 UCUUUGCACACCGCAGCGAUAUCG GUGCUAAAGUUCG GUUUGGGGAGAAA 1051-1100 GGUGUGAUGGACGAGAUUCACCGGG AAGAAGAUAUGGCCAGACAAAGAAA 1101-1150 GAGAUCUGCUACUGCAGGCGGACUGCCGGCCACGCAACCUACCGACCCAC 1151-1200 GUGGUUGGCUAGCGGAAUAUCAAGAU AGAGAUUAUAAAUUCACACAGUCA 1201-1250 GAGGUGGAUGAUCUAAACAAAGUGAUUC AAGGUAUACCACAGGUACGCGC 1251-1300 CGGCACGGUGGGAGAGUGGGUGAAG UUGAACUUCAUCUCAAACCUUCCGA 1301-1350 AAGGGUUUGAUGUAACUCGGGUGUGACAGCUCAUGAAAAAAAUACAAAUA 1351-1400 AUCAAGAUAAAAUACAACUCCAAAGAGCCAACAUCACCAUCACACACCAC 1401-1450 AGCCAGGACUUGGAAAAGCACUAAC AACAAAGAUGAAUCUGACUCAGGAG 1451-1500 GAUGAACGAAGGCUUGCACUACUACGUGAGAGAGCCGCUGCCUGGGAUCA 1501-1550 UGCAAAUGCCUUGGGUGAGUUUGAAAG CAAUGCAGAAUUGGAUGUAUACC 1551-1600 AUGAAGGUAUUGCGUAUCCCCCCUCGAUCGACGAGAUGGAGGUAGGUGAG 1601-1650 GAUAGACCCACUUUAGACAAGAUCUUGGAAUCCUCCAUAGGCUCGGAAUC 1651-1700 CUUACCGAAUCUAGAGGAUGGACCACCAGAUCAAGAGGCAGCCUACCAUG 1701-1750 AAGGUGAGAUGUAUUCGACGAUGACAACCAGCUACAUGGACCAGGGUAAU 114

PAGE 115

Table 4-5. Continued. 1751-1800 GAGAGCAUGGACCUAAGUGCAGGACCAACCUGCUCUACUCCUCAACCUAU 1801-1850 UACCGGGAUCCAUCAUAAAGAGACCA UCAAUGGUCGGGUUACCCAGAUGG 1851-1900 UUUACAUCGGUGAAACCUCUGAUAUG GCACUCAUCAAUAGAUUGCAAGAU 1901-1950 GAAUUUGAUGUUUUGUAC AACAUGUGUCCGAAUCGUAACAGGCGGGGAGA 1951-2000 GAUUGUACACUUGAUAGUGAUGUGGAGCACCCCUCUUAAACCGAGCACUC 2001-2050 GCUAUUCGUUAUCGCCAGCCCAACC CCCCGGCAAGAAGAGGAAGAUUGUU 2051-2100 CAACACUCAUCAGACCCAGAGCCAGAGAUAGUAUUUGUCCCGGAACCCAC 2101-2150 UUCAUCAGAACAGUCAUUACCUGAGCGUCAGAUAAAAGCGUAUCUCCAGC 2151-2200 GGGGGAUAGAACUUAAGGGGAAGUCGG CGGAAGUUCCCAACUUCAAGAUC 2201-2250 ACCAACGAGACAUUGGGUUUCUCUGAUGCGAACCUUAAGUUAUUAUACCC 2251-2300 CGACACCUCUUUGUAUCCGGACAACCCCGUGUUUAUAUUAGAAGAUGUGU 2301-2350 UCACAUCCACCAGACAAUUAAAUAAGAUCAAGGUAAAUUACCAGUUGUAU 2351-2400 GAACCGGUACUCCCCUCUAUAUAGGGGA GCGACCCCAUCCAUGCACCGUG 2401-2450 CUUUACCAUGAAAAAAAUGUCUCAACAACAAAUCAAAAUUCCCUCCACAA 2451-2500 UGUGCCACAGAACCUCAGUUCUACCCUUAGUCCUUUCCUUUUUCCAUCUA 2501-2550 UAUUAUUAUGGACAUGUUGCAUCUG UGUCUAUCCCGUUAUCAAUAAAUAU 2551-2600 AACAGACAGGAUAUUGCCAGUACAAUUGAACACUACAAUCUUGAUGGAGG 2601-2650 GUGCAGAAGAGCUGAAUAUUACUCAGCAGUUAAAUGUGCAGCACUUCCAA 2651-2700 GAGGGCCGGCUCACACGGGAGGAAAAUUUUGUCUGUAGUCACAUGACAAC 2701-2750 AACAGAUCUAGCCACUGAUUUCUUCCAACUUUCAAAACUAGCCGGAGCUU 2751-2800 AUGAUGCCGCGGGUAGCCAAGACCAAAGAGUGUACUCUCCCUCUUCCAUC 2801-2850 UCCACGUUGGCCCCAAUAUUUCCUGAAACCACACCUGCAACCCUCAGCCC 2851-2900 UGAUGGAGUGACGCUAAACAUCAGGACCUCGGUGAUGGAUAUUAACCCAU 2901-2950 UGACUCGCUUCACCAAAGCAGUUA AAGAUCUUAGGGAAGACAUUACACCA 2951-3000 GGGUAUGAUCUGACCAACGGAAGGAAGUUUGGUCGGGAGUUUUUAAGAGC 3001-3050 CCUUGAGGAGAAUGGGGUACGAAGCAGGCACAAGAGGGAUUCAACUACUG 3051-3100 AUAUGCUGCAAAUGACACUGGAUGUG ACUGAGUCUAGAGAAAAAUACGUU 3101-3150 GAAAUCAGGGACCAGUGUGCAAGCG UGCUAGGUGCUCUGAUCAAGAGUAU 3151-3200 GGUUCCAUCAGUGUCGGAACAUUGCUUAUAUCACUAUCACAUUAAGAAUA 3201-3250 UAGUAAAUUGUUUCACUAGUCAUUACAU CAGACUCCACAAUGAGCCAGAU 3251-3300 UUCCCGUUGGUUAUGGCAUUAUACUCUGAGUACCUAUCACUCACAGUGAA 3301-3350 GAUACAUUACAUGGACGAGAUU UUAGUCAGCACCCUAAGGGCAUGAAAAA 3351-3400 AACUAGCAACAUCAAUUACAAACAUCACUGACAAAUUCUCUGUGAAACCG 3401-3450 CCAUGAAUAAGAUGAACCAAUUGG UCAGGUUUGUUAAAGACACCGUUGCG 3451-3500 GUGAGAAAACCACAGAGUGAAGAUC AAUUAUAUCUCCCUAUCCCCAGCAC 3501-3550 GAUCGGAGGACAUGAGGUGGACUCACCAUUCGCCGAGCCGACCGCCCCGA 3551-3600 CACUAGGAAUCAUCCAAUCCAAGUG CAAAAGGGCCGACUGGCUGAUCAAG 115

PAGE 116

Table 4-5. Continued. 3601-3650 UCGCAUCUCACGAUCACAACUAAUUA UGAGAUAAAAGAGUGGGAGACCUG 3651-3700 GGACAGAGCAAUUAGUGAUAUUCUGGA UCUGUAUGAUGGCAAUCCAGUAU 3701-3750 UCAAGCCUAUUUUAUUGUUCGUGUAUUAUGUAUUAGCCUACAACGCCCGG 3751-3800 AAAAUUCCUGGCCCAUCCAACGGGGU AAGGUAUGGGGCAUAUUUUGACGA 3801-3850 ACUAACUACUGUAUGGCAUGCCAUC CCGGAACUAAUGAACCAGGAGACUG 3851-3900 AUUAUUCAUACAAUCA UCGGGUAAUCCAUAGAAAGAUCCAGUAUGUCAUC 3901-3950 UCAUUUAAGAUUCAGAUGUCAUCUACUAAGCGGCGAACUUCACCUAUCGA 3951-4000 GUCCUUCAUUGAGGUUACAUCUGAAGGACUUAAACACACUCCACAGUUUA 4001-4050 CCACCAUUCUAGAUCGUGCACGGUUUG UCUAUUCGUUGACAGGAGGGCGU 4051-4100 UAUGUUAUCCAUCCUUUCUAAAACU AACCUACGCAGUAACAACAGAUUAU 4101-4150 UCCAUCUGCAUGAAAAAAACUCCUUCUGCUGAAAUGGUACAUUACGAGAC 4151-4200 UCAUAUUCUAUUCAUCCAUCUCUGGAUGUUGGCACUGAUAUUUAUCACCA 4201-4250 CCUCAGUAUGGUUGGCCGCCAGCCAAAAAACUUUUACCCCGGAUUUGGUA 4251-4300 UUCCCUGAAAUGAAUCGUAAUAGUUCUUGGAGUGUGGCAAACUAUGGAGA 4301-4350 GAUAUUAUGCCCAACAUCCUUUCAGUC AUACGACCCCAAGAAGCACCAGA 4351-4400 UCCUGACACGUGUCCUAGUUGAGAGACCAAGCCUAAACACUGACACAAAG 4401-4450 GUGGAGGGCUACACAUGCCAUAAGGUGAAGUAUGAGACCAUCUGUGAUAU 4451-4500 GCCAUGGUAUUUCUCCCCUACCAUUUCACACUCCAUCUCUCCGCUUAGGG 4501-4550 UAAAAGAGUCUGAGUGUAAAGACGCUA UAGCAGAGCACCAAUUGGGCACC 4551-4600 CAUGUCUCGCUGAGCUUUCCCCCUGAGGACUGUUCUUGGAACUCAGUGAA 4601-4650 CACGAAGGCGUACGAGGAUAUUAUAGUUAAAGACCAUCCUGUAAUGUUGG 4651-4700 AUCCAUACACUAAUAAUUAUGUUGACGCAAUAUUUCCAGGUGGUAUUUCA 4701-4750 AGCCCAGGGAUGGGGGGGACCAUUCAUGACGAUAUGAUGUGGGUAUCCAA 4751-4800 GGAUCUGGCAGUGAGCCCAGAAUGCUCU GGGUGGCAACAAAGUAUGGGGC 4801-4850 UUAUUUAUUCAUCUCGGCUGUAUGGAG AGCGGGAGCCGAUGCUUGAAGUA 4851-4900 GGCUCCAUCCACAUCGAGGGCCACAGGGACAAGAACCUCACUUUAGCCUG 4901-4950 CCGGAUUUCCUUCUGUGGUGAGAUAGGGGUGAGGUUUCAUGAUGGGGAGU 4951-5000 GGAUGAAGGUAUCUGUAAAUCUUGACCAUCCUAACUCUGUGACAUUCCAA 5001-5050 GUAACUGAUUUCCCUCCGUGUCCUCCCGGUACCACUAUCCAAACUGCCGU 5051-5100 UGUGGAAAAUAUAAACCCAGAGAUUC AAGAGCUUACUGUCAACAUGAUGU 5101-5150 ACAGAUUGAAGUGUCAAGAGACCA UCUCAAAAAUGGUAUCU GGGCUUCCG 5151-5200 ACCUCUGCCCUUGAUCU CUCCUAUUUGAUUCAAGUGCAGGAGGGACCUGG 5201-5250 UAUUGUUUAUAAGAGAGAGAAGGGUG UCUUGUACCAAAGUGUCGGAAUGU 5251-5300 ACCAAUACAUAGACACAGUAACAC UGAACAAGGAGGAGAACCAACUGGGU 5301-5350 GAAAACGCGAGGGGGCAAAAGGUCUUUU GGACCGAAUGGAGCGAUUCACC 5351-5400 GACACGCCCCGACCUCCAGGAAGGG AUCAAUGGGAUUGUUAAAUAUGAGG 5401-5450 GGCAGAUAAGAGUGCCUUUAGGAAU GUCUCUGAGAUUAGAGGCUGCAACA 116

PAGE 117

Table 4-5. Continued. 5451-5500 GAAUUAAUGUGGGGACACCCUGUCC ACACCGUGUCUCACCCUAUACUUCA 5501-5550 UGUUAUCAGUAAUCACACAGAACAGUCUGUAACAACUUGGAACAGGGGUG 5551-5600 UUAAUUCCACAAACCUGAUAGGCCUGGCAACUCGAUCAAUAUCAGGUUUC 5601-5650 UACAAUGAUCUGAAGUUGUAUUUGAUCUUGGCACUGAUAGUCGUUUCAAU 5651-5700 UGUCGCACUGGUUGUCUUAGAUGUGAUCCCCUUUAAGUAUAUUUUAUUUA 5701-5750 UACUAUGCCCACCUCUGCUGUUGUGUAGGUUCAUAAAAUGUUCACGAAGG 5751-5800 AGGCCUGAAACUGGGGACAGGUAUCA CGUAGAAUAUAACCGACCUGGACA 5801-5850 AGUGUCCAGCGCGUUCUAAGGU GGUCAACAUGGCCAUGAAAAAAACAACA 5851-5900 CCCAUUAAGCUGUGACAACUCCAGGUCAUCUUGCAAUUAUCAUGGAUUUU 5901-5950 GAAAUAGAAGAUCCAUAUGAUCCGU UCUCUAUGGACACA UAUUUGGACCC 5951-6000 ACAAGACCCAUCCUUC GGGGAUCUUGAGUCAAUGAGGCAUUUAAGUAAUG 6001-6050 UUGACUACUCUCUGAAUUCUCCGAUGAUUGCAGACGAGUUGGAGGCGUUC 6051-6100 AUUCGUUGGUUGCAAUGC GGAUGCACUGACCCCC GAUGGAACGAGGAUCG 6101-6150 AUGGGUCCGGACGAAACAAGGACUGUU CUCUGGGCAAUCUC CUACAACAA 6151-6200 UAGAAGGUGCAGCUACUUUCACCG GAUGGUUUGGAAACUUUAAUCUAAAA 6201-6250 CGCAGGUACUAUAUUGUGAGACA GUUUAAAAUGAUACUUGAAAAAGCUCA 6251-6300 GGCAGACUCAGAAGAGACCAAACCUGU CGUGGAUGCAUUCCUACGGGGAU 6301-6350 GGAUCAAUCACAAAGGAGUCACCCUGACAUCUAAGAUCACCCUGCCAGAG 6351-6400 GAAGAGUUGAAAUGGGGAU AUUAUUUUUGGGAAUUG CACAUAGUGACUCU 6401-6450 CCACCUAAAUUGUACAACCGACC AAGAGAGGACCCACUUGAUCAAAAGCU 6451-6500 UCAAAUCCAAGUCUCGUGGGUUGCCCGAUGUGUUCGAUUUCACAUUGUAC 6501-6550 ACGAGAAACUUCGGGCCCUUGAGUAUAGCAGGGGGGUAUGUUUACAUGUU 6551-6600 UGAUCACAACAGGAUGCUGGACCGCAAU GCAAUUCUGAUGAUGAAGGAUA 6601-6650 CCUACGUUGCUCGGUUCAACUCAUUCCU UGCACUAAGCAAUAGGGCAGAC 6651-6700 UGUGUCUUCCCUGAAGACGCCCAUUACCGACUACAAAUGCUAUAUGAGAU 6701-6750 AGGGGACAUGGUCUUGGAUGAGGGCG GCACAUCUGGAUACAAUGGUCUGA 6751-6800 AGUUGCUGGAGGCUAUGUGCAGUAGCAGGAUUACAGACUUGGCACAGUCU 6801-6850 AAGAAGCCUCUAAUACCUGACUUUCCUGACUUCCGUUCCCAUGUUAGAGC 6851-6900 CAAGGUAAGAGAAGAGUCUGCGAA CACACCAUCGAUUGGUAAAAUGUAUG 6901-6950 AGCUGAUUGAAGGAACAACCAGCUAUG ACACUCUUCUCACAUUCUAUGGG 6951-7000 UCCUUUCGUCACUGGGGACACCCUUAUAUAAAUUACUUAGCCGGGUUAGA 7001-7050 GAAAUUGUACAUCCAGACGACGGU GGAGAAAGAGAUUGACCAGGAGUACG 7051-7100 UAGAGAAAUUGGCCAGUGAUCUCGCGUUCCUUGUUAUCCAGGAUAGGUUU 7101-7150 CGUAAAACAAAGAAAUGGCCC GUUGAUCCUCUU CUAAUAGAUAAGGACCA 7151-7200 UCCUCUGGUAGAGUAUAUCCGGACAUCAUCAUGGCCCAACAACUCCAUCA 7201-7250 UUAAAAAUUUUGGGGAUGGAUGGCACAC AUUACCUUUGACUAAAUGUUAU 7251-7300 GACAUCCCUGACGUGAUUGAUCCAUCACUCUUGUAUUCAGACAAGAGUCA 117

PAGE 118

Table 4-5. Continued. 7301-7350 UUCUAUGACGAGAUCGGAAGUCAG GGGAUGGAUGACAUCACACCCUGGGA 7351-7400 AACCGAUCCCCUCCCGGAAAGUGCUGUCAACUCUCUUAAACUCCCCAUCG 7401-7450 ACGAACUGGCCCGUGUUCUUACAACA GGUGAACGAUUCAGGCAUACCCAU 7451-7500 UGAACAGCUUAUUAUCGGAUUGAUGGCAAAAGAAAGAGAACAGAAAAUAG 7501-7550 ACGGGAGGUUCUUCUCUUUAAUGUCUUGGGACAUCCGCGACUAUUUUGUC 7551-7600 AUGACAGAAUACCUCAUUAAGACCC ACUUUGUCCCGU UAUUCAAAGGGCU 7601-7650 GACCAUGGCAGAUGACCUCACAACAGU GAUUGGUAAGAUAUUGGAGAACU 7651-7700 CCCGGGGUCAAGGAGAAGCAGACUAUGA GAACUUGACUAUCACUGAUCAC 7701-7750 AUUGACUAUGAAAAAUGGAACAACCA UCAGAGAGGGGAAGCAAAUAAUCC 7751-7800 AAUUUUUUUGGUCAUGGGGAAGUUUUUA GGUUACCCGCAUCUCAUAGAAC 7801-7850 GGACCCAUGAGAUAUUUGAGAAAUCAUGGAUCUAUUAUCUAAAUAGGGCU 7851-7900 GACCUGAUGGACUUCGAUGGAGAGGGC AAUCUGAUGAAUAGAACCGAACU 7901-7950 CCGGGUCUGCUGGAAUGGGCAGAAGGGUGGACUGGAGGGGCUGAGACAAA 7951-8000 AGGGAUGGAGCAUUUGCAACUUGCUA GUAUUGAGGAGAGAGAGCCUAGCU 8001-8050 ACUAACACAGUGGUUAAAACGCUCGC CCAAGGAGAUAAUCAGGUAUUGAG 8051-8100 UUCUAGAUAUAGGAUACGCACAUCACGAGAUCAGAAUCAGUUGCAGUCGA 8101-8150 ACAUUGAGGAUAUCUGUAGAAAUAA UCGAAGCUUAAUGGAGAGAAUUAGG 8151-8200 AUAGGGACAGGGAAACUAGGACUCAUUA UAAAUCAUGAUGAGACCAUCAA 8201-8250 AAGCACCGAGUACAUGAAUUACGGG AAGACUUGUGUCA UCCAUGGUAAUA 8251-8300 UUAGAAACCUUGAGACAAAGAGAUGGUCUAGAGUAACCU GUGUAACUAAU 8301-8350 GAUCAGUUACCUACCCUAUCAAAUGUCAUGGCCACGAUUGGAAGCAAUGC 8351-8400 CCUAACAGUCUCCCACUACUCUGACUCCCCGAUCAAUUCGAUGUAUCAUU 8401-8450 AUAAUUUUAUGGGUAAUUUUGUGAGGAUCAUGAAUGAGAUACACAACCCA 8451-8500 GCAUUGCGUGGACCAGUGUCCUCUA UAGAAGGGGUAACCGGACAAUCAUU 8501-8550 CAGCCGACUGUCUUACUUAUUGGCGGUCUUGUACUUAGACCCCUCUAUGG 8551-8600 GAGGUGCAUGUGGUAUGUCUUUAACCAGGUUCCUGAUUAGGAUGUUCCCG 8601-8650 GACCCGAUAACAGAAAGUC UUACUUUCUUGAGGAUCGUAGCGAUGAAUGU 8651-8700 UCACUCUGAUGAGGUCCGACAGACAUUUAUCCAGUUUGGUAAUCCUAAAC 8701-8750 UGAAGACAUUCUCACCGGAAGACUUAU CUAAGCUUUUAGAAGACCCUCUC 8751-8800 UCUAUCAACGUACCGAAAGGUUUGAGUGCUACUAAUCUGAUAAAGGAUGC 8801-8850 UAUAAAACUGUCGCUACA CAGAUCAGUUGACGAAA UUGCCAAUGAAAUCA 8851-8900 UUGCAGAAGCUGUGAUCCACCAAAAAG AUCAUGAAGAAGGAUUCCUCAUG 8901-8950 CAUCUGACACAGAUCAGUCCCUUGUUCCCGAGAUUUUUGAGUGAAUAUAA 8951-9000 AGCUGGAACUUAUUUGGGCAUAGCUGAGGGUCUAAUCGGAUUAUUUCAAA 9001-9050 ACUCUAAAACCAUCAGGAACCAAUUUAGGAGGAACCUCGAUAUUGGAUAU 9051-9100 GACAGUAUUGUGAUUAAGUCUGAGAUUGCAACGAUUAGAGACUUGACUGG 9101-9150 GUACCGAUUAGAGGAUGCUGAAAGAGUGGAGAUGUGGCCAUGUUCAUCCA 118

PAGE 119

Table 4-5. Continued. 9151-9200 CGCAAGCAGAUUACUUGAGGAGAGUGUCAUGGCAGCAAGUGGUUUAUGGG 9201-9250 GCGACAAUACCUCACCCUGCCGAGUUAUUCGGACUUCCUCUAAGGGCCGC 9251-9300 CCCGACUUGCCCAAAUUGCACUACAACUUUCCCGAUGAACCUGUACAUCU 9301-9350 CUGUAUUGAUUCCACUUGGGUUCAA GGGUUUGAAAGAUACCCGAGGAACA 9351-9400 UGUGUGGCAUAUUUGGGAUCUAGUAC GACAGAAUCCACAGGUAUCGUUAA 9401-9450 UCCAUGGGAGAAGGAGGCGGUUGUG CCAGUGAUCAAAAGGGCAGCAUCGC 9451-9500 UGAGAAAUGGGAUAGGAUGGUUCAUU GAACCUGGAUCAAACCUGGCACAG 9501-9550 AGUAUUUUAAACAACUUGCAAUCUUUGA CAGGCGAGUCCUGGAGCCAAAA 9551-9600 CUCCGGUGGGGUGAGACGAACUGGAUCUGCAUUGCAUCGAUUCAGUUGUU 9601-9650 CUCGGCAGAGUGGUGGUGGUUACACAGCUCAGAACCCGUCGAAGUUGACA 9651-9700 AGGAUGAUUGCCACGACCAAUUAUCUUGCUGAUCUGGGAGAUGAGAACUA 9701-9750 UGAUUUCAUGUACCAGAGCUGCUUGCUCAACGCUUUAAUAUCAGUGGGGG 9751-9800 AGAUCCACCCUGUAGAUGGGUCUCAAGGUUAUUAUCAUCAACACGUCAAU 9801-9850 UGUACGUCCUGCCUCAGACCUAUUAA GGAGGUGACACUGGAAAGCCCUGC 9851-9900 ACCUUAUUCCCAUACCAUAACAUCCAAUCUUCUGGACAAAUGGAAGCCAG 9901-9950 AUGGCUCGAAGUGGUCAGUCUCCCGACCCUCUAUUCCAUUGAGAUCAGGG 9951-10000 AAGUGGGAAUGUGUGUCUCAUGAUCGUCAAUCCUAUCAUGUUGGCUUCAU 10001-10050 ACAAGGGUUCAUUUAUGGAGAUUCUGUAUGGGGAAUCCGAUCUAUGGCUG 10051-10100 AUGAUCCCGCAUUAUUCCCAUUGAGUUUCCGGAAUAAGGUUAACCCUCGA 10101-10150 GCCUACCUACUGGGAAUCUUGCACGGACUACUUCGCAGCUGCACGGUUUC 10151-10200 GGUUGUCCAUCAACGAUGUUUUCGAAG CAGUAGAGCCGUGAAACAAACCA 10201-10250 CCCUCGGACUCUGUAGCAUGACUGUG AGUAGGCUGGUU CAAAACGAUGGA 10251-10300 UUUCUAAACAUACUCCGGGAUGAGCA GUUUACGGCAGUUUUCAGAUCGAU 10301-10350 CCCACACAGGAUACCUCCUUCUUAUCCGAUGGUGACCCAAGACAUAGGGG 10351-10400 ACUUGGCAUCAAACUACCUAAAGCGCCAGCUGAUGACUG AAGGGCUGGCC 10401-10450 UACUUCAGGUCUGUCA AAUCGGGAUCUGGAAAUG AAGCAUGGAUCUUUGC 10451-10500 UGAUGCUAACCACCCUAUGAUCGUGAGCCUAAUAACAGUUAGUGGGCUAA 10501-10550 UGACCAAAAUUUUAGCAAAGGACAUA UGGCAGAAGAGAG ACCUGGAAGAA 10551-10600 UUGAAGGAGCUUCGAAGCUUGAGCGU ACAGGCGCGAGAACAGGAUGACCC 10601-10650 GCAUACUGACAUCACCGCCAUGGAGUCACUCGCCGCUGAGUGGGUUGUUU 10651-10700 GCUCUGAUCAGGAGACCAGACACGCAG UAAAGUAUAACACUCGAAUAGAG 10701-10750 UCCGUAACAGACAGCGGGCUGACAGUG ACAUGGGGUGAUGAAUACACUUC 10751-10800 AUCUGCUGAUUUCGUAACAGUGGUGUUCUCUGUAGAAGAAAUCCGUACUC 10801-10850 CACCGGGUAUGCUGAUCCCCAGGAUUCAAAACCCUUUAAUAUCUGGCCUA 10851-10900 AGAACAGCACAAAUUGCAACUGGCUCUCAUUACAAACUAAGGAGUAUUCU 10901-10950 UUCAAAACUCCGCCUCAAU GUAAGGGGCGCCCUAGUGGGAGGAGACGGGU 10951-11000 CAGGGGGGCUGACUGCCUUAGUGUGUAGAAUGUACCCCACUAGCAGAGUA 119

PAGE 120

Table 4-5. Continued. 11001-11050 AUCUUUAACAGUAUAUGCGAUUUCUCA GAUGUACGCCUGAAGGGCACAAC 11051-11100 UCCUGCACCUCCGUCUGCCCUCUCUCACUCAUUGAAUGAUUGUACCCAAG 11101-11150 UUGUGAAUUAUUCCGAUAGCUGGGCACAUCCUAGCGAUCUUACUGACACA 11151-11200 AAGACAUGGAAAUACUUUGUGGACAU AACUAAAUCUAAAUCCAUACAAGU 11201-11250 GGAUCUGAUAAUACUGGAUAUGGAAGUAGUAGACGAGAGCUCAAUAUCCA 11251-11300 AGAUUGAAGACAACUUGAUGAGGUAU GGACCUCAACUGCUGACAAGAGAU 11301-11350 GGAGUGAUCCUAUUUAAAACGUACCU CACCAGGAUAUUCAAGGCCCAGGA 11351-11400 GAUGAUAUUGACCAAAUGUGGACAUGUCUUCUCUCAGGUCGAGCUAUGGU 11401-11450 ACUCAGAUCUAUCUUCAUCCCAAACAUCAGAGGUUUACGUCUUGAUGUCA 11451-11500 GGGCAGACAAAACUCUCCCACCUUCAG GUGCGUAAG CCGGAUCUGAUGCA 11501-11550 AUUGAGGACAGAUGUCUCAUCCUUCCCUGUCUUUGCACCUCCGAUAGUGG 11551-11600 AAUUCAAGCGGGCCAGAAAGGUCGCCGGG UUGGAUUUGACAGUAGGGGUA 11601-11650 CCCCCUCUACUACUGCCAGAACCAGGCCCUGAAAUGAUUAACCUCUUAUC 11651-11700 AUCAUUAGGAGUUCGCCCUGAUAUAUCAUUUUCGAUCACUAGAUCAUUUG 11701-11750 GGGACAACAUGUCAACAG AAUUCCUACCUCUACAC UUGUUUCUACUAGCG 11751-11800 UUGAAUGGGAUAUACGAUGUCACAACCGGCUACAGGACCCCUCCUGGCCC 11801-11850 UCCUUCCGACCAAGUGUGUAUAAAGUUGGGCAUCUGGAUAGUGGGAUAUA 11851-11900 GGAUAUGGUCAGGAUAUGUCCAAGAU UCAUACGCAAAAGCC UCAUUCGGU 11901-11950 CAAAAACUCAUUGAUAAC UUUGUUCCGUUC AACUUUUGGAACAAACAGAU 11951-12000 UGGACGAUCUUGGUUCCCGCAUUGGUCACUUAUUAAACCACAGACAUAUA 12001-12050 CGAAAAAUAUACAAUUGGAUUCUGAGAU GGCCCAUAUUGGUUCAGUGAUA 12051-12100 CGUAUCCUGCAUCGAAAUUUCCCCAAAGCCAGAAGAAACCCUCCCGGUGA 12101-12150 UUUGGUCGAUAAUAACUGUUCUCGAAUCAACAAGGGGAUAUCCUGUCUCA 12151-12200 ACAUGGACAUGAAAACGGGGAUAUUAC GGUGGAUGGGGGAUGGGCGAGAC 12201-12250 CUGUCUGGAGUUAAUGUCACCACCCGGAUCUCAAACUUCUAUGUCACCCA 12251-12300 AGAUGAUCAAACCGCUAGCUUCAG AAAUUAAUGGACUCCUCGUCCCUUUU 12301-12350 GCCUCUCAUGAAAAAAACAGUUCUGGAAG CACUAUCCAUGCAAUCGAUUC 12351-12400 CGCAGGCAUCACUGAAUAACAGAU UCCGCGUACAAAAUUGAAUCAGU 120

PAGE 121

121 DSimSV-E 177 DMelSV-E 332 DMelSV-E 349 DMelSV-E 216 DMelSV-E 343 DMelSV-E 238 DSimSV-E 91 DMelSV-E302 DMelSV-A DSimSV-E 155 DSimSV-E 52 DSimSV-E 151 100 100 46 95 88 67 68 28 95 Figure 4-1. Maximum likelihood phylogenetic relationship am ong ancestral sigma virus from Drosophila melanogaster (DMelSV-A) and sigma virus evolved in naive native hosts (DMelSV-E) and in novel D. simulans host (DSimSV-E) lineages. Bootstrap values are indicated at the branches.

PAGE 122

CHA PTER 5 TESTING ATTENUATION BY CROSS-INFE CTING NAVE NATIVE AND NOVEL HOSTS WITH EXPERIMEN TALLY-EVOLVED VIRUS Background Interspecific host shifts are a major source of zoonosis in humans. Recent examples of shifts into humans include HIV, SARS and influenza ( CLAAS et al. 1998 ; GAO et al. 1992 ; HOLMES and RAMBAUT 2004 ). The debate continues as to what biological processes govern the evol ution of virulence; it is clear that a one size fits all theory is unlikely to exist ( ALIZON et al. 2009 ; ANDRE and HOCHBERG 2005 ; EBERT 1998 ). Thus, we need to continue exploring the evolut ion of virulence unde r different conditions and in natural as well as model systems. The evolution of virulence following hos t shifts has been studied primarily in serial pass age experiments (SPEs) ( EBERT 1998 ). Two major conclusions from SPE studies are that pathogen virulence increases in the novel host (as observed in Chapter 3) and that this increase in virulence in the novel host is accompanied by attenuation in the native host ( EBERT 1998 ). Attenuation is defined as a reduction or even elimination of the ability of the evolved path ogen to re-infect, replicate and induce virulence on is native host. Since pathogen success is guaranteed under SPE regimens at every generation, which could influenc e the experimental outcomes ( EBERT 1998 ), and the majority of SPE e xperiments involve single-celle d novel hosts, the degree to which SPE findings will be replicated in natural pathogen evol ution in a complex multicellular host is unknown. The Drosophila /sigma system provides a testi ng ground for the hypothesis that attenuation to the native host occurs followin g a shift to a novel host in a complex, multicellular host. Drosophila melanogaster is the native host of this virus and is an 122

PAGE 123

excellent model organism with many genet ic and molec ular tools available ( SCHNEIDER 2000 ). The DMelSV rhabdovirus of D. m elanogaster (DMelSV) is an RNA virus that is very similar to the causative agents of many human disorders, such a rabies ( ROSE and WHITT 2001 ). DMelSV is readily passed from parent to offspring ( FLEURIET 1988 ) and can be successfully shifted onto new host species ( As in Chapter 3; LONGDON et al. 2011 ). This virus/host system thus elim inates the need for experimenter-assisted passage ac ross host generations. In a previous study (Chapter 3), DMelSV wa s artificially host shifted onto a novel host (D simulans ). Many infected novel host lines were produced. The first three generations after injection underwent selection for transmission effici ency. After this, the virus was passaged naturally (without experim enter selection) for 13 host generations. While transmission efficiency declined in many of these infected host lines, four sustained a steady infection with high DSim SV titers and infectivity (Table 1 and Chapter 3). High titer was def ined as at least 5.0 x 105 viral genome copies per microgram of purified host RNA. We regarded the virus strains in these fly lines as welladapted to their novel host. Here we tested whether attenuation would occur in these four strains by injecting the D. simulans evolved (DSimSV-E) viru s strains, as well as the ancestral, unselected isolate from D. melanogaster into nave D melanogaste r and D. simulans hosts. We tracked infection success, transmissibility and virulence across three host generations. Materials and Methods Nave Hosts We infected one effectively isogenic line of D simulans (MD106; www.dpgp.org) and one highly inbred line of D melanogaster ( Diallel 75 ; YANG and NUZHDIN 2003 ) with 123

PAGE 124

DSimS V-E. MD106 is one of the effectively is ogenic lines included in the Drosophila Population Genomics Project ( www.dpgp.org ). Diallel 75 is a highly inbred line ( 40 generations of full sib inbreeding; YANG and NUZHDIN 2003 ). These lines were used previously to generate the virus evolved in D melanogas ter and D simulans (Chapter 3). Inocula and Injections Inocula were prepared from f our DSimSV-E-infected fly lines exhibiting high virus titers and infectivity of D simulans -evolved virus (Chapter 3; see Table 1 for titers and inoculum names). These viral genotypes were the result of 16 generations of experimental evolution (first 3 generations with selectio n for transmission efficiency; subsequent 13 generations under relaxed selection) in novel ( D simulans ) hosts. Each inoculum was prepared from ~200 females fr om each evolved line as in Chapter 3. The D. melanogaster and D. simulans flies (1-2 days old females) were injected as described previously (Chapter 3). Brie fly, a Sutter instruments XenoWorks Analog Microinjector equipped with a 100 l syringe and needles made from Narishige GD-1 blanks were used. Each inoculum was thawed individually on ice and 0.1 l was injected into the abdomens of females fr om each host species. In total, 720 D melanogaster and 720 D simulans survived the injection procedure and were placed on food. Additionally, 144 D melanogaster and 144 D simulans were injected with the ancestral un-evolved inoculum (DMelSV-A) harvested from wild-collected flies. Finally, 20 females from each species were mock -injected with homogenization buffer to control for any effects of the injection pr ocedures. The mock-inje cted flies produced the uninfected control lines. 124

PAGE 125

Rearing Conditions and Artific i al Selection for Infection After the injection procedure, each fly was placed in a vial with 7 ml of Jazz-Mix Drosophila food that was prepared following the manufacturers instruction, and prepared with additional Bakers yeast and a strip of KimWipe. A male fly was added to each vial. This is referred to as generation zero (G0). All flies were housed at 24C with a 16:8 light:dark photoperiod. Artificial selection was conducted for trans mission efficiency, retaining only those lines which transmitted successfully. The offs pring (G1) of the injected flies (G0) emerged and were transferred to vials of fr esh food 20 days after the injection date. These G1 flies were flipped onto new vials again 10 days later to prevent them from mixing with their offspring. When the G1 flies were 25 days old ( 5 days), they were set up as 1 + 1 and females were allowed to oviposit for two days. After this oviposition period the flies were removed from the vi als, and their fecundity and hatchability was assayed as described below. Finally, the parent flies infection status was determined by CO2 assay. The vials containing the eggs of infected females were retained, while the vials containing the eggs from uninfected fe males were discarded. This process of 1 + 1 setup, aging, flipping, fitness assaying and gassi ng was continued for three generations. At each generation onl y the offspring of the infected females were kept. Fecundity and Hatchability Egg production and per cent hatchability we re determined by allowing flies to oviposit for 24 hours on standar d food tinted with green food co loring (for ease of egg detection; McCormick & Co., Inc, Hunt Va lley, MD) which had been poured in a thin layer onto standard glass microscope slides Approximately 60% of each slide was covered with 1ml of the colored diet. The slides were sprayed with a 4mg/ml yeast 125

PAGE 126

solution and provided to the flies for ov iposition within the confines of a wide Drosophila vial. Subsequently, the female was gassed to determine her infection status. All of the eggs from i nfected flies were counted and the slides were incubated at room temperature for >24hrs. Subsequently, the e ggs that failed to hatch on these same slides were counted to determine per cent hatchability. Over 3000 females were set up on slides and scored for infection but many were uninfected (did not die upon gassing). In total, 630 flies were identified as being infected and had their fecundity and hatchability assayed. Ribonucleic Acid (RNA) Purification Ribonucleic acid from each inoc ulum was isolated using TRIzol (www.invitrogen.com), following the manufacture rs directions, as described previously (Chapter 3). Briefly, a 50 l al iquot of inoculum was mixed wi th 200 l of TRIzol on ice. The mixture was vortexed briefly and incubated at room temperature (RT) for 5 min. Each sample then received 40 l of chlo roform and was vortexed for five seconds, incubated at RT for 2-3 min, then centrif uged at 12,000g for 10 min at 2C. The RNA in the aqueous phase was precipit ated with isopropanol at -20C for 10 min. The precipitate was pelleted by centrifugation at 12,000g for 10 min at 2C. The pellet was washed with cold (4C) ethanol, air dried for five min at RT, resuspended in 29 l of DEPC-treated water with 1 l of 40 M RNAse inhibitor (www.neb.com) and incubated at 55C for 10 min to aid resuspen sion. Samples were stored at -80C until used. Reverse Transcription Reverse transcription of the DMelSV virus genome was performed using the SuperScript III First-Str and cDNA synthesis system ( www.invitrogen.com ) following the manufacturers recommendations. Briefly, 200ng of inoculum RNA were mixed with 1 l 126

PAGE 127

of primer (whole1F: 5 TA GAAGCATCCTCGGCTTTC 3) and 1 l of 10mM d NTP mix and the reaction volume was brought to 10 l with DEPC-treated wa ter. This reaction (part 1) was incubated at 60C for 5 min and then placed on ice. A cDNA synthesis mix (2 l of 10X RT buffe r, 4 l of 25mM MgCl2, 2 l of 0.1mM DTT, 1 l of 40U/ l RNAseOUT and 1 l of 200U/ l SuperScript III enzyme, part 2) was prepared and added to part 1 and the mixture was incubated at 50C for 50 min, then at 85C for 5 min to stop the reaction and then chilled on ice. RNAse H (1 l) was added to the tube and the synthesis reaction was incubated at 37 C for 20 min. The cDNA was stored at 20C. Statistics The proportion of injected flies that bec ame infected, and the proportion of these flies that transmitted the virus to their offspring, were evaluated using G tests. All of the samples were larger than five, thus no data corrections were applied. Because the residuals for infection, fecundity and hatchab ility data were not normal, the data were analyzed using nonparametric statistics as re levant (descriptions in text) with JMP version 10.0. Results Cross-Infection Success and Se lection for Transmission We injected nave D melanogaster and D simulans females with the four high titer strains of sigma resulting from passage for 16 generations in D simulans (DSimSVE), as well as the ancestral inoculum from which they were derived (DMelSV-A; Chapter 3). Each of the five inocula was injected succ essfully (i.e., the injected flies survived) into 144 females from each species (720 fema les successfully injected). As observed previously (Chapter 3), the DMelSV-A inoculum infected D simulans more effectively 127

PAGE 128

than D m elanogaster (P < 0.0001; Figure 5-1). T he DSimSV-E and the DMelSV-A inocula had similar infectivity in D simulans (G = 0.8361, P = 0.3605; Figure 5-1). Interestingly, the DSimSV-E and t he DMelSV-A inocula also infected D melanogaster similarly ( G = 3.15, P = 0.0759; Figure 5-1), suggesting that the evolved virus was not attenuated to its native host with res pect to within-host replication. Since a successful host shift requires that the pathogen not merely replicate within novel hosts, but be transmitted between novel hosts, we determined the transmission efficiency of the infected flies (Figure 5-2). D simulans flies infected by DSimSV-E transmitted the virus to their offspring at significantly higher frequencies than conspecifics infected with DMelSV-A ( G = 6.57, P = 0.0104). However, there was quite a bit of variation among viral strains within D. simulans, with two strains (DsE155 and DsE177) generally higher t han DMelSV-A, while the confi dence intervals of the other two strains (DsE52 and DsE91) overl apped DMelSV-A (Figure 5-2).Transmission efficiency for the D melanogaster females that were infected with DSimSV-E was statistically indistinguishable from conspecifics infected with DMelSV-A ( G 1.016, P = 0.3135). The nature of the data (het eroscedastic residuals) precludes a straightforward parametric analysis of transmission efficien cy; however, several meaningful results were clear. DSimSV-E was transmi tted at higher efficiency in D simulans than in D melanogaster, but only at G1 (Figure 5-3). Infe ction rates at each generation were nearly identical for D melanogaster infected with DSimSV-E and DMelSV-A virus: 95% confidence intervals for DSimSV-E (G1: 57.7, CI 36.4-78.9; G2: 88.9, CI 58.3-119.4; G3: 87.5, CI 42.1-132.9) overl apped with the point estimate as well as the 95% CI for 128

PAGE 129

DMelSV-A (G1: 58.3, CI 19.3-98.3; G2: 100.0, CI 100. 0-100.0; G3: 100. 0, CI 100.0100.0; see Table 5-2). As observed previously (chapter 3), the few D. m elanogaster lines that were infected with DMelSV-A t end to respond to selection for transmission efficiency more quickly than the D. simulans lines infected with DM elSV-A (Figure 5-3, Table 5-2). Components of Host Fitness as a Proxy for Virulence Two components of host fitness, fecundi ty and hatchability, were examined for three generations following infection by inject ion. Data for both fitness components were standardized to the uninfected means of the relevant host species to correct for differences in environmental heterogeneity between generations. Fecundity as a Proxy for Virulence Neither the error bars nor the confidence intervals for DSimSV-E-infected D. melanogaster and D. simulans overlap unity in G1. Thus, the virus is virulent with respect to fecundity in both hosts (that is, fe cundity is lower in infected animals than in uninfected, sham-injected conspecifics; T able 5-3, Figure 5-4). DSimSV-E is also virulent in the D. simulans host in G2, but not in D. melanogaster. Fecundity is not reduced in either host in G3. Interestingly, fecundity of DSimSV-E infected D. melanogaster is lower than that of DSimSV-E-infected D. simulans in G1 ( D. melanogaster : 0.132, CI -0.013-0.278; D. simulans : 0.480, CI 0.128-0.833), but higher in G2 ( D. melanogaster : 0.956, CI 0.2651.647; D. simulans: 0.504, CI 0.131-0.876), and in distinguishable in G3 ( D. melanogaster : 0.656, CI -0.801-2.114; D. simulans : 0.899, CI 0.099-1.698; see Table 53, Figure 5-4). Interestingly, although the small sample size (three strains in G1) precludes formal Wilcoxon signed rank te sts, an analogous examination of the data 129

PAGE 130

shows that D. sim ulans has higher fecundity in G1 for a ll three of the viral strains that were transmitted successfully in both hosts (Tabl e 5-3). This difference is not trivial, with a fecundity reduction of 32% in D. melanogaster Fecundity in D. melanogaster rebounds in the next generation, at least in the two strains that are present in both species in G2 (fecundity in D. melanogaster is more than twice as high as in D. simulans which is unchanged; Figure 5-4). Thus, al though the sample size is small, the data are consistent with rapid change between G1 and subsequent generations in D. melanogaster but not in D. simulans. D rosophila melanogaster infected with DSimSV-E had lower relative fecundity at G1 (0.132, CI -0.013-0.278) than did conspecifics infe cted with DMelSV-A (0.755 CI 0.037-1.473; i.e., standardized fecundity for DSimSV-E standard errors did not overlap with DMelSV-A; Table 5-3, Figure 5-4). Again, DSimSV-E-infected D melanogaster appear to rebound quickly, as their relative fecundity is higher than that of DMelSV-Ainfected conspecifics at G2 (0.956 CI 0.265-1.647 and 0.315 CI -0.065-0.695, respectively; Table 5-3) and is indistingu ishable at G3 (0.656, CI -0.801-2.114 and 0.678, CI .855-2.212, res pectively; Table 5-3). Errors for DSimSV-E in D. simulans do not overlap with point estimates for DMelSV-A in conspecifics, in any of the three generations though the means appear to be closer to DMelSV-A than in D. melanogaster (Table 5-3). Additiona lly, errors for the DSimSV-E-infected D. simulans do not overlap 1 in G1 or in G2 ( i.e ., fecundity is reduced in infected flies relative to uninfected flies; the virus is virulent). However, they do include 1 in G3, again suggesting that host, virus, or both have changed in such a manner as to reduce virulence (Table 5-3). 130

PAGE 131

Hatchabilit y as a Proxy for Virulence Hatchability for D. melanogaster and D. simulans infected with DSimSV-E was similar in G1 and G3 ( D. melanogaster : G1: 0.757, CI 0.036-1.479; G2: 1.443, CI 0.2922.594; G3: 0.316, CI -0.271-0.904; and D. simulans: G1: 0.709, CI 0.346-2.000; G2: 0.553, CI 0.098-1.009; G3: 0.346, CI -0.176-0.868; Table 5-4, Figure 5-5), but higher for D. melanogaster in G2. Hatchability was not decreased in D. melanogaster infected by DSimSV-E relative to sham-injected, uninfected conspec ifics in the first two generations, but was in the third ( i.e. both confidence intervals and error bar s overlapped unity; see above and Table 5-4, Figure 5-5). Su rprisingly, hatchability of D. simulans was reduced relative to uninfected controls but only at G3 (G1: 0.709, CI; 0.346-2.000, G2: 0.553, CI 0.0981.009; G3: 0.346, CI -0.176-0. 868). Thus, DSimSV-E virulent with respect to hatchability in the host in which it had evolved for 16 generations ( D. simulans ), but not in the ancestral host, D. melanogaster Further evidence that DSimSV-E was avirulent with respect to hatchability in D. melanogaster comes from the observation that our point estimate of hatchability for D. melanogaster infected by DMelSV-A (0.936) was included in both the confidence intervals and the error bars for conspecif ics infected with DSimSV-E for all three generations (above and Table 4). Hatchability for D. simulans infected by DSimSV-E was similarly indistinguishable from conspecifics infected by DMelSV-A, so the protracted period of evolution in this species did not result in rescue of virulence with respect to hatchability. 131

PAGE 132

Discussio n We tested the generality that, after a host shift, a pathogen becomes attenuated and its infectivity toward its native host is reduced. We compared the ability of the evolved virus and the unevolved ancestral virus to infect nave native hosts ( D. melanogaster ) after 16 host generations of evolution in a novel host, D. simulans The ancestral virus (DMelSV-A) and the evolved vi rus (DSimSV-E) infected injected flies at similar rates. This suggests that the in fectivity toward the native host was not attenuated, despite 16 host generations in the novel host. However, successful infection requires t he successful infection of the next host. Thus we examined whether the flies that were successfully infected by injection were able to produce infected offspring. We also tested whether native host females infected with DSimSV-E transmitted the virus to their o ffspring with lower efficiency than the flies infected with DMelSV-A, as predicted by the attenuation hypothesis (EBERT 1998 ). As seen previously (Chapter 3), D. melanogas ter infected by injection with DMelSV-A transmit the virus to their offspring at relatively low levels. Interestingly, D. melanogaster females infected by injection with DSimSV-E actually transmitted the evolved virus to their offspring at higher levels (12.1% vs 7. 0%; not statistically significant) than their conspecifics transmitted the ancestral virus. Finally, we tested whether transmission efficiency would be similar across three host generations for native hosts infected wit h either DMelSV-A (ancestral virus) or DSimSV-E (evolved virus). Infection levels were nearly identical at every generation for D. melanogaster females infected with either ancestral or evolved virus, consistent with lack of attenuation. 132

PAGE 133

Howev er, absence of attenuation does not necessarily mean absence of adaptation in the novel host. D. simulans females injected with DSimSV-E had uniformly higher infection levels at G1 than did the D. melanogaster flies that received the same virus type suggests that DSimSV-E did, in fact, adapt to the novel host but that this adaptation did not come at the cost of attenuation in the native host. Taken together, the results on the infectivity of evolved virus do not support the attenuation generality that was developed based on results from SPE studies ( EBERT 1998 ). Instead of the predicted attenuation, the evolved virus actually was more infective than the ancestral virus during the early stages of the infection. Subsequent infection rates were indistinguishable at each generation. There are few, if any, other examples where infectivity was not att enuated but instead increased when a pathogen evolved on a novel host was returned to its native host ( ANTONOVICS et al. 2002 ). We also tested the whether virul ence would be attenuated by comparing component s of host fitness in hosts infected with either ancestral or DSimSV-E (Figure 5-4 and 4-5). In D. simulans, fecundity was unaffected by vi rus type (Figure 5-4): both ancestor and DSimSV-E reduced fecundity similarly. However, in D. melanogaster DSimSV-E was more virulent than the ancestor in G1, but less virulent than the ancestor in G2 and indistinguishable in G3. Higher virulence than the ancestor in the native host is once again inconsistent with attenuation. Moreover, the fecundity cost incurred by infection with DSimSV-E was higher in D. melanogaster than in D. simulans. Thus, even though fecundity in the novel host was not significantly different for the ancestral virus and the evolved virus, the vi rus may have evolved in a way that was (in 133

PAGE 134

the context of this study, not in an abs olute sense) neutral in the novel host but deleterious on reintroduction to the native host. In contrast, hatchability of D. melanogaster infected with the ancestral virus or with DSimSV-E was indistinguishable; mor eover, hatchability wa s the same between infected and uninfected flies in this species. Interestingly, in D. simulans hatchability was adversely and similarly affected by infection with DSimSV-E and DMelSV-A. Again we did we not see attenuation. This result is contrary to the SPE conc lusion of attenuation but is not entirely novel (EBERT 1998 ). A hypervirulent bacterial insect pathogen ( Xenorhabdus nem atophila ) exhibited significant increases in virulence when evolved on a novel host and then returned to its native host ( CHAPUIS et al. 2011 ). Our results lead us to conclude t hat the predicted outcome of pathogen attenuation following a host shift is not ubi quitous, as demonstrated when the DMelSV rhabdovirus is evolved in a novel host ( D. simula ns) and reintroduced into its native host, D. melanogaster The lack of attenuation is contra ry to the predictions developed from SPE studies and may lend support to the suggestion that the experimenters in SPE studies may be inadvertently influencing at least some experimental outcomes by guaranteeing the success of the pathogens being studied ( EBERT 1998 ). Another factor that could explain our contra ry results is the use of a complex multicellu lar host. Pathogens infecting complex host are challenged different ly than pathogens infecting unicellular host and, as such, we might expect different evol utionary strategies. At the least, more work with complex hosts is nece ssary to further explore these results. 134

PAGE 135

Table 5-1. The virus lines used for cross in fection inoc ulum produ ction exhibited high virus titers. Titer Infectivity (%) DSimSV-E 91 5.18 x106100 DSimSV-E 52 3.54 x106100 DSimSV-E 155 1.77 x106100 DSimSV-E 177 1.77 x106100 Lines of virus evolved in Drosophila simulans (DSimSV-E) that exhibited high virus titers (genome copies/g of host RNA) and infect ivity (>50% of flies infected) in D. simulans after passaging in this novel host for 16 host generations. Table 5-2. Confidence intervals for per cent infection in Drosophila simulans (DSim) and D. melanogaster (DMel) adults infected with ei ther evolved (DSimSV-E) or ancestral (DMelSV-A) virus. Generation Estimate StDev StErr Upper CI Lower CI N 1 76.9 25.0 1.7 101.4 52.4 4.0 2 90.5 19.7 1.6 109.8 71.3 4.0 DSimSVE/DSim 3 94.7 15.4 1.8 112.2 77.3 3.0 1 57.7 18.8 5.2 78.9 36.4 3.0 2 88.9 22.0 7.3 119.4 58.3 2.0 DSimSVE/DMel 3 87.5 23.1 8.2 132.9 42.1 1.0 1 62.8 22.0 3.2 106.0 19.6 1.0 2 89.3 20.7 2.8 129.9 48.7 1.0 DMelSVA/DSim 3 76.3 25.6 5.9 126.6 26.0 1.0 1 58.3 20.4 8.3 98.3 18.3 1.0 2 100.0 0.0 0.0 100.0 100.0 1.0 DMelSVA/DMel 3 100.0 0.0 0.0 100.0 100.0 1.0 135

PAGE 136

Table 5-3. Confidence intervals for fecundi ty standardiz ed to the uninfected control mean in Drosophila simulans (DSim) and D. melanogaster (DMel) females infected with either evolved (DSimSV-E) or ancestral (DMelSV-A) virus. Generation Estimate StDev StErr Upper CI Lower CI N 1 0.480 0.360 0.024 0.83 0.13 4.0 2 0.504 0.380 0.031 0.88 0.13 4.0 DSimSVE/DSim 3 0.899 0.707 0.081 1.70 0.10 3.0 1 0.132 0.129 0.036 0.28 -0.01 3.0 2 0.956 0.499 0.166 1.65 0.26 2.0 DSimSVE/DMel 3 0.656 0.744 0.263 2.11 -0.80 1.0 1 0.610 0.466 0.068 1.52 -0.30 1.0 2 0.392 0.261 0.035 0.90 -0.12 1.0 DMelSVA/DSim 3 1.114 0.793 0.182 2.67 -0.44 1.0 1 0.755 0.366 0.150 1.47 0.04 1.0 2 0.315 0.194 0.052 0.69 -0.07 1.0 DMelSVA/DMel 3 0.678 0.782 0.319 2.21 -0.86 1.0 Table 5-4. Confide nce intervals for hatc hability standardized to t he uninfected control mean in Drosophila simulans (DSim) and D. melanogaster (DMel) females infected with either evolved (DSimSV-E) or ancestral (DMelSV-A) virus. Generation Estimate StDev StErr Upper CI Lower CI N 1 0.709 0.371 0.026 0.760 0.83 0.13 2 0.553 0.465 0.039 0.631 0.88 0.13 DSimSVE/DSim 3 0.346 0.461 0.058 0.461 1.70 0.10 1 0.757 0.638 0.202 1.213 0.28 -0.01 2 1.443 0.830 0.277 2.082 1.65 0.26 DSimSVE/DMel 3 0.316 0.300 0.134 0.689 2.11 -0.80 1 0.781 0.380 0.055 0.893 1.52 -0.30 2 0.557 0.419 0.060 0.677 0.90 -0.12 DMelSVA/DSim 3 0.391 0.530 0.132 0.673 2.67 -0.44 1 0.936 0.451 0.184 1.409 1.47 0.04 2 1.812 1.194 0.331 2.533 0.69 -0.07 DMelSVA/DMel 3 0.722 0.574 0.287 1.635 2.21 -0.86 136

PAGE 137

DMelSV-A DSimSV-E Mean SE DSimSV-E 177 DSimSV-E 155 DSimSV-E 91 DSimSV-E 52 Injected flies that became infected (%) 0 20 40 60 80 100 DMelSV-A/ DMel DSimSV-E/ DSim DMelSV-A/ DSim DSimSV-E/ DMel Figure 5-1. Infectivity of virus passaged in the novel host, Drosophila simulans (DSimSV-E), was not attenuated to the native host, D melanogaster Per cent of nave injected flies that became infected is shown for each inoculum x host combination. DSimSV-E and the or iginal ancestral virus (DMelSV-A) infected nave D melanogaster at similar levels ( P = 0.3605). DMelSV-A infected D simulans more effectively than D melanogaster (P < 0.0001), while DSimSV-E and DMelSV-A infected D simulans or D melanogaster similarly ( P = 0.3605 and P = 0.079, respectively). 137

PAGE 138

Transmission to G1(%) 0 20 40 60 80 100 DMelSV-A DSimSV-E Mean SE DSimSV-E 177 DSimSV-E 155 DSimSV-E 91 DSimSV-E 52 DMelSV-A/ DMel DSimSV-E/ DSim DMelSV-A/ Dsim DSimSV-E/ DMmel Figure 5-2. Transmission efficiency of virus evolved in Drosophila simulans (DSimSV-E) was not attenuated to the native host, D melanogaster Per cent of infected flies that successfully infected their o ffspring is shown for each inoculum x host combination. DSimSV-E-infected f lies transmitted the virus to their offspring at significantly higher levels than the D melanogaster flies initially infected with ancestral virus (DMelSV-A, P = 0.0104). DSimSV-E and DMelSV-A infected D melanogaster females transmitted virus to their offspring at similar levels ( P = 0.3135) 138

PAGE 139

Generation 123Infected adults (%) 20 40 60 80 100 DSimSV-E/DMel DMelSV-A/DMel DSimSV-E/DSim DMelSV-A/DSim Figure 5-3. The ability to produce increasing in fection levels in successive generations was not attenuated when virus evolved in Drosophila simulans (DSimSV-E) and re-introduced to its native host, D melanogaster when compared to the performance of the unevolved ancestral virus (DMelSV-A ) in the native host. Per cent of infected adult flies shown for each generation. Infection levels were statistically indisti nguishable at each generation for D. melanogaster infected with either DSimSV-E (black circ les) or DMelSV-A (white circles). Additionally, DSimSV-E was more infective in D simulans (black triangles) than in D melanogaster but only at G1. 139

PAGE 140

Generation ( Drosophila melanogaster ) 123Fecundity (Standardized to uninfected mean) 0.0 0.5 1.0 1.5 DMelSV-A DSimSV-E Generation ( Drosophila simulans ) 123 DMelSV-A DSimSV-E Figure 5-4. Contrary to predi ction, the virus passaged in Drosophila simulans (DSimSVE, white circles) appears to be mo re virulent in the native host ( D melanogaster ) than the ancestral vi rus (DMelSV-A, black circles). Fecundity standardized to the control mean is show for D simulans and D melanogaster infected with DSimSV-E or DMelSV-A. DSimSV-E-infected D. melanogaster had significantly lower fecundity at G1 and conspecifics infected with DMelSV-A. DMelSV-A-infected D melanogaster appear to rebound quickly as their fecundity is significantly higher that DSimSV-E infected conspecifics at G2 ; no diffe rence in fecundity was observed for these two groups at G3. Fecundity of DSimSV-E and DMelSV-A-infected D. simulans was statically indistinguishable at all three host generations. 140

PAGE 141

Generation (Drosophila melanogaster ) 123Hatchability (Standardized to uninfected mean) 0.0 0.5 1.0 1.5 2.0 2.5 3.0 DMelSV-A DSimSV-E Figure 5-5. Hatchabilit y was indistinguishable for Drosophila melanogaster receiving unevolved ancestral virus (black circles) or virus evolved in D. simulans (white circles). 141

PAGE 142

CHA PTER 6 CONCLUSIONS Summary of Findings Understanding the evolution of virulence continues to be an important area of research as we strive to prevent and cure diseases of humans, non-human animals, and plants. Of particular interest are the events that occur after a pathogen invades a novel host (i.e., a host shift). Most of the currently accepted theor y on the evolution of virulence following a host shift is based on data developed using single celled hosts in serial passage experiments ( EBERT 1998 ). Based on these data, it is expected that after a host shift, a pathogen will be highly virul ent to its novel host and will grow progressiv ely more attenuated (i.e., less infe ctionand replication-competent) to its native host as time passes. Studying host shifts in nature is difficult because natural host shifts happen in the absence of observation and because the sta nding genetic variation present in the pathogen population immedi ately prior to the shift is unknown. Therefore it becomes difficult to know whether the polymorphism s observed in the novel host are de novo mutations or simply genetic variants that were present in those parasites of the original host that successfully colonized the novel host. An additional complication is whether or not these ancestral variants are actually advantageous in the new host, or simply increased in frequency due to founder effects; resolution of this conundrum is beyond the scope of the current work. Additionally, the host shifts of interest (particularly those with potential to infect humans) involve comp lex multicellular hosts. To address these issues, I developed a tractable model system using Drosophila spp. and the sigma rhabdovirus, which, respectively, are good models for dipteran disease vectors and 142

PAGE 143

dipteran-carried pathogenic RNA viruses. Usi ng this system, the genetic variation in sigma virus from natural populations of the native host was determined using next generation SOLiDTM sequencing, a shift onto a novel host was induced and tracked in the laboratory, and the resulting evolved viru s was returned to its native host. Finally, virus lines that successfully colonized the novel host were sequenced. I found that the sigma rhabdov irus has lower genetic variation than is typically expected of RNA viruses and that a large part of the variation likely is due to ADAR activity. The sigma virus was more infective on the novel host than on the native host and, as predicted by serial passage studies (SPE), it was also more virulent on the novel host. Contrary to predi ction, I found no evidence of attenuation when the evolved virus was returned to its native host; in fact, the opposite was true early in the infection. Sequencing the major variant from virus stra ins that successfully infected the novel host identified some viruses that were identical or nearly identical to the ancestral virus, while other viruses that were genetically drastica lly different from t he ancestral virus. However, both types of virus successfu lly achieved and maintained high transmission efficiency on the novel host. These results i ndicate that the evol ution of virulence following a host shift in multicellular animals is more complex than what is seen in microbial systems, and indicate that this new model system can be used successfully to learn about the evolution of virulence. Adenosine-Dependent RNA Deaminases Although a large part of the genetic variat ion identified through sequencing was attributable to ADAR acti vity, a better understanding of ADAR activity within D. melanogaster is needed to determine whether or not this is an antiviral response favorable to host fitness and deleterious to viral fitness, as previously suggested 143

PAGE 144

( CARPENTER et al. 2009 ). In D. melanogaster dADAR activity helps control nervous system integrity and function in adult flies and is not necessary during development ( PALLADINO et al. 2000b ). The eggs and sperm of infected flies contain virions ( BRUN and PLUS 1980 ); however, the detailed mechanics of inter-generational and inter-cellular transmission of sigma virus are unknown. De termining whether or not dADAR-mutated virions are transmitted to the next generatio n of flies and moreover successfully replicate in them would first require dete rmining conclusively that the production of transmitted virions occurred in the absence, or at least decrease, of dADAR activity. Additionally, the major variant from the contingent of virions that are vertically transmitted within flies for 15 successive generations could be sequenced and compared to determine whether dADAR affected the transmitted viral genomes, in other words, if A-to-G changes are detectable when the viral genomes present at each generation are compared. In vertebrates, ADAR1 does help modulate the host response to infection ( GEORGE and SAMUEL 2011 ; WARD et al. 2011 ); however, no such activity has been determined for ADAR2, which is the homolog of dADAR ( GEORGE and SAMUEL 2011 ; PALLADINO et al. 2000a ). Moreover, there may be host genetic variation for dADAR activity. Further analysis will be required to determine whic h, if any, mutations in the virus genomes following 16 host generations of host evolution resulted from ADAR changes. It would be particularly interesting to attribute differential vi rulence and/or titer to the presence or absence of ADAR change s; however, this remains to be seen. Transmissibility and Virule nce Following a Host Shift The novel host ( D. simulans ) was easier to infect than the native host ( D. melanogaster ), but many lines of evidence sugges ted that selection against the virus was stronger in the novel host than the native hos t. Virus-host interactions at the cellular 144

PAGE 145

level are critical for successful co lonization of novel host lineages ( DENNEHY et al. 2007 ; PERLMAN and JAENIKE 2003 ); this is one possible explanat ion for why more infections failed on the novel hos t than on the native host. Although the virus was successful in reaching new hosts in both species, response to selection for transmission efficiency was faster on the native host and decreased more rapidly in the novel host following relaxation of selection for trans mission efficiency. Thus, it is likely that the adaptive walk required fewer steps (mutati ons) in the nave native host than in the novel host ( ANDRE and HOCHBERG 2005 ; BULL and OTTO 2005 ). This makes sense given that DMelSV is more mutationally robust than previously expected ( BRUSINI et al. 2012 in review ; CLAAS et al. 1998 ). As such, the virus would be less abl e to adapt to a novel host than to a nave native host when all else is equal ( ANDRE and HOCHBERG 2005 ; MOYA et al. 2000 ). Once successful transmission was established, the virulence of the virus on the native and hovel hosts was measured. The virus wa s virulent on both hosts but, as predicted by SPE results ( EBERT 1998 ), it was more virulent on t he novel hos t. Interestingly, in both hosts, groups with higher viral titer had higher virulence than groups with lower titer. A positive associ ation between titer and transmission efficiency has been observed for many horizontally transmitted parasites ( EWALD 1983 ; FRANK 1996 ; LENSKI and MAY 1994 ). Higher virulence is expected to be associated with lower tra nsmission, even though higher titer may be associated wit h higher transmission; accordingly, an association between titer and virulence should result in a virulence/transmission tradeoff ( LIPSITCH et al. 1996 ; SHARON et al. 1999 ). In DMelSV, a vertica lly transmitted parasite, we see that virulence decreased over time but transmission also decreased over time. These observations are not consistent with our expectation that virulence and 145

PAGE 146

transmission should be negatively associated. It is possible that we are not effectively capturing the primary cause of virulence with our m easurements of hatchability and fecundity, or that there is sufficient envir onmental variation that we are unable to distinguish signal from noise statistically, but selection (unconstrained by concerns of normality) can do so. It is also possible t hat for this system, the interaction between transmission and virulence was not a tradeoff but a compromise. Sequencing Evolved Virus The expectation from evolution of virul ence theory is that a virus evolved on a novel host will possess genetic differences that correspond to its adaptation on that host ( EBERT 1998 ). This suggests that conspecif ics pathogens should c onverge on a common genotype if given enough time. Here we saw that two major genotypes occurred in viral strains that were equally successful in the novel host. One of the genotypes was nearly indistinguishable from the ancestral virus while the second was quite different. The three strains that had t he different genotype suggest that either convergence occurred or some other mechanism was at play. It is likely that that different genotype was favored early in the infection process and became the predominant variant in the three lines were it was present. Conversely, that genotype was either absent or outcompeted in the tw o novel host lines where the major variant was very similar to the ancestral virus. In eith er case it is clear that this was not a case of simple convergence onto a main genotype. Testing Attenuation The expectation from evolution of virul ence theory is that a virus will become attenuated to its native host as a re sult of passage on a novel host ( EBERT 1998 ). Substitutions that facilitate transmission in the novel host could literally replace 146

PAGE 147

substitution s that facilitated transmission in the native host. Attenuation is thus defined as a decreased ability to infect and/or persist on the native host. Accordingly, we tested attenuation by comparing the ability of the evolved virus (16 generations on the novel host) and the unevolved ancestral virus to infect nave native hosts ( D. melanogaster ). The ancestral virus (DMelSV-A) and the evolved virus (DSimSV-E) infect injected D. melanogaster (native) flies at similar rates. Surp risingly, native host females infected with DSimSV-E transmitted the virus to their offspring with higher efficiency than the flies infected with DMelSV-A. Neither infectivity nor transmission then are consistent with the predictions of the attenuation hypothesis ( EBERT 1998 ). A similar but different results was observe in a hypervirulent bacterial insect pathogen ( Xenorhabdus nem atophila ) instead of being attenuated actually grew more virulent toward its native host ( CHAPUIS et al. 2011 ). Therefore it is clear that the attenuation results from SPE studies do not hold for all systems, further w ork is needed to broaden our understanding of pathogen attenuation in complex multicel lular organisms following host shifts. In conclusion, it is clear that the questions on standing genetic diversity within pathogen populations and questions on the evol ution of virulence and attenuation following a host shift require more attention. Although our study was small (one genotype of one novel host), it is compelling and suggests that conclusions generated in one model system may not be directly applic able to all other systems in question. Therefore, these questions need to be tested further on a larger scale using more genotypes of novel and native hosts, more s pecies of host from within the same genus as the native host, hosts from different gener a within the same family, or even using hosts from different families within the same order. With such a hierarchical approach, 147

PAGE 148

one can consider standing genetic variation in both host and parasite populations and thoroughly address questions of virulence and attenuation. T hese studies will further our understating of the processes regulating the evolution of virulence, particularly as it relates to the phylogenetic length/distance that is spanned by a host shift. Using such a design may also permit us to evaluate over what taxonomic distances generalities may be constructed, or may be believable. Incr easing our understanding in this area is critical because all of the par asites that jump into humans come from genera other than our own. In reality, only a few known case s of viral host shifts originated within the Hominidae [HIV for example ( CALVIGNAC-SPENCER et al. 2012 ; GAO et al. 1992 )] and while some zoonoses originated in mammal s [SARS, for example ( HOLMES and RAMBAUT 2004 )], diseases like influenza originated from multiple steps: from other vertebrate orders to other mammals, eit her through direct c ontact with humans or through ins ect vectors that themselves became infected ( BAIGENT and MCCAULEY 2003 ; GODDARD et al. 2003 )]. Thus far, o ur work has demonstrated that at least for this system, some of the generalities from microbial systems do not hold. Our results are important because over a quarter of the emerging infectious diseases cataloged in a 50 year period were vectorborne ( JONES et al. 2008 ); although many insects groups can vector disease, most insect vectors of significant human importance are dipterans just li ke D. melanogaster ( BEATY and MARQUARDT 1996 ). Therefore, this study paves the way for larger future studies that can include a wide range of host species within the Insecta. Such studies will provide information on the evolution of virul ence as it relates to host shifts in general 148

PAGE 149

and as it relates to pertinent taxa in parti cular, like the Culic idae, whose members are the vectors of many diseases of humans and livestock. APPENDIX A MOLECULAR BIOLOGY PROTOCOLS RNA Isolation from One Fly RNA isolation was conducted using TRIzol and following the manufacturer's guidelines (www.invitrogen.com). 1. Homogenization a. Chill the fly and put it into a clean t ube with 200 l of TRIzol (for multiple flies add 100 l per extra fly up to 1ml; up to 50 flies can be done in 1 ml of TRIzol. If doing more than one fly, ra mp all volumes accordingly (except for ethanol washes and final susp ension, those stay constant). b. Homogenize ON ICE with Kontes homogenizers until onl y the wings are distinguishable. c. Incubate the homogenized sample s for 5 min at room temp. 2. Phase separation a. Add 40 l of chloroform for the first fl y and 20 l more for each additional fly. b. Cap sample tubes securely. c. Vortex tubes for 5 seconds and incubate them at room temp erature for 2 to 3 min. d. Centrifuge the sample at no more t han 12,000 g for 10 mi n at 2 to 8C. Following centrifugation, the mixture separates into a lower red, phenolchloroform phase, an intermediate phase, and a colorless upper aqueous phase. RNA remains exclusively in t he aqueous phase. The volume of the aqueous phase is about 60% of the vo lume of TRIZOL Reagent used for homogenization. e. Transfer the aqueous phase to a fresh tube. Do not cont aminate the aqueous phase with the organic phase. 3. RNA precipitation a. Add 100 l of isopropyl alcohol (50 l more for each additional fly) to the aqueous phase and shake vigorously but do not vortex 149

PAGE 150

b. Incubate samples at -20C for 10 min. c. Centrifuge at 12,000g for 10 min at 2 to 8C. i. This should produce a gel-like pel let on the side and bottom of the tube. ii. Remove and discard the supernatant. 4. Ethanol wash a. Add 1 ml of cold (-4C) 75% et hanol (this volume remains the same whether you do 1-10 flies). b. Mix the sample by vortexing and c entrifuge at no more than 7,500 g for 5 min at 2 to 8C. RNA can be left in alc ohol at 4C for 1 week or longer at20C. c. Discard alcohol d. Quick spin the tubes and use a pipette to discard extra alcohol at bottom of tubes. 5. Re-suspension of RNA a. Air-dry the RNA pellet for 5 min. ( pop the lid and leave in rack on bench top). b. Dissolve RNA i. add 29 l of DEPC-treated RNAse-free water ii. add 1 l of RNAse inhibitor and mi x by flicking the tube (DO NOT mix by pipetting) iii. Centrifuge the tube briefly to co llect the solution at the bottom. iv. Incubate for 10 min at 55C. c. Cool the tube on ice, give a qui ck spin and store it at -80C. Reverse Transcription The whole genome of the sigma rhabdovir us was reverse transcribed using a primer complementary to the 3' end of the RNA genome (whole1F: 5 TAGAAGCATCCTCGGCTTTC 3) and Superscript III first strand synthesis kit (www.invitrogen.com). 150

PAGE 151

1. Combine 200ng of genomic RNA (4 l of 50ng/ l), 1 l of whol e1F primer, 1 l of 10mM dNTP mix and 7 l of DEPC-treated water for a final volume of 13 l. 2. Incubate this mix at 65C for 5 min and then chill on ice. 3. Combine the following items from the Invitrogen first st rand synthesis kit: 4 l of the 5X first strand buffer, 1 l of DTT, 1 l of RNAseOU T, 1 l of Superscript III RT enzyme. 4. Add the mixtures from steps #2 a nd #3 and incubate at 50C for 50 min. 5. Successful reverse transcription can be confirmed by PCR amplifying a fragment of the viral genome from each end of the viral genome using the newly synthesized cDNA as template; subsequently. The Polymerase Chain Reaction Routine PCR was performed using a bas ic master mix ( www.promega.com). 1. Reactions were composed of 12.5 l of 2X master mix, 1 l of forward and 1 l of reverse primer, 1 l of template a nd 9.5 l of DEPC-treated water. 2. Reactions were incubated in a thermal cycler for two min at 94C followed by 40 cycles of 94C for 30 seconds, 55C for one min, and 72C for one min. 3. Subsequently, the reac tions were incubated for an additional 5 min at 72C and then stored at 4C. 4. PCR reactions were electrophoresed in 1% agarose gels made with 1X Tris acetate EDTA (TAE, 96.8 g Tris, 22.84 mL Acetic Acid,14.88 g EDTA), 2L of dH20). 5. The gels were submerged in 1X TAE and electrophoresed at 80 volts for 45 min. Quantitative Polymerase Chain Reaction The N gene was amplified from each vi ral cDNA template using a tagged primer (NplusTag: 5 GCAGTATCGTGAGTTCGAGTGT CCGATGACCTGTCCGTAAC3', 22bp of non-Sigma sequence (italicized) follow ed by 20bp of Sigma-specific sequence). Virus titer quantification was achieved us ing strand-specific quantitative PCR and a Taqman primer/probe set des igned against the sigma virus N The forward primer matched the tag sequence from the RT primer (tag: 151

PAGE 152

5GCAGTATCGTGAGTTCGAGTGTt3; derived from P urcell et al. ( 2006 ) while the Taqman probe (Tprobe: 5CATG AGATGGAGGAAC TTTCTCTCCCA3') and the reverse primer (R: 5GAGTCGCAGCTTTGGAGTTC3 ) were specific to the Sigma N gene. The tagged primer was complementary only to t he viral genome and, as such, allowed for quantification of genomic RNA but not mRNA or the full length genome intermediate (both are complement ary to the genome). 1. 2 L of cDNA (equivalent to a quantity of 50ng of total RNA) were added to 25 L of Real-Time PCR master mix (Applied Biosystems, Foster City, CA), 0.9 L of both of primers at 50mM, 1.25 L of the Taqman pr obe at 10mM and 19.95 L of DEPC water for a total volume of 50 L. 2. This reaction was split into 15 l triplicates. 3. A serially diluted PCR am plicon containing the region being quantified was used for absolute quantitation using a five point serial dilution standard curve(107 to 103). 4. The reactions were amplified using a StepOnePlus RealTime PCR System (Applied Biosystems, Foster City, CA). 152

PAGE 153

APPENDIX B FLY HANDLING PROTOCOLS Drosophila Rearing All flies were reared at 24C with a 16:8 light:dark photoperiod in a Percival I36VL incubator (www.percival-scientific.com). The flie s were reared on Jazz-Mix Drosophila food (www.fishersci.com) using st andard narrow polystyrene drosophila vials. 1. The food was prepared according to the manufacturer's recommendations. a. 226.8 grams of food were mixed with 1.2 liters or water, brought to a boil and cooked at a low rolling boil for 10 min. b. The food was poured into narrow polyst yrene vials (7ml per vial) using a Wheaton Unispense (www.fisher.com pump unit and a 3mm diameter hose. The food was cooled on the bench top overnight. 2. The vials were prepared to receive flies. a. Each vial was sprinkled with some baker's yeast (Fleischmann's brand was always used; 4-6 grains). b. A piece of Kim Wipe (1/8 of a sheet ) was inserted into each vial until it reached the bottom (between the f ood and the side of the vial). c. The vial was capped with a cotto n ball and refrigerated until needed. 3. New vials were set up with flies on a 21 days schedule with constant density and sex ratio (5 + 5 ). a. Adults were added on day 1 and the females were allowed to oviposit. b. The adults were cleared from the vials on day 7. c. The next generation was se tup on day 21 (new day 1). Artificial Injection of Female Drosophila Inoculum was prepared according to Clark et al. ( 1979 ). Flies were collected such that only flies < 2 days old were in jected. This is because at this age the abdomen is still soft and will distend to accept the inoculum. All flies were injected using a Sutter 153

PAGE 154

instruments XenoWorks Analog Microinjector http://www.sutter.com/products/ product_sheets/Xenowor ks_Analog_Injector.html ) equipped with a 100 l syringe. The needle blanks used were Narishige GD-1 (S erial #08NG1025) purchased from Tritech Research, Inc. All blanks were pulled using a Kopf Model 720 vertical needle puller. Inoculum Production 1. Place 100 adults in a 1.5ml Eppendorf t ube. If you have more than 100 adults per line, make multiple homogenizations. 2. Add 500 l of cold (4C) homogeniza tion buffer (HB, 0.005M Tris, 0.25M sucrose, pH 7.5) 3. Homogenize flies using the Kontes homogenizer until the flies are disrupted and only wings are visible. Vortex the samp le as needed to dislodge flies from the bottom. 4. Centrifuge the homogenate for 15 min at1200xg and 0C and collect the supernatant. 5. Centrifuge the super natant from step 4 at 6000xg and 0C. 6. Filter through a 0.45m spin column. a. Add up to 800 l of supernatant to the column. b. Spin at 1000 x g for 1 min. c. Collect filtrate. d. Homogenate that does not pass through the column should be transferred into a clean column and re-spun. e. Aliquot the filtrate into 100 l tubes (30 l per tube). f. Store at -80C. Artificial Injection 1. The microinjector tubing, needle holder, and the needle were back-filled physiological grade mineral oil (Sig ma M8410-100ML; lot #019K0125). The system was ready for injection when t he oil meniscus within the needle moved without lag, which would have suggested an air bubble (when an air bubble is present in the system the meniscus will not move exactly when dispensing occurs and the meniscus will continue to move after the mechanical dispensing 154

PAGE 155

action has ceased. This is because of the bubble contr acting and then relaxing slowly. 2. The inoculum was thawed on ice ( and kept on ice) and the needle was backfilled. The tip of the needle was placed into the inoculum and the plunger on the syringe that was mounted on the microinjector was reversed. 3. An anesthetized fly was placed on the microscope stage with its head facing the person injection and its legs downward. This facilitated inserting the needle between the second and fourth abdominal tergites and dispensing 0.1 l of inoculum. After the inocul um was dispensed the needle was left in place for about 5 seconds. In preliminary injections it was found that when the needle is removed immediately after the inoculum is dispensed some of the inoculum would leak out of the fly. Fecundity and Hatchability Fecundity and hatchability were done us ing food tinted with green dye that was poured directly onto a st andard glass slide to produce a thin layer of food over one half of the slide. The females were allowed to oviposit, the eggs were counted and then incubated at room tem perature for 24 hours. Then the eggs that failed to hatch were counted. 1. The food was prepar ed as detailed above. 2. Green food dye (McCormick brand) from local grocery store was added (9-10 drops per 100ml of food. 3. The food was dispensed using a repeating pipette and corresponding combitips (www.eppendorf.com). One millilit er of food was dispensed at one end of each glass slide. 4. The food was allowed to set (~60 min), covered in pl astic wrap and stored at 4C until its use. 5. On the day of use the slides were l abeled and sprayed with a 4g/L yeast solution. 6. The slides were placed into a large polystyrene Drosophila vial that had been prepared earlier by the addition of 2ml of 2% agar (as the bottoms of the vials are not flat, this was necessary to prevent t he flies from moving to the side of the slide that did not have food). 7. One male and one female were added to the vial on the side of the slide that contained the food. 155

PAGE 156

156 8. The vials were kept at st andard rearing conditions (see above). 9. The slide was replaced every twelve hours for up to three days. 10. The eggs on the slides removed from t he vial were counted (fecundity measure). 11. The slides were sprayed with the yeas t solution to keep them moist, lightly covered with plastic wrap and kept at the room temperature for 24hrs. 12. After 24hrs the slides were kept at 4C and the unhatched eggs were counted as time permitted within the week following the experiment [(total eggs-unhatched eggs)/total eggs = percent hatchability].

PAGE 157

LIST OF REFERENCES ALIZON, S., 2008 Transmission-recovery trade-offs to study parasite evolution. Am Nat 172: E113-E121. ALIZON, S., A. HURFORD, N. MIDEO and M. VAN BAALEN, 2009 Virulence evolution and the trade-off hypothesis: history, current state of affairs an d the future. J Evol Biol 22: 245-259. ANDERSON, R. M., and R. M. MAY, 1979 Population biology of infectious diseases. Nature 280: 361-367. ANDERSON, R. M., and R. M. MAY, 1982 Coevolution of hosts and parasites. Parasitology 85: 411-426. ANDRE, J.-B., S. GANDON and J. KOELLA, 2006 Vaccination, within-host dynamics, and virulence evolution. Evolution 60: 13-23. ANDRE, J. B., and M. E. HOCHBERG, 2005 Virulence evolution in emerging infectious diseases. Evolution 59: 1406-1412. ANTONOVICS, J., M. HOOD and J. PARTAIN, 2002 The ecology and genetics of a host shift: Microbotryum as a model system. Am Nat 160: S40-S53. BAIGENT, S. J., and J. W. MCCAULEY, 2003 Influenza type A in humans, mammals and birds: Determinants of virus virulence, host-range and interspecies transmission. Bioessays 25: 657-671. BARBOZA, P., A. TARANTOLA, L. LASSEL, T. MOLLET, I. QUATRESOUS et al. 2008 Viroses mergentes en Asie du Sud-Est et dans le Pacifique. Md Maladies Infect 38: 513-523. BASS, B. L., 2002 RNA editing by adenosine deami nases that act on RNA. Annu Rev Biochem 71: 817-846. BASS, B. L., K. NISHIKURA, W. KELLER, P. H. SEEBURG, R. B. EMESON et al. 1997 A standardized nomenclature for adenosine deaminases that act on RNA. RNA 3: 947-949. BEACH, N. M., R. B. DUNCAN, C. T. LARSEN, X. J. MENG, N. SRIRANGANATHAN et al. 2009 Persistent infection of turkeys with an av irulent strain of turkey hemorrhagic enteritis virus. Avian Dis 53: 370-375. BEATY, B. J., and W. C. MARQUARDT, 1996 The biology of disease vectors University Press of Colorado, Niwot, Co. BENAGLIO, P., and C. RIVOLTA, 2010 Ultra high throughput sequencing in human DNA variation setection: a comparative st udy on the NDUFA3-PRPF31 region. PLoS ONE 5 BOLKER, B. M., A. NANDA and D. SHAH, 2010 Transient virulence of emerging pathogens. J Roy Soc Interface 7: 811-822. 157

PAGE 158

BRUN, G., and N. PLUS, 1980 The viruses of Drosophila pp. 625-702 in The genetics and biology of drosophila edited by M. ASHBURNER and T. R. F. WRIGHT. Academic Press, London. BRUSINI, J., Y. WANG, L. F. MATOS, L.-S. SYLVESTRE, B. M. BOLKER et al. 2012 in review Virulence evolution in a host-parasite syst em in the absence of viral evolution. Am Nat. BULL, J. J., 1994 Perspective Virulence. Evolution 48: 1423-1437. BULL, J. J., M. R. BADGETT, H. A. WICHMAN, J. P. HUELSENBECK, D. M. HILLIS et al. 1997 Exceptional convergent evol ution in a virus. Genetics 147: 1497-1507. BULL, J. J., and S. P. OTTO, 2005 The first steps in adapt ive evolution. Nat Genet 37: 342-343. BULLER, R., B. ARIF, D. BLACK, K. DUMBELL, J. ESPOSITO et al., 2005 Poxviridae in Virus Taxonomy, The VIIIth Report of the In ternational Committee on Taxonomy of Viruses. (http://www.ncbi.nlm.nih .gov/ICTVdb/Ictv/fr-fst-g.htm.). edited by M. M. FAUQUET CM, MANILOFF J, DESSELBERGER U, BALL LA. Elsevier Academic Press, Oxford, UK. BURCH, C. L., and L. CHAO, 1999 Evolution by small steps and rugged landscapes in the RNA virus phi 6. Genetics 151: 921-927. CALVIGNAC-SPENCER, S., S. A. J. LEENDERTZ, T. R. GILLESPIE and F. H. LEENDERTZ, 2012 Wild great apes as sentinels and sources of infectious disease. Clin Microbiol Infect 18: 521-527. CARPENTER, J., 2008 The evolution and genetics of Drosophila melanogaster and the sigma virus, pp. The Un iversity of Edinburgh. CARPENTER, J., L. KEEGAN, L. WILFERT, M. O'CONNELL and F. JIGGINS, 2009 Evidence for ADAR-induced hypermutation of the Dr osophila sigma virus (Rhabdoviridae). BMC Genetics 10: 75. CARPENTER, J. A., D. J. OBBARD, X. MASIDE and F. M. JIGGINS, 2007 The recent spread of a vertically transmitted virus through populations of Drosophila melanogaster Mol. Ecol. 16: 3947-3954. CHAPUIS, ., S. PAGS, V. EMELIANOFF, A. GIVAUDAN and J.-B. FERDY, 2011 Virulence and pathogen multiplication: a serial passa ge experiment in the hypervirulent bacterial insect-pathogen Xenorhabdus nematophila PLoS ONE 6: e15872. CLAAS, E. C. J., A. OSTERHAUS, R. VAN BEEK, J. C. DE JONG, G. F. RIMMELZWAAN et al. 1998 Human influenza A H5N1 virus related to a highly pathogenic avian influenza virus. Lancet 351: 472-477. CLARK, M. A., W. B. MCCRADY and C. L. FIELDING, 1979 Independent production of CO2 sensitivity by nuclear gene-Dly and a tr ansmissible agent from delayed-recovery Drosophila melanogaster Genetics 92: 503-510. COLLIGNON, P. J., 2010 A pandemic response to a disease of predominantly seasonal intensity. Med J Aust 192: 358-358. 158

PAGE 159

COMBES, C., 2005 The art of being a parasite Council of Sci. Eds. Virulent [Internet]. MS N Encarta U.S. Englis h Dictionary. [cited 2011 Jan 28]. Available from: http://encarta.msn.com/dictionar y_1861734303/definit ion.html, pp. COWPERTHWAITE, M. C., J. J. BULL and L. A. MEYERS, 2006 From Bad to Good: Fitness reversals and the ascent of deleterious mutations. PLoS Comput Biol 2: e141. CUEVAS, J. M., S. F. ELENA and A. MOYA, 2002 Molecular basis of adaptive convergence in experimental populations of RNA viruses. Genetics 162: 533-542. DARWIN, C. R., 1909-14 The voyage of the beagle, pp. in The Harvard Classics edited by C. E. ELLIOT. P.F. Collier & Sons. www.barte lby.com/29/. (Jan. 26, 2011), New York, NY. DENNEHY, J. J., 2009 Bacteriophages as model organisms for virus emergence research. Trends Microbiol 17: 450-457. DENNEHY, J. J., N. A. FRIEDENBERG, R. D. HOLT and P. E. TURNER, 2006 Viral ecology and the maintenance of novel host use. Am Nat 167: 429-439. DENNEHY, J. J., N. A. FRIEDENBERG, Y. W. YANG and P. E. TURNER, 2007 Virus population extinction via ecol ogical traps. Ecol Lett 10: 230-240. DOLGIN, E., 2010 Better forecasting urged to avoid drug waste. Nat Med 16: 5-5. DOMINGO, E., and J. J. HOLLAND, 1997 RNA virus mutations and fitness for survival. Annu Rev Microbiol 51: 151-178. EBERT, D., 1998 Experimental evoluti on of parasites. Science 282: 1432-1436. EBERT, D., and J. J. BULL, 2003 Challenging the trade-off model for the evolution of virulence: is virulence managem ent feasible? Trends Microbiol 11: 15-20. EWALD, P. W., 1983 Host-parasite relations, ve ctors, and the evolution of disease severity. Annu Rev Ecol Syst 14: 465-485. FAUQUET CM, M. M., MANILOFF J, DESSELBERGER U, BALL LA, 2005 Virus Taxonomy, The VIIIth Report of the International Committee on Taxonomy of Viruses. (http://www.ncbi.nlm.nih.gov /ICTVdb/Ictv/fr-fst-g.htm.). Elsevier Academic Press, Oxford, UK. FINE, P. E. M., 1975 Vectors and ve rtical transmission: epidemiologic perspective. Ann N Y Acad Sci 266: 173-194. FISHER, R. A., 1930 The genetical theory of natural selection Oxford University Press, Oxford., Dover, New York. FLEURIET, A., 1981a Comparison of various ph ysiological traits in flies ( Drosophila melanogaster ) of wild origin, infected or uninf ected by the hereditary rhabdovirus sigma. Arch Virol 69: 261-272. FLEURIET, A., 1981b Effect of overwintering on t he frequency of flies infected by the rhabdovirus sigma in experimental populations of Drosophila melanogaster Arch Virol 69: 253-260. 159

PAGE 160

FLEURIET, A., 1982 Transmission efficiency of the si gma virus in natural populations of its host, Drosophila melanogaster Arch Virol 71: 155-167. FLEURIET, A., 1988 Maintenance of a hereditary virus: the sigma virus in populations of its host, Drosophila melanogaster Evol Biol 23: 1-30. FLEURIET, A., 1996 Polymorphism of the Drosophila melanogaster sigma virus system. J Evol Biol 9: 471-484. FLYNN, P., 2010 The Handling of the H1N1 P andemic: More Transparency Needed, pp. in Parliamentary Assembly of the Council of Europe. http://assembly.coe.int/CommitteeD ocs/2010/20100329_MemorandumPandemie _E.pdf. FRANK, S. A., 1992 A kin selection model for the evolution of virulence. Proc R Soc Lond, Ser B: Biol Sci 250: 195-197. FRANK, S. A., 1996 Models of parasite virulence. Q Rev Biol 71: 37-78. FRANK, S. A., and P. SCHMID-HEMPEL, 2008 Mechanisms of pathogenesis and the evolution of parasite vi rulence. J Evol Biol 21: 396-404. FRASER, C., C. A. DONNELLY, S. CAUCHEMEZ, W. P. HANAGE, M. D. VAN KERKHOVE et al. 2009 Pandemic potential of a strain of influenza A (H1N1): early findings. Science 324: 1557-1561. GAO, F., L. YUE, A. T. WHITE, P. G. PAPPAS, J. BARCHUE et al. 1992 Human infection by genetically diverse SIVsm-related HIV-2 in west Africa. Nature 358: 495-499. GENTON, B., I. BETUELA, I. FELGER, F. AL-YAMAN, R. F. ANDERS et al. 2002 A recombinant blood-stage malaria vaccine reduces Plasmodium falciparum density and exerts selective pressure on parasite populations in a phase 1-2b trial in Papua New Guinea. J Infect Dis 185: 820-827. GEORGE, C. X., and C. E. SAMUEL, 2011 Host response to polyomavirus infection is modulated by RNA adenosine deaminase ADAR1 but not by ADAR2. J Virol 85: 8338-8347. GODDARD, L. B., A. E. ROTH, W. K. REISEN and T. W. SCOTT, 2003 Vertical transmission of West Nile virus by three California Culex (Diptera: Culicidae) species. J. Med. Entomol. 40: 743-746. GOULD, S. J., 1990 Wonderful Life: The Burgess Shale and the Nature of History W. W. Norton. GOULD, S. J., and R. C. LEWONTIN, 1979 Spandrels of San-Marco and the panglossian paradigm: a critique of the adaptationist program. Proc R Soc Lond Ser B-Biol Sci 205: 581-598. HARISMENDY, O., P. C. NG, R. L. STRAUSBERG, X. Y. WANG, T. B. STOCKWELL et al., 2009 Evaluation of next generation sequencing platforms for population targeted sequencing studies. Genome Biol 10. HOGENHOUT, S. A., M. G. REDINBAUGH and E.-D. AMMAR, 2003 Plant and animal rhabdovirus host range: a bug's view. Trends Microbiol 11: 264-271. 160

PAGE 161

HOLMES, E. C., and A. RAMBAUT, 2004 Viral evolution and the emergence of SARS coronavirus. Philos Trans R Soc Lond, Ser B: Biol Sci 359: 1059-1065. ISAACS, D., 2010 Lessons from the swine fl u: Pandemic, panic and/or pandemonium? J Paediatr Child H 46: 623-626. JONES, K. E., N. G. PATEL, M. A. LEVY, A. STOREYGARD, D. BALK et al. 2008 Global trends in emerging infectious diseases. Nature 451: 990-U994. KEEGAN, L. P., A. GALLO and M. A. O'CONNELL, 2001 The many roles of an RNA editor. Nat Rev Genet 2: 869-878. KELLY, H. A., 2010 A pandemic response to a disease of predominantly seasonal intensity. Med J Aust 192: 81-83. KIMURA, M., 1968 Evolutionary rate at t he molecular level. Nature 217: 624-626. KIRCHER, M., and J. KELSO, 2010 High-throughput DNA sequencing concepts and limitations. Bioessays 32: 524-536. KNIPE, D. M., and P. M. HOWLEY (Editors), 2001 Fundamental virology Raven Publishers, Ltd., New York. KOMURIAN-PRADEL, F., M. PERRET, B. DEIMAN, M. SODOYER, V. LOTTEAU et al. 2004 Strand specific quantitative r eal-time PCR to study replic ation of hepatitis C virus genome. J Virol Methods 116: 103-106. KUIKEN, T., E. C. HOLMES, J. MCCAULEY, G. F. RIMMELZWAAN, C. S. WILLIAMS et al. 2006 Host species barriers to infl uenza virus infections. Science 312: 394-397. L'HERITIER, P. H., 1958 The hereditary virus of Drosophila Adv Virus Res 5: 195-245. L'HERITIER, P. H., and F. HUGON DE SCOEUX, 1947 Transmission par greffe et injection de la sensibilite herditaire au gaz carbonique chez la Drosophile Bull Biol Fr Belg 81: 70-91. LANDER, E., and M. WATERMAN, 1988 Genomic mapping by fingerprinting random clones: a mathematical analysis. Genomics 2: 231-239. LAWRENCE, P. O., and L. F. MATOS, 2005 Transmission of the Diachasmimorpha longicaudata rhabdovirus (DlRhV) to wasp offspring: an ultrastructural analysis. J. Insect Physiol. 51: 235-241. LEHMANN, K. A., and B. L. BASS, 2000 Double-stranded RNA adenosine deaminases ADAR1 and ADAR2 have overlapping specificities. Biochemistry 39: 1287512884. LENDERBERGER ET AL., 1992 Emerging infections: Microbial threats to health in the United States, pp. Institute of Medicine, Washington. LENSKI, R. E., and R. M. MAY, 1994 The evolution of virulence in parasites and pathogens reconciliation between two competing hypotheses. J. Theor. Biol. 169: 253-265. LEVIN, B. R., 1996 The evolution and maintenance of virulence in microparasites. Emerg Infect Dis 2: 93-102. 161

PAGE 162

LEVIN, B. R., and J. J. BULL, 1994 Short-sighted evolut ion and the virulence of pathogenic microorganisms. Trends Microbiol. 2: 76-81. LEVIN, B. R., and C. S. EDEN, 1990 Selection and evolution of virulence in bacteria: an ecumenical excursion and modest suggestion. Parasitol 100: S103-S115. LEVIN, B. R., M. LIPSITCH and S. BONHOEFFER, 1999 Population Biology Evolution, and Infectious Disease: Convergence and Synthesis. Science 283: 806-809. LIPSITCH, M., M. A. NOWAK, D. EBERT and R. M. MAY, 1995 The population dynamics of vertically and horizontally transmitted paras ites. Proc R Soc Lond, Ser B: Biol Sci 260: 321-327. LIPSITCH, M., S. SILLER and M. A. NOWAK, 1996 The evolution of virulence in pathogens with vertical and horizontal transmission. Evolution 50: 1729-1741. LITTLE, T. J., D. M. SHUKER, N. COLEGRAVE, T. DAY and A. L. GRAHAM, 2010 The coevolution of virulence: tolerance in perspective. Plos Pathog 6: e1001006.doi:1001010.1001371/journal.ppat.1001006. LONGDON, B., J. D. HADFIELD, C. L. WEBSTER, D. J. OBBARD and F. M. JIGGINS, 2011 Host phylogeny determines viral persistence and replication in novel hosts. Plos Pathogens 7 LUSTIG, S., U. OLSHEVSKY, D. BEN-NATHAN, B. E. LACHMI, M. MALKINSON et al. 2000 A live attenuated West Nile virus strain as a potential veterinary vaccine. Viral Immunol 13: 401-410. MACKENZIE, J. S., K. B. CHUA, P. W. DANIELS, B. T. EATON, H. E. FIELD et al. 2001 Emerging viral diseases of Southeast Asia and the Western Pacific. Emerging Infect Dis 7: 497-504. MACKINNON, M. J., S. GANDON and A. F. READ, 2008 Virulence evolution in response to vaccination: The case of malaria. Vaccine 26: C42-C52. MATROSOVICH, M., N. ZHOU, Y. KAWAOKA and R. WEBSTER, 1999 The surface glycoproteins of H5 influenza viruses isolated from humans, chickens, and wild aquatic birds have distinguishable properties. J Virol 73: 1146-1155. MAY, R. M., S. GUPTA and A. R. MCLEAN, 2001 Infectious disease dynamics: what characterizes a successful invader? Ph ilos Trans R Soc Lond, Ser B: Biol Sci 356: 901-910. MEYERS, L. A., B. R. LEVIN, A. R. RICHARDSON and I. STOJILJKOVIC, 2003 Epidemiology, hypermutation, within-host evolution and the virulence of Neisseria meningitidis Proc R Soc Lond, Ser B: Biol Sci 270: 1667-1677. MOERDYK-SCHAUWECKER, M., S. I. HWANG and V. Z. GRDZELISHVILI, 2009 Analysis of virion associated host proteins in vesicu lar stomatitis virus using a proteomics approach. Virol J 6 MORITZ, R. F. A., C. W. W. PIRK, H. R. HEPBURN and P. NEUMANN, 2008 Short-sighted evolution of virulence in parasitic honeybee workers (Apis mellifera capensis Esch.). Naturwissenschaften 95: 507-513. 162

PAGE 163

MOYA, A., S. F. ELENA, A. BRACHO, R. MIRALLES and E. BARRIO, 2000 The evolution of RNA viruses: A population genetics vi ew. Proc Natl Acad Sci U S A 97: 69676973. MOYA, A., E. C. HOLMES and F. GONZALEZ-CANDELAS, 2004 The population genetics and evolutionary epidemiology of RNA viruses. Nat Rev Microbiol 2: 279-288. MUELLER, S., D. PAPAMICHAIL, J. R. COLEMAN, S. SKIENA and E. WIMMER, 2006 Reduction of the rate of poliovirus protein synthesis through large-scale codon deoptimization causes attenuation of vi ral virulence by lowering specific infectivity. J Virol 80: 9687-9696. MULLER, M. P., and A. MCGEER, 2007 Severe acute respiratory syndrome (SARS) coronavirus. Semin Respir Crit Care 28: 201-212. NISHIKURA, K., 2010 Functions and regulation of RNA editing by ADAR deaminases, pp. 321-349 in Annu Rev Biochem edited by R. D. KORNBERG, C. R. H. RAETZ, J. E. ROTHMAN and J. W. THORNER. OHARA, P. J., S. T. NICHOL, F. M. HORODYSKI and J. J. HOLLAND, 1984 Vesicular stomatitis virus defective interfering particles can contain extensive genomic sequence rearrangements and base substitutions. Cell 36: 915-924. ONDOV, B. D., A. VARADARAJAN, K. D. PASSALACQUA and N. H. BERGMAN, 2008 Efficient mapping of Applied Biosyst ems SOLiD sequence data to a reference genome for functional genomic applic ations. Bioinformatics 24: 2776-2777. ORR, H. A., 2005 The genetic theory of adaptation: a brief history. Nat Rev Genet 6: 119-127. PALLADINO, M. J., L. P. KEEGAN, M. A. O'CONNELL and R. A. REENAN, 2000a dADAR, a Drosophila double-stranded RNA-spec ific adenosine deam inase is highly developmentally regulated and is itse lf a target for RNA editing. RNA 6: 10041018. PALLADINO, M. J., L. P. KEEGAN, M. A. O'CONNELL and R. A. REENAN, 2000b A-to-I premRNA editing in Drosophila is primarily involved in adult nervous system function and integrity. Cell 102: 437-449. PERLMAN, S. J., and J. JAENIKE, 2003 Infection success in novel hosts: An experimental and phylogenetic study of Drosophila -parasitic nematodes. Evolution 57: 544557. PETSCHEK, J. P., M. J. MERMER, M. R. SCHECKELHOFF, A. A. SIMONE and J. C. VAUGHN, 1996 RNA editing in Drosophila 4f-rnp gene nuclear transcrip ts by multiple A-toG conversions. J Mol Biol 259: 885-890. PIALOUX, G., B. A. GAUZERE, S. JAUREGUIBERRY and M. STROBEL, 2007 Chikungunya, an epidemic arbovirosis. Lancet Infect Dis 7: 319-327. PIZZATTO, L., C. M. SHILTON and R. SHINE, 2010 Infection dynamics of the lungworm Rhabdias pseudosphaerocephala in its natural host, th ecan toad (Bufo marinus), and in novel hosts (Native Au stralian frogs). J Wild Dis 46: 1152-1164. 163

PAGE 164

PURCELL, M. K., S. A. HART, G. KURATH and J. R. WINTON, 2006 Strand-specific, realtime RT-PCR assays for quantification of genomic and positive-sense RNAs of the fish rhabdovirus, Infectious hematopoi etic necrosis virus. J Virol Methods 132: 18-24. RAMBAUT, A., D. POSADA, K. A. CRANDALL and E. C. HOLMES, 2004 The causes and consequences of HIV evol ution. Nat Rev Genet 5: 52-61. ROBERTS, D. B., 2006 Drosophila melanogaster : the model organism. Entomol Exp Appl 121: 93-103. ROKYTA, D. R., P. JOYCE, S. B. CAUDLE and H. A. WICHMAN, 2005 An empirical test of the mutational landscape model of adaptat ion using a single-stranded DNA virus. Nat Genet 37: 441-444. ROOSSINCK, M. J., 2010 Lifestyles of plant viruses. Philos Trans R Soc Lond, Ser B: Biol Sci 365: 1899-1905. ROSE, J. K., and M. A. WHITT, 2001 Rhabdoviridae: the viruse s and their replication, pp. 665-688 in Fundamental virology edited by D. M. KNIPE and P. M. HOWLEY. Lippincott Williams & Wilkins, Philadelphia, PA. SALVAUDON, L., V. HERAUDET and J. A. SHYKOFF, 2007 Genotype-specific interactions and the trade-off between host and parasite fitness. Bmc Evolutionary Biology 7 SAMUEL, C. E., 2011 Adenosine deaminases acting on RNA (ADARs) are both antiviral and proviral. Virology 411: 180-193. SCHNEIDER, D., 2000 Using Drosophila as a model system. Nat Rev Genet 1: 218-226. SCULLY, L. R., and M. J. BIDOCHKA, 2006 Developing insect m odels for the study of current and emerging human pathogens, pp. 1-9 in FEMS Microbiol Lett SHARON, L. M., I. J. MOLINEUX and J. J. BULL, 1999 Virulence evolution in a virus obeys a trade-off. Proc R Soc Lond, Ser B: Biol Sci 266: 397-404. STURTEVANT, A. H., 1920 Genetic studies on Drosophila simulans I. Introduction. Hybrids with Drosophila melanogaster Genetics 5: 488-500. TENINGES, D., F. BRAS and S. DEZLE, 1993 Genome organization of the sigma rhabdovirus: six genes and a gene overlap. Virology 193: 1018-1023. TSUKAMOTO, K., T. IMADA, N. TANIMURA, M. OKAMATSU, M. MASE et al. 2007 Impact of different husbandry conditions on contac t and airborne transmission of H5N1 highly pathogenic avian influenza virus to chickens. Avian Dis. 51: 129-132. VAN DER MEULEN, K. M., M. B. PENSAERT and H. J. NAUWYNCK, 2005 West Nile virus in the vertebrate world. Arch Virol 150: 637-657. VERMEIJ, G. J., 2006 Historical contingency and the purport ed uniqueness of evolutionary innovations. Pr oc Natl Acad Sci U S A 103: 1804-1809. VODOVAR, N., C. ACOSTA, B. LEMAITRE and F. BOCCARD, 2004 Drosophila : a polyvalent model to decipher host-pathogen in teractions. Trends Microbiol 12: 235-242. 164

PAGE 165

165 WAHL, L. M., and D. C. KRAKAUER, 2000 Models of experimental evolution: the role of genetic chance and selective necessity. Genetics 156: 1437-1448. WARD, S. V., C. X. GEORGE, M. J. WELCH, L.-Y. LIOU, B. HAHM et al. 2011 RNA editing enzyme adenosine deaminase is a restrictio n factor for controlling measles virus replication that also is r equired for embryogenesis. PNAS 108: 331-336. WAYNE, M. L., G. M. BLOHN, M. E. BROOKS, K. L. REGAN, B. Y. BROWN et al. 2011 The prevalence and persistence of sigma virus, a biparentally transmitted parasite of Drosophila melanogaster Evol Ecol Res 13: tbd. WICHMAN, H. A., M. R. BADGETT, L. A. SCOTT, C. M. BOULIANNE and J. J. BULL, 1999a Different trajectories of parallel evol ution during viral adaptation. Science 285: 422-424. WICHMAN, H. A., M. R. BADGETT, L. A. SCOTT, C. M. BOULIANNE and J. J. BULL, 1999b Different trajectories of parallel evol ution during viral adaptation. Science 285: 422-424. WOLFE, N. D., C. P. DUNAVAN and J. DIAMOND, 2007 Origins of major human infectious diseases. Nature 447: 279-283. WOODS, R., D. SCHNEIDER, C. L. WINKWORTH, M. A. RILEY and R. E. LENSKI, 2006 Tests of parallel molecular evolution in a long-term experiment with Escherichia coli Proc Natl Acad Sci U S A 103: 9107-9112. YAMPOLSKY, L. Y., C. T. WEBB, S. A. SHABALINA and A. S. KONDRASHOV, 1999 Rapid accumulation of a vertically transmitted par asite triggered by relaxation of natural selection among hosts. Evol Ecol Res 1: 581-589. YANG, H. P., and S. V. NUZHDIN, 2003 Fitness costs of Doc expression are insufficient to stabilize its copy number in Drosophila melanogaster Mol Biol Evol 20: 800-804.

PAGE 166

BIOGRAPHICAL SKETCH Luis Fernando Matos was born in the Azor es Islands (Portugal) and immigrated with his family to California when he was thirteen. Luis attended Chatom Junior High, where his English speaking ski lls improved significantly, Turlock High School, Modesto Junior College, and the Califor nia State University at Stanislaus where he studied biology and in 1999 became the fi rst member of his extended family to earn a Bachelors degree. Luis entered the masters of entomology pr ogram at Washington State University (Pullman, WA) where he examined the intera ction between a pestiferous weed and one of its biological control agents. He created an artificial diet fo r the immature stage of this insect and developed mass reari ng protocols using said diet. Luis received his Masters in entomology in May of 2002. During his ti me in Pullman Luis met Joanna Joyner and they were married in the summer of 2002. Luis entered the doctoral program in entomol ogy at the University of Florida soon after. His dissertation research focus ed on understanding the evolution of virulence following host shifts using a new model system that he helped develop. During his program Luis had the opportunity to partici pate in an NSF-funded GK-12 program for graduate students. This program convinced Luis that he w anted to teach (and conduct research) at a small university where he coul d apply the inquire-based skills he acquired in during this NSF training. Additionally, Luis became the father of Lucas David Matos in November of 2006 and successfully fought test icular cancer into remission during 2008. He successfully defended his dissertation in November of 2012. 166