UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

Characterization of nucleopolyhedroviruses infecting western (Malacosoma californicum pluviale) and forest… Cooper, Dawn M. 2001

Your browser doesn't seem to have a PDF viewer, please download the PDF to view this item.

Item Metadata

Download

Media
831-ubc_2001-0361.pdf [ 4.12MB ]
Metadata
JSON: 831-1.0089978.json
JSON-LD: 831-1.0089978-ld.json
RDF/XML (Pretty): 831-1.0089978-rdf.xml
RDF/JSON: 831-1.0089978-rdf.json
Turtle: 831-1.0089978-turtle.txt
N-Triples: 831-1.0089978-rdf-ntriples.txt
Original Record: 831-1.0089978-source.json
Full Text
831-1.0089978-fulltext.txt
Citation
831-1.0089978.ris

Full Text

Characterization of nucleopolyhedroviruses infecting western (Malacosoma californicum pluviale) and forest tent caterpillars (Malacosoma disstria) in British Columbia, and detection of sublethal infection in Held populations of forest tent caterpillars by By Dawn M. Cooper B.Sc., University of Regina, 1997 A thesis submitted in partial fulfillment of the requirements for the degree of Master of Science In the Faculty of Graduate Studies (Department of Zoology) We accept this thesis as conforming to the required standard The University of British Columbia October 2001 © D a w n Cooper, 2001 In presenting this thesis in partial fulfilment of the requirements for an advanced degree at the University of British Columbia, I agree that the Library shall make it freely available for reference and study. I further agree that permission for extensive copying of this thesis for scholarly purposes may be granted by the head of my department or by his or her representatives. It is understood that copying or publication of this thesis for financial gain shall not be allowed without my written permission. Department of 2-(Or? In The University of British Columbia Vancouver, Canada DE-6 (2/88) ABSTRACT Viral diseases are an important feature of fluctuating lepidopteran populations. Epizootics of baculovirus disease can kill large numbers of hosts, making them attractive agents for biological control. The drive to improve virus efficacy by genetic manipulation has focused attention on the behaviour of baculoviruses in nature. I examined the genetic diversity present in nucleopolyherovirus (NPV) populations infecting tent caterpillars in British Columbia. Natural baculovirus isolates are genetically diverse in many different lepidopteran systems. I used R E N analysis to survey genetic diversity in six different N P V populations infecting the western tent caterpillar, Malacosoma californicum pluviale (Dyar). In total, 14 genetic variants were found among the six populations sampled. Genetic variation was structured at the population level. Bioassays detected subtle differences in pathogenecity associated with different N P V populations, suggesting that local genetic variation in N P V could influence local host dynamics. The use of baculoviruses as biopesticides has emphasized a need to determine host range for a number of different viruses. I characterized the NPVs infecting two closely related host species, the western tent caterpillar, Malacosoma californicum pluviale (Dyar), and the forest tent caterpillar, Malacosoma disstria (Hiibner), and determined the cross infective potential of each virus. M. c. pluviale N P V (McplNPV) and M.-disstria N P V (MadiNPV) were genetically distinct. Cross infection bioassays demonstrated that 3 r d instar M. c. pluviale larvae were not susceptible MadiNPV. In the reciprocal cross infection, only a small proportion of M. disstria larvae was susceptible to McplNPV. The majority of mortalities were the result of an unexpected MadiNPV infection. P C R failed to detect MadiNPV contamination in the M c p l N P V stock. Laboratory contamination and contamination of egg mass surfaces were unlikely. I suggest the virus was present as either a latent or sublethal infection that was vertically transmitted from females to larvae within the eggs themselves. These findings suggest that N P V may influence host dynamics. First, genetic variability may be associated with life history traits important for virus survival and may therefore influence local host dynamics. Second, the detection of a latent or sublethal infections provides a mechanism for both enhancing virus dispersal and maintaining virus during low host densities. ii TABLE OF CONTENTS ABSTRACT.. . . II T A B L E O F CONTENTS HI LIST O F TABLES V LIST OF FIGURES VI A C K N O W L E D G E M E N T S VII 1.0 G E N E R A L INTRODUCTION 1 1.1 INTRODUCTION 1 1.2 STATEMENT OF PURPOSE 1 1.3 VIRUS AND TENT CATERPILLAR NATURAL HISTORY 2 2.0 M O L E C U L A R CHARACTERIZATION OF T H E NUCLEOPOLYHEDROVIRUS (NPV) INFECTING WESTERN T E N T CATERPILLARS, MALACOSOMA CALIFORNICUM PLUVIALE (DYAR), IN SOUTHWESTERN BRITISH COLUMBIA 4 ABSTRACT 4 2.1 INTRODUCTION 5 2.2 METHODS 7 Collection and Rearing of NPV infected larvae 7 Extraction of McplNPVfrom Infected Larvae 8 Characterization of McplNPV genome 8 Bioassays 9 Statistical Analysis 10 Bioassays 10 F statistics 11 2.3 RESULTS 12 Incidence Data 12 Restriction Enzyme Analysis (REN) 12 Genetic Variation Hindll l 13 i i i Hierarchical analysis of virus diversity 13 Comparisons between years '. 14 Genetic variation with EcoRI 14 Bioassay results 15 REN analysis of progeny virus from mixed genotype bioassays 15 2 . 4 DISCUSSION 1 6 3.0 EVIDENCE FOR THE ACTIVATION OF LATENT OR SUBLETHAL NPV INFECTION IN FIELD POPULATIONS OF THE FOREST TENT CATERPILLAR, MALACOSOMA DISSTRIA (HUBNER) 34 ABSTRACT • 3 4 3.1 INTRODUCTION 3 5 3 . 2 METHODS 3 8 Insects 38 Virus • 38 Bioassays 38 Virus purification and restriction endonuclease analysis 39 PCR amplification ofMdNPV specific sequences 40 Statistical Analysis 40 3 .3 RESULTS 4 1 Genetic comparison of McplNPV and MadiNPV 41 Bioassays 41 M. disstria larvae: MadiNPV challenge 41 Cross infection: M. disstria larvae: McplNPV challenge 42 M. californicum pluviale larvae: McplNPV challenge 42 Cross infection M. c. pluviale larvae: MadiNPV challenge 4 2 Bioassays: examination of progeny virus 4 2 PCR examination of McplNPV stock 43 3 . 4 DISCUSSION 4 3 4.0 GENERAL CONCLUSIONS 57 4 .1 SUMMARY OF FINDINGS 5 7 4 . 2 FUTURE RESEARCH 5 8 REFERENCES 60 APPENDIX 1.0 THE INFLUENCE OF HOST PLANT ON BACULOVIRUS EFFICACY 69 APPENDIX 2.0 MOLECULAR CHARACTERIZATION OF THE NUCLEOPOLYHEDROVIRUS INFECTING THE FOREST TENT CATERPILLAR, MALACOSOMA DISSTRIA (HUBNER), IN PRINCE GEORGE BRITISH COLUMBIA 75 iv LIST OF T A B L E S TABLE 2.1. INCIDENCE OF NPV INFECTION IN WESTERN TENT CATERPILLAR POPULATIONS IN SOUTH WESTERN BRITISH COLUMBIA IN 1998 21 TABLE 2.2. ESTIMATED SIZES (KB) OF RESTRICTION FRAGMENTS OF THE MCPLNPV GENOME. FRAGMENT SIZES AND GENOME SIZES ARE THE MEAN SIZE OF 10 RESTRICTION ENDONUCLEASE PROFILES (VIRAL ISOLATES) FROM TWO TO 6 GELS 22 TABLE 2.3. FREQUENCIES OF HINDVII GENETIC VARIANTS FROM THE 1998 MCPLNPV FIELD COLLECTED VIRAL ISOLATES. VARIANT FREQUENCIES WERE USED TO ASSESS VIRAL POPULATIONS FOR STRUCTURE 23 TABLE 2.4. F STATISTICS AND ASSOCIATED P VALUES ASSOCIATED WITH EACH OF THE SCENARIOS USED TO ASSESS THE MCPLNPV POPULATIONS IN SOUTHWESTERN BRITISH COLUMBIA FOR POPULATION SUBSTRUCTURE 24 TABLE 2.5. LD 5 0 s FOR THE JEN, SATURNA, MANDARTE, AND MONTAGUE VIRUS TYPES FROM THE BIOASSAYS CONDUCTED IN 1999 AND 2000 25 TABLE 2.6. THE REN VARIANTS COLLECTED FROM INSECTS CHALLENGED WITH MIXED 25 GENOTYPE VIRUS STOCKS 25 TABLE 3.1. VIRUS TYPE AND DOSAGE, IN OBS, USED TO CHALLENGE M. DISSTRIA AND M. C. PLUVIALE LARVAE IN ...48 THE 1999 BIOASSAYS. NUMBERS IN PARENTHESIS REPRESENT THE NUMBER INSECTS CHALLENGED AT EACH DOSE 48 TABLE 3.2. ESTIMATED REN FRAGMENT SIZES (KB) OFM. DISSTRIA NPV AND M. CALIFORNICUM PLUVIALE NPV GENOMES TABLE 3.3. DOSE AND INSTAR OF THE SIX M. DISSTRIA LARVAE SUCCUMBING TO MCPLNPV INFECTION TABLE A. 1.1 DOSES AND NUMBER OF LARVAE FED FOR ASPEN/ALDER FOOD BIOASSAYS TABLE A. 1.2. ANCOVA OF THE EFFECTS OF FOOD PLANT ON VIRUS-INDUCED MORTALITY EXPERIENCED BY 3RD INSTAR M. DISSTRIA TABLE A. 1.3. ANCOVA OF THE EFFECTS OF FOOD PLANT ON VIRUS-INDUCED MORTALITY EXPERIENCED BY 4™ INSTAR M. DISSTRIA TABLE A.2.1 FREQUENCIES OF HINDIU GENETIC VARIANTS FROM MADINPV ISOLATES COLLECTED FROM FIVE M. DISSTRIA POPULATIONS AROUND PRINCE GEORGE, BRITISH COLUMBIA 50 50 72 72 V LIST OF FIGURES FIGURE 2.1. POPULATION DENSITIES (NUMBER OF TENTS (LOG) PER YEAR) OF WESTERN TENT CATERPILLAR POPULATIONS IN SOUTHWESTERN BRITISH COLUMBIA 26 FIGURE2.2. M. C. m/vMLECOLLECTION SITES IN SOUTH-WESTERN BRITISH COLUMBIA 27 FIGURE 2.3. RESTRICTION DIGESTS (HINDIII) OF 39 OFTHE1998 FIELD ISOLATES 28 FIGURE 2.4. HINDIII RESTRICTION DIGESTS OF MANDARTE ISLAND ISOLATES FROM 1998, 1999 AND 2000 COLLECTIONS 29 FIGURE 2.5. HINDIII RESTRICTION DIGESTS OF MANDARTE ISLAND ISOLATES FROM 1998, 1999 AND 2000 COLLECTIONS 30 FIGURE 2.6. £ t o R I RESTRICTION DIGESTS OF WILD-TYPE ISOLATES OF M C P L N P V FROM MANDARTE ISLAND (VEI) , SATURNA ISLAND (VE2), AND GALIANO ISLAND (MONTAGUE PROVINCIAL PARK (VE3,4,5), ROADSIDE (VE6) AND JEN (VE7) 31 FIGURE 2.7 A,B. LEVELS OF MORTALITY EXPERIENCED BY WESTERN TENT CATERPILLARS EXPOSED TO A RANGE OF DOSES AND FOUR DIFFERENT M C P L N P V VIRUS STOCKS (JEN, SAT, M A N , MONT) AT THE THIRD INSTAR (A) AND 4™/5™ INSTAR (B) STAGE 32 FIGURE 2.8. HINDIII RESTRICTION DIGESTS OF GENETIC VARIANTS PRODUCED FROM LARVAE FED VIRUS STOCKS IN 2000 33 FIGURE 3.1. XHOI (A), E c o R I (B), HINDIII (C) R E N DIGESTS OF M A D I N P V (1) AND M C P L N P V ( 2 ) 51 FIGURE 3.2. MORTALITY EXPERIENCED BY 3R D AND 4™ INSTAR M. DISSTRIA LARVAE CHALLENGED WITH M A D I N P V (OPEN CIRCLES) AND M C P L N P V (OPEN DIAMONDS)..... 52 FIGURE 3.3. MORTALITY EXPERIENCED BY 5™ INSTAR M A D I N P V CHALLENGE WITH MADINPV(CIRCLES) AND M C P L N P V (OPEN DIAMONDS) 53 FIGURE 3.4. MORTALITY EXPERIENCED BY 3R D INSTAR M. CALIFORNICUM PLUVIALE LARVAE CHALLENGED WITH M C P L N P V (OPEN DIAMONDS) AND M A D I N P V COLOURED CIRCLES) 54 FIGURE 3.5. HINDIII R E N DIGESTS OF PROGENY VIRUS FROM 11 M. DISSTRIA LARVAE SUCCUMBING TO VIRAL INFECTION DURING M C P L N P V CROSS INFECTIONS 55 FIGURE 3.6. EXAMINATION OF M C P L N P V STOCK FOR M A D I N P V CONTAMINATION USING P C R 56 FIGURE A . 1. MORTALITY EXPERIENCED BY 3R D AND 4™ INSTAR M. DISSTRIA LARVAE CHALLENGED WITH M A D I N P V ON EITHER ASPEN (•••) AND ALDER (O) 74 FIGURE A.2.1. XHOI RESTRCITION DISGESTSOF 1998 M A D I N P V ISOLATES 78 FIGURE A.2.2. RESTRICTION DIGESTS (ECORI) OF M A D I N P V ISOLATES COLLECTED FROM FIELD POPULATIONS OF M. DISSTRIA NEAR PRINCE GEORGE BRITISH COLUMBIA 7' FIGURE A.2.3. HINDIII RESTRICTION DIGESTS OF M A D I N P V ISOLATES COLLECTED FROM FIVEM. DISSTRIA SITES AROUND PRINCE GEORGE BRITISH COLUMBIA 8i vi A C K N O W L E D G E M E N T S I would like to thank my supervisor, Dr. Judy Myers, for inspiring my interest in disease ecology and for her intellectual and practical support during my studies. I would also like to thank Judy for her tremendous patience while I slowly came to the realization, after many broken bones, sprains and strains, that I wasn't immortal. I am grateful to my research committee, Drs. Martin Adamson, Sally Otto, and David Theilmann for their ideas and direction during the writing of this thesis and the course of my studies. Specifically, I would like to thank David Theilmann for answering my many questions and emails, and for providing logistical support, without which this project would not have succeeded. Special thanks to Sally Otto for being a rich resource and an excellent role model. Many other people were of great help to me over the course of my studies. Special thanks to Dr. Rick Taylor and his lab crew (Janelle Curtis, Steve Latham, Dave OBrien, Mike Stamford, and Cole Burton) who provided me with lab and freezer space, and who helped me get my molecular work off the ground. Special thanks also go to Dr. Martin Adamson and his crew (Amanda Brown, Tammy Laberge, and Allyson Miscampbell) who also provided lab space and tremendous amounts of technical support. Drs. Jenny Cory and Maynard Milks were an invaluable resource and went beyond the call of duty to help me with both the molecular and statistical aspects of my thesis work. Jenny Cory was especially helpful with new ideas and comments on this thesis. Thanks to the many graduate and undergraduate students in Judy's lab (Leonardo Frid, Madelon Denoth, Alida Janmaat, Penny Liu , Shezadi Zara, Jessica Ware, Ilia, Cody, and Kendra Coutts). Many of these students helped me collect and rear caterpillars. Special thanks to Leo Frid for having to endure thousands of caterpillars and many bioassays with me. Many thanks to my family (Gordon, Iris and Garett Cooper) and friends who have supported me over the past four years. Thanks to Allyson Miscampbell, Leanna Warman, Joel Sawada, and John Hagen for helping me keep things in perspective. And special thanks to Steve Martell and Carl Walters for encouraging me to not accept second best. Steve was a tremendous help in putting this document together. vii Lastly, thanks to folks at the Z C U , Alistair Blanchford and Jens Haeusser for help with all my computer problems. 1.0 G E N E R A L I N T R O D U C T I O N 1.1 INTRODUCTION Nucleopolyhedroviruses (NPVs) are highly virulent pathogens commonly associated with collapsing populations of many forest Lepidoptera (Myers, 1988). Viral disease is both predicted and observed at high host densities which suggests that pathogens may regulate host dynamics (Anderson and May, 1981). Epizootics can kill large numbers of hosts within weeks, making them attractive agents for biological control (Cory, 2000). The possibility of improving N P V efficiency by genetic manipulation has focused investigations on natural N P V populations. Natural N P V populations are found to be genetically heterogeneous at number of different ecological scales (Shapiro et al., 1991). Heterogeneity appears to be maintained over successive host generations, suggesting that variation may be important for virus survival (Gelernter and Federici, 1986). Although the mechanisms producing variation are well known, the effects of mixed genotype infections are not well understood. Examining variation and its role in the biology of NPVs may provide insight into virus-host evolution and may help to identify more effective strains for biological control. The use of genetically modified baculoviruses as bio-pesticides has prompted the need to determine host ranges for a number of different viruses, to allow for predictions regarding the effects of virus on non-target species. Currently, predicting effects on non-target species is difficult because there is no discernable pattern for baculovirus host range. Baculoviruses isolated from different hosts vary in the number of host species they can infect; some appear to be species-specific, where as others appear to infect a range of hosts. In general, major commercial targets for baculoviruses are generally crops affected by groups of pests, therefore a baculovirus with a broad range is preferable. This further emphasizes the need for extensive host range testing. 1.2 S T A T E M E N T O F PURPOSE The overall goal of this thesis is to examine the natural variability present in the natural N P V populations isolated from two tent caterpillar species, the western tent caterpillar, Malacosoma californicum pluviale (Dyar) and the forest tent caterpillar, Malacosoma disstria (HUbner), and to determine the cross infective potential of each virus. I have divided my thesis into two chapters. Chapter one, called " Molecular characterization of the nucleopolyhedrovirus (NPV) infecting western tent caterpillars, Malacosoma californicum pluviale, in southwestern British Columbia", examines the NPVs isolated from five western tent caterpillar sites in southwestern British Columbia. Genetic variation is examined using restriction endonuclease analysis (REN). The distribution of genetic variants is then used to examine the virus populations in southwestern British Columbia for genetic population structure. Bioassays were used to detect phenotypic differences among virus isolates from different host populations. Chapter two is titled "Evidence for the activation of latent or sublethal N P V infection in field populations of the forest tent caterpillar, Malacosoma disstria (Hiibner)", compares the NPVs isolated from western and forest tent caterpillars. R E N analysis was used to characterize the genomes of both McplNPV and MadiNPV, and bioassays were used to determine the cross infective potential of each virus. Examination of the progeny virus from M. disstria larvae cross infected with M c p l N P V revealed an unexpected MadiNPV infection. Contamination of the M c p l N P V stock used to challenge M.disstria larvae and both laboratory contamination and contamination of egg mass surfaces was unlikely, suggesting the infection was present within the M. disstria larvae themselves. There are two appendices to my thesis. Appendix one, entitled "The influence of host-plant of baculo virus efficacy" examines the effects different food plants may have on virus-host interactions. Appendix two is entitled "Molecular characterization of the nucleopolyhedovirus (NPV) infecting forest tent caterpillars, Malacosoma disstria, in the Prince George region of British Columbia". In this appendix, I use R E N analysis to examine N P V isolates from five M. disstria sites around Prince George, British Columbia. A n overall summary of important results and implications for further study are in given in the "General conclusion" section. 1.3 V IRUS AND TENT CATERPILLAR NATURAL HISTORY Nucleopolyhdedrosis viruses (NPV) are members of the Baculoviridae family, a group of naturally occurring insect pathogens reported in more than 400 species, the majority of which occur in Lepidoptera. NPVs are large double-stranded D N A viruses in which rod shaped nucleocapsids are either singly or multiply embedded within a membranous envelope and are occluded within a protein matrix, polyhedrin, to form occlusion bodies (OBs) (Miller, 1997). . Infection occurs through ingestion of O B contaminated foliage. Polyhedra are quickly dissolved 2 in the alkaline p H of the insect mid-gut to release virions, which pass through the peritrophic membrane and infect mid-gut epithelial cells (Washburn et al., 1998). Infection occurs by the spread of virions along the trachea and other susceptible tissues. Later stages of infection are marked by rupture of the host cuticle to release OBs into the environment. Transmission within a host generation occurs primarily through environmental contamination (leaf material, bark, tent material, soil, or egg mass surfaces). Transgenerational transmission may be a combination of environmental contamination and the transfer of virus from parent to offspring (transovarial virus transmission) (Kukan, 1999). The western and forest tent caterpillars (M. c. pluviale (Dyar) and M. disstria (Hiibner) are common defoliators of deciduous trees in British Columbia. M. c. pluviale feeds primarily on red alder, Alnus rubra, as well as ornamental trees such as black cherry and crab apple. M. disstria feeds on trembling aspen, Populus tremuloides. Tent caterpillars have one generation per year with females laying eggs (100-300) in a single egg mass on the host tree after adult emergence and mating in midsummer (Myers, 2000). Both species are gregarious and feed in family groups. M. c. pluviale family groups (all individuals from one egg mass) build communal silken tents which provide protection and a surface on which to bask in the sun. M. disstria does not build silken tents but does form family groups on the trunk of a tree. Tent caterpillar populations typically experience cyclic dynamics which follow a pattern of 3 to 4 years at low host densities, 2 to 3 years of increasing densities and 2 to 3 years at peak densities (Myers, 1988). N P V is commonly observed at outbreak densities. 3 2.0 M O L E C U L A R C H A R A C T E R I Z A T I O N O F T H E N U C L E O P O L Y H E D R O V I R U S (NPV) I N F E C T I N G W E S T E R N T E N T C A T E R P I L L A R S , MALACOSOMA CALIFORNICUM PLUVIALE (DYAR), IN S O U T H W E S T E R N BRITISH C O L U M B I A ABSTRACT Natural baculovirus isolates are shown in many species to be genetically diverse. Here we used restriction endonuclease (REN) analysis and bioassays to examine the genetic diversity of Malacosoma californicum pluviale N P V (McplNPV) among host populations in southwestern British Columbia. The population dynamics of M. californicum pluviale are well known and epizootics of N P V infection occur periodically. In total, 14 genetic variants were found among 39 isolates examined using Hindlll enzyme. Genetic variation was found within populations and within islands. Variant frequencies were used to examine genetic variation for population structure. M c p l N P V populations were found to be structured on two levels. Viral isolates within host family groups were more similar than isolates from other family groups, and isolates within populations were more similar than isolates from other populations. The occurrence of certain variants in more than one population and on more than one island suggests the natural movement of virus between host populations. The virus with the lowest pathogenicity (measured here as L D 5 0 ) was derived from caterpillars at Montague Provincial Park. The Montague host population is thought to be a sink population because it only occurs during peak densities in surrounding areas. It is likely that virus was brought with immigrating females and was vertically transmitted to larvae through egg masses. Vertically transmitted virus might be associated with reduced pathogenicity and sublethality. Whether selection occurs on the genetically diverse virus as host density changes will be investigated over the next cycle of tent caterpillar populations. 4 2.1 INTRODUCTION Nucleopolyhedroviruses (NPV) are naturally occurring insect pathogens that primarily infect Lepidoptera. Epizootics of viral disease are often observed during high host densities, which suggests that pathogens may play a role in regulating insect dynamics (Cory et al., 1997; Anderson and May, 1980). Viral life history traits, such as virulence, viral load (amount of virus produced) and persistence, in conjunction with resistance development in the host will determine the effect of N P V on host population dynamics (Milks, 1997). The potential of viral disease to terminate insect outbreaks has made NPVs attractive agents for biological control. In recent years, the possibility of improving N P V efficiency by genetic manipulation (e.g. Cory et a l , 1994) has focused research on the genetics of natural virus populations. Several studies have shown that wild N P V populations can be genotypically diverse at different ecological scales. Restriction endonuclease (REN) analysis has been an efficient method of distinguishing genotypic variants and has shown that variation occurs from between different geographic regions to within single agricultural fields (Smith and Summers, 1978; Vlak and Groner, 1980; Gettig and McCarthy, 1982; Allaway and Payne, 1983; Cherry and Summers, 1985; Shapiro et al., 1991; Ebling and Kaupp, 1995). Virus populations within single host individuals have also been shown to be genotypically diverse (Knell and Summers, 1981; Maeda et al., 1990). Baculovirus studies have examined field populations for variation at either large ecological scales (across regions, countries, or states) or smaller scales (within single populations or individual larvae). Few studies have compared the degree of variation among various ecological levels and examined field populations for population substructure. The ability to discern geographic patterns allows inferences to be made about the migration of adult moths and the movement of virus within the environment. Recombination, point mutations, D N A insertions and deletions, and acquisition of host D N A are common in baculovirus genomes (Crozier et al., 1988; Crozier and Ribero, 1992; Martin and Weber, 1992), however, the extent to which each of these factors contributes to variability in nature is unknown (Crozier and Ribero, 1992). Genetic differentiation may also result from the natural selection of different variants in different populations, chance differences in genotypic frequencies among the different founders of each population, genetic drift, or any of the above combined. Field studies of the natural variability of Spodoptera exigua N P V indicate that genetic heterogeneity is maintained through successive host generations (Gelernter and 5 Federici, 1990; Munoz and Caballero, 2000). The persistence and maintenance of minority genotypes in natural populations indicates that heterogeneity may be important for virus survival, but few studies have focused on the phenotypic consequences of such genetic variability. Changes in viral phenotype such as virulence or the amount of virus produced may alter viral-host interactions. Phenotypic differences in traits that affect viral fitness will select genotypes with the highest fitness and may in turn alter host population dynamics. Understanding the role of genetic variability is essential to understanding baculovirus evolution and may aid in the development of more realistic population models and biological control agents. Here we examine virus isolates from 12 populations of western tent caterpillars (Malacosoma californicum pluviale (Dyar)) to determine if N P V varies genetically and phenotypically among host populations. M. c. pluviale is a common defoliator of deciduous trees, particularly red alder, Alnus rubra, in northwestern North America. This species has one generation per year with first instar larvae hatching in early to mid-April. Oviposition occurs in mid-summer when adult females lay all their eggs in one egg mass. M. c. pluviale larvae are gregarious and feed and build silken tents as family groups. Tents provide the caterpillars with protection from predators and rain and a surface on which to bask in the sun. These conspicuous tents make it possible to detect caterpillars at both low and high densities. M. c. pluviale populations have been monitored in southwestern British Columbia since 1976 (Myers, 1990, 2000). Caterpillar populations occur on the southern Gulf Islands (Galiano, Mandarte, and Saturna Islands), on Vancouver Island (Sydney), and on the mainland (Westham and Cypress mountain). Populations cycle through peak densities every 6-11 years (Myers, 2000). The population trends for four M. c. pluviale populations over one cycle are shown in Figure 2.1. N P V is commonly observed in M. c. pluviale populations at outbreak densities (Wellington, 1960; Myers, 1990; Myers and Kukan, 1996; Myers, 2000). The Gulf Island populations have shown increased levels of infection with increasing population density although infection rates apparently declined on Mandarte Island before the start of the most recent population decline. N P V infection was closely related to density in mainland populations and virus was relatively high prior to the declines in 1992 and 1993 at Cypress and Westham. M. c. pluviale populations in southwestern British Columbia are unique in that their distribution in the southern Gulf Islands and south-western mainland allows for the simultaneous comparison of viral isolates at several ecological scales. The gregarious nature of tent caterpillars provides the opportunity to compare viral isolates from individuals within the same 6 family group, as well as to compare the isolates between different family groups and between different populations. Collections made over a three-year period allowed for the comparison of isolates between years. This study has three main objectives. The first is to use restriction endonuclease analysis (REN) analysis to describe the genetic variation present in wild M c p l N P V . The second is to use this molecular information to compare viral isolates at different ecological scales in order to discern geographic patterns or population subdivisions. The third is to use bioassays to assess pathogenicity of virus stocks made from four of the host virus populations. 2.2 METHODS Collection and Rearing of NPV infected larvae In the spring of 1998, larvae were collected from 12 sites in six geographic locations in British Columbia: three sites at Cypress Mountain and two sites at Westham; five Gulf Island sites; (three sites on Galiano Island, one site on Mandarte Island, and one site Saturna Island); and two sites on southern Vancouver Island (Figure 2.2). We considered viral isolates from each host site a virus population. Approximately 20 larvae were collected from each tent (family), from as many tents as possible per site in low density host populations, and to a maximum of 31 individuals per tent and 25 tents per site from high density populations. Family groups were transferred to paper coffee cups and transported to the laboratory where they were fed decontaminated alder leaves (10% bleach for 60 seconds, dH20 for 60 seconds). The number of larvae alive, as well as the number of larvae dead from virus, parasitoids, or other causes, was recorded daily. Time from infection to death is approximately six to ten days at room temperature. N P V infection was diagnosed by the physical characteristics of dead larvae such as a fragile integument that was easily ruptured to release pinkish liquid. The presence of occlusion bodies (OBs) in dead larvae was verified by microscopic inspection. A l l dead individuals were stored in Eppendorf tubes at - 2 0 ° C . N P V infection rates per site are expressed as the number of families with infected larvae / the total number of families sampled. During the springs of 1999 and 2000, a total of 19 infected individuals were collected from Mandarte Island in the manner described above. These samples provided the opportunity to examine virus from the same host population over a three year period. Both infection levels and host populations were at low densities in 1999 and 2000. Nineteen additional infected 7 individuals were collected from Galiano Island and Westham (mainland) during the spring of 2000. Extraction of McplNPV from Infected Larvae Viral D N A was extracted from individual caterpillars by first grinding larva with a plastic pestle and re-suspending the homogenate in distilled water (dH.20). Occlusion bodies (OBs) were separated from insect debris with a series of three washes and low speed centrifuge spins (1000 rpm for 30 sees). OBs were pelleted with a high speed centrifuge spin (14,000 rpm for 20 minutes) and re-suspended in 1 ml of dH_>0. Aliquots of the O B solution were then stored for bioassay experiments. Dissolution of polyhedra was carried out by treatment with an alkali lysis solution (1M N a 2 C 0 3 , 150 m M NaCl, 0.01 m M E D T A p H 10.8) at 3 7 ° C for 60 minutes. Virions were pelleted with a high speed centrifuge spin (14000 rpm for 30 minutes). D N A pellet was re-suspended in 500 u.1 of proteinase K buffer ( lOmM Tris p H 7.4, l O m M E D T A , 150 m M NaCl 0.4% SDS) and 20 ul of proteinase K (20 mg/ml) and incubated at 3 7 ° C overnight. A protein precipitation solution (Puregene™) was used to remove proteins from the D N A solution. D N A was precipitated with sodium acetate (3 M , p H 8.0) and 100% ethanol at - 2 0 ° C for 3 hours. This was followed by a wash with 70% ethanol for 1.5 hours. D N A was pelleted with a high speed spin (14,000 rpm for 30 minutes) and re-suspended in 65 u.1 of T E buffer at room temperature. D N A was stored at 4 ° C . Characterization of McplNPV genome Viral D N A (~2ug) was digested with Hindlll, EcoRl, and Xhol (Gibco B R L ) for 12 hours at 3 7 ° C using manufacturer's recommended conditions. A subset of 13 samples were digested with each of the three enzymes to determine both the size of the genome and the extent of genetic variation present in M c p l N P V isolates (isolate refers to the virus collected from one larval host). Ten samples were digested twice to ensure restriction patterns were repeatable. Restriction fragments were run on 0.7% agarose gels for 24 hours at 65 V (lx T B E ) . Both gel and running buffer (lx T B E ) contained ethidium bromide at a final concentration of 0.5%. D N A fragment size was estimated by comparison with X-Hindill molecular marker (NEB Biolabs). 8 Genomic D N A from M c p l N P V isolates was often difficult to digest with EcoRI and XhoI. The most reliable and repeatable digests were obtained with HindIII enzyme therefore the majority of restriction fragment profiles were produced with Hincflll. Restriction fragments in each profile were assigned letters A - P (Figure 3). New fragments falling between two previously labeled fragments were given a letter/number combination. Each unique fragment pattern was considered a new genetic variant. The virus isolated from one larva was only assigned a single (dominant) genetic variant. Although submolar bands were detected in a small number of the M c p l N P V R E N profiles they were not incorporated into the process of assigning variants. A small number of Mandarte Island samples allowed for the comparison of viral isolates from the same population over a three year period (1998-2000). Samples collected during the spring of 2000 (Mandarte Island (four individuals), Galiano Island (three individuals), and Westham (one individual)) were used to assess the amount of variation, present within the viral populations during low host densities. Bioassays Bioassays were conducted during the summers of 1999 and 2000 to assess phenotypic differences associated with virus isolates collected from different host populations in 1998. Differences in pathogenicity, measured here as L D 5 o (dose required to kill 50% of hosts), were compared between virus isolated from four infected host sites; Saturna Island, Mandarte Island, and two sites on Galiano Island (Montague Provincial Park and Jen). The virus stocks used as inoculum were made by combining 300 /u. 1 of OBs from four infected caterpillars from each of the host sites (1998 field samples), as described above. Stock concentrations were estimated Q from ten counts using a Nuebauer haemocytometer. The mean concentrations were 3.6x10 OBs/ml (2 .5x l0 8 - 4.8x10 8 OBs/ml), 3.2xl0 8 OBs/ml (2 .3x l0 8 - 4.1x10 s OBs/ml), 1.9xl0 8 OBs/ml (0.7xl0 8-3.2xl0 8 OBs/ml), and 2.8xl0 8 OBs/ml (1.5xl0 8-3.3xl0 8 OBs/ml) for Jen, Saturna, Montague and Mandarte respectively. Serial dilutions of 105, 106 and 107 OBs/ml were made for each virus stock using dH^O. Larvae used in the 1999 bioassays originated from egg masses collected from Pender and Galiano Islands (southern British Columbia) during the spring of 1999. Each egg mass was surface sterilized with 10% bleach (60 seconds) and rinsed with dt^O (60 seconds) to eliminate 9 any virus or bacteria that may have been present as surface contaminants. Egg masses were left to hatch at room temperature. Once larvae began to emerge, family groups were placed in 1 L paper popcorn cups where they were fed and monitored daily. Larvae were removed for dosing immediately following the 3 r d instar molt and starved for 24 hours. Each larva was individually fed an alder leaf disc (50mm2) to which a controlled amount of viral OBs were added. Caterpillars consuming less than 95% of the leaf disc in 24 hours were removed from the study. Each bioassay consisted of four virus treatments that ranged from 120 -7800 OBs/larva. Control larvae received leaf discs treated with dtLzO. The number of larvae used in each dosage treatment varied from 15 to 26 individuals. Larvae from the same dosage treatment were reared as a group in 1L paper cups. In 1999 there was no replication at any treatment level. Each bioassay was monitored for 14 days. Infected and dead individuals were removed daily. N P V infection was diagnosed as described previously. A l l virus-killed larvae were placed in individual Eppendorf tubes and stored at - 2 0 ° C . Larvae used for the bioassays in 2000 were 4 t h and 5 t h instars collected as larvae from cherry trees in the vicinity of the U B C campus. Virus dose was increased to account for the older instars used (1680-90000 OBs/larva). The number of virus doses was increased from four to six, and each dose was replicated three times. Larvae in each replicate were reared in groups of 8-12 individuals. Progeny virus from each of the bioassays conducted during 2000 was examined using R E N analysis to determine the number and type of variants produced from each of the mixed genotype virus stocks. Extraction and restriction protocols were as described above. Statistical Analysis Bioassays Bioassay data were analysed using Proc logistic in SAS (1990). Slopes and intercepts were compared as described by Collet (1991). This involved fitting three models; the first fitted model constrains all lines to have the same slopes and intercepts, the second fits a common slope but allows intercepts to vary among lines, the third model allows both the slope and intercept to vary among lines. The difference between the deviance of the third and second model has a chi-square distribution under the null hypothesis that the slopes are identical. The degrees of freedom are given by the number of parameters estimated in the third model minus the second one. A l l tests of significance were done at a=0.05. 10 LD50S and associated confidence intervals were calculated using the trimmed Spearman-Karber method (Hamilton et al., 1977) and U S E P A software (Ecological monitoring research division 1992b). F statistics The frequency of viral variants was used to examine the virus populations for structure. The genetic variation present in M c p l N P V populations was quantified using the frequencies of each genetic variant to test the null hypothesis that variants are randomly distributed. Analysis of molecular variance, A M O V A s (Arlequin (VI . 1)), were used to produce estimates of the variance components and F-statistics associated with each hierarchical level tested. We used A M O V A s to test three scenarios. The first scenario compared variant frequencies within populations to that among populations and generated an FST- The second scenario tested the importance of family groupings and generated three F statistics; the first, Fcr, is an index of the variant frequencies within populations relative to that of the total population (all variants within the region sampled), the second, F S c , is an index of the variant frequencies among the family groups within each population, and the third, F S T , is an index of the variant frequencies within family groups to that among family groups within the same population. The third scenario was used to determine how the composition of genetic variants within family groups would affect estimates at the broader levels. This scenario was divided into two parts; 3A and 3B. Scenario 3A used only one individual from those families represented by only one genetic variant. Comparisons were then made between the variant frequencies within families to the total population, which in this case were all variants within a population. In scenario 3B, we randomly chose only one individual from each family group within a population and compared those to the total population, again represented by all genetic variants within a population. Scenario 3B was run ten times. F- values, not significantly different from zero, will be interpreted as random distribution. A l l tests of significance were conducted at p=0.05. Small F -values, not significantly different from zero, will be interpreted as random distribution. A l l tests of significance were conducted at p=0.05. F-statistics calculated here are used only as an index of the distribution of variant frequencies among and within the various groups tested. These data are not estimates of nucleotide divergence, therefore, provide no information concerning relatedness of individual variants. A caveat of using the analysis for this system is that F-statistics are interpreted under 11 the assumption that restriction analysis is performed on non-recombining D N A . Here, R E N analysis was performed on total viral D N A , which is often subject to recombination (Maeda et al., 1993). Although we know recombination to be common in baculoviruses, we do not know the precise effects of recombination on variant diversity. Given that the production of certain genetic variants may be strongly influenced by recombination, the detection of population structure in this analysis will be interpreted loosely as the likelihood of sampling a given genetic variant from a given population. Population differentiation is therefore based solely on the presence or absence of viral variants. 2.3 RESULTS Incidence Data Infection by M c p l N P V was found in caterpillars from both Victoria and all Gulf Island populations (Table 2.1) in 1998. No virus was found in the mainland populations. Two sites on Galiano Island had the highest level of N P V infection (as measured by the proportion of infected tents) with Montague Provincial Park at 50%, Roadside (N) at 33%, and Jen at 30%. On Saturna Island 24% of families had infected individuals, and on Mandarte Island there were 14%. Victoria (A and B) had the lowest levels of infection (11%) among the island populations. Restriction Enzyme Analysis (REN) Variation in restriction profiles was seen with all enzymes used. Restriction fragment profiles for M c p l N P V differed from the previously published R E N patterns of other baculoviruses (Lee and Miller 1978; Smith and Summers, 1978; Kislev and Edelman, 1982; Cherry and Summers,1985; Shapiro et al., 1991; Keddie and Erlandson 1995). Restriction fragments were labeled alphabetically for ease of comparison. Approximate sizes of each restriction fragment are listed in Table 2.2. The M c p l N P V genome is approximately 122.5+/-0.13 kb. It should be noted that this approximation may be an underestimate because of difficulties in detecting fragments smaller than 1 Kb. 12 Genetic Variation H ind / / / The majority of restriction profile comparisons were made using the HindIII enzyme. R E N patterns for the 39 field isolates collected during 1998 are shown in Figure 2.3. A total of 13 genetic variants were found among the virus isolates (most common, or dominate virus isolated from a single larva) (Table 2.3). Variation was detected among isolates within the same population as well as among isolates from different populations. The most common genetic variants were V10 (found in six isolates from one population), V8 (found in five isolates from two populations), and V12 (found in five isolates from one population) while the least common variants were V4, V5 , and V6 found in one isolate each. The predominant genetic variants in each population varied (Table 2.3). Fragments common to every isolate were A , B , C , D , E , F , G , J, K , N , and O (Figure 2.3). Variation occurred primarily in three regions of the genome; G - H (6.1,5.7 kb), I-J (4.6, 4.3 kb) and the K - L - M (4.1, 3.2, 3.2 kb). The addition or deletion of most fragments could not be accounted for because bands smaller than 0.92 were difficult to detect. The exception is the loss of the H fragment, which appears to result in the gain of a 4.25 kb J+l band (see V4, V6 , and V13 in Figure 2.3). Most fragments were found in individuals from each of the host sites with the exception of the M + l band, which was specific to samples from Galiano Island. No R E N profiles were successfully obtained from the Victoria samples. Hierarchical analysis of virus diversity The distribution of HindIII viral variants was examined by pooling sample site data into several different hierarchical arrangements. A M O V A s were used to test two scenarios. The first scenario involved the comparison of variant frequencies within populations to that among populations. The analysis produced an FST of 0.334 (Table 2.4), significantly different from zero, suggesting that the distribution of genetic variants is not random and the McplNPV from host populations is structured at the population level. Visual examination of restriction profiles suggested that overall, viral isolates from larvae within single host family groups were more similar than those from other family groups (Figure 2.4). In a few populations, all viral isolates from single family groups were identical. Since sampling within populations was limited, we were concerned that the frequencies of genetic variants within family groups may artificially inflate the FST. Therefore, the second scenario was 13 used to determine the effects of family groupings on population structure by designating each population as a group (i.e all variants within a host population represent the total population) and the family groups within each host population as subpopulations. The FST produced by this grouping, 0.444, was statistically different from zero (Table 2.4) suggesting that the variants within families were more similar than variants between families. The comparison among families within a population (Fsc) was 0.262 and was also statistically different from zero. Since family effects were significant, we ran two additional scenarios. The first used only one representative from those families where all individuals produced the same genetic variant. The FST produced here was lower (0.162) but still statistically significant from zero (Table 2.4). In the second scenario 3B, we randomly chose one variant to represent each family and repeated each scenario 10 times. Although this greatly reduced our sample size, the F S T values in each of the 10 trials ranged from 0.090-0.438 (Table 2.4) and were all statistically significantly different from zero. This suggests family groupings did not affect comparisons at the population level. Comparisons between years Restriction fragment patterns for the 1999 and 2000 Mandarte Island isolates are shown in Figure 2.4. Host densities were low and infection was limited to only one family per year in 1999 and 2000. Viral variants in each family group were identical, while variants between families differed among the three years with V8 and V 9 occurring in 1998, V14 in 1999 and V3 (Figure 2.4) in 2000. R E N patterns for isolates collected from Mandarte and Galiano Islands, and Westham (mainland) in 2000 are shown in Figure 2.5. A l l Galiano Island isolates (Gal 1,2,3) are identical to V3 from the 1998 Galiano Roadside samples. The Westham isolate is identical to V10 found on Saturna Island in 1998. This suggests that even at low host densities and among relatively isolated island populations, N P V varies genetically to some degree. Genetic variation with EcoRI EcoKL digest patterns for seven genetic variants are shown in Figure 2.6. Seven variants were found from the thirteen isolates sampled. Submolar bands were present in at least three genetic variants (Ve5, Ve6, and Ve7). Twelve fragments, A , B , C , D , E , F, G , H , J, K , L , and M were common to all viral isolates. Variation occurred in the following four fragments of the genome; F (9.6 kb), I (7.4 kb), L (5.2 kb), and N (3.2 kb) region. The Jen, Roadside and Saturna isolates shared the addition of a 7.6 kb fragment (D+l). In addition, Saturna isolates had a 6.8 kb (E+l) and a 7.2 kb (E+2) addition. Many of the additions or deletions, such as the K 14 and N fragments, could not be accounted for because bands smaller than (1 kb) were often too faint to detect. Hierarchical comparisons were not performed using EcoRl variants because of limited sample sizes. Variation was found both within and between populations. No viral isolates from the same host family group were examined using EcoRl. Interestingly, samples from Montague Provincial Park appear more similar to samples from Mandarte Island than to samples from the other Galiano populations (Jen and Roadside). In contrast, the Jen and Roadside isolates were more similar to the Saturna isolates. Bioassay results Pathogenicity of N P V did not vary significantly among virus stocks (Figure 2.7). Mortality increased with increasing virus dose in all bioassays. Interestingly, mortalities greater than 50% were not observed when 3 r d instar larvae were challenged with the Montague virus stock. LD 5 oS from the 3 r d instar bioassays suggest that the Jen, Saturna, and Mandarte virus stocks were more pathogenic than the Montague virus stock (Table 2.5). The L D 5 0 f o r the Montague stock was nearly seven times greater than that of the Jen stock. A similar pattern was obtained from the 4 t h/5 t h instar bioassays. The Montague stock was again the least pathogenic, with an LD50 nearly five times greater than the Jen stock. The LD50 for the Mandarte virus stock was nearly as high as the Montague stock for 4 t h/5 t h instars (Table 2.5). REN analysis of progeny virus from mixed genotype bioassays R E N digests were used to assess the amount of variation produced during mixed genotype bioassays. Virus D N A was extracted from 10-15 infected larvae from each of the four bioassays (Jen, Saturna, Mandarte, Montague) and restricted with HindIII. A total of five R E N variants were produced from the Montague virus stock, three were produced from the Jen stock, three from the Mandarte stock and two from the Saturna stock (Table 2.6). The two variants produced by the Saturna virus stock are identical to V I and V10 from the 1998 field isolates (Figure 2.8). Both V I and V10 were equally expressed among larvae fed the Saturna virus stock (Table 2.6). V 4 , along with two new variants were produced from the Mandarte stock (Figure 2.8). Variants identical to V I , V5, V6, and V12, along with two new variants were found in larvae infected with the Montague virus stock. V10, V12, and one new variant were found in 15 larvae fed the Jen virus stock. Only a proportion of samples from each bioassay were examined, therefore the number of variants produced within each of the bioassay groups may be greater. 2.4 DISCUSSION The occurrence of genetic variants in natural baculovirus isolates is well documented (Lee and Miller, 1978; Smith and Summer, 1978; Gettig and McCarthy, 1982; Kislev and Edleman, 1982, Shapiro et al., 1991). Because N P V variation has been found within a single host (Lee and Miller, 1978; Gettig and McCarthy, 1982; Stiles and Himmerich, 1998), variation within a host population is not unexpected; however, natural populations of baculoviruses have shown varying degrees of restriction profile variation. The present study confirms that genetic variants are present within M c p l N P V wild type populations. O f all 48 M c p l N P V isolates examined (across all three years), 14 variants were identified. Genetic variation was found within and between populations as well as between islands. R E N analysis detected certain variants at more than one site and in particular, on more than one island, which suggests that M c p l N P V isolates can be moved between sites either by the movement of adult moths (Pair et al., 1986) or the dispersal of virus by birds or wind (Lautenshlager et al., 1980), or that new viral variants are constantly being produced. M c p l N P V from host populations in south-western British Columbia were found to be structured at two levels. Viral isolates within a host site were more similar than isolates from other host sites and isolates within a host family group were more similar than isolates from different family groups within the same host site. Significant population structure indicates that virus distribution was not random. If distribution were random we would expect to detect a larger proportion of all viral variants in all host sites. Significant population structure, as detected here, should be interpreted carefully. Firstly, sample sizes were limited. More intensive sampling will provide a better estimation of population structure. Secondly, these F-statistics do not reflect the relatedness of individual genetic variants and therefore can not be used to determine the relatedness of populations. Our interpretation is limited to concluding that viral isolates within populations are more similar than viral isolates from other populations based solely on the presence of viral variants. Estimates of nucleotide divergence would provide more information about the genetic relatedness of the variants between populations and may provide information concerning viral movement. 16 Our findings are similar to those reported for Spodoptera frugiperda (SfNPV) (Shapiro et al., 1991). Here, R E N analysis and nucleotide divergence was used to compare 22 SfNPV isolates over a large ecological scale (15 isolates within Lousiana to seven isolates outside Louisiana (Georgia, Colombia, Venezuela, Mexico) and a smaller scale (isolates within single agricultural fields). Isolates from within Louisiana were found to be more similar than those outside Louisiana. Significant SfNPV genetic variation was also found within single agricultural fields, though variation within agricultural fields was not structured. R E N analysis did detect foreign SfNPV isolates interspersed within Louisiana isolates suggesting that virus can be moved between sites over relatively large distances In this study, the presence of the same variants on both Mandarte and Galiano Island (Montague Provincial Park), and Saturna Island and a mainland site suggest that the mainland (Westham) and all Gulf Island populations may be within adult dispersal capabilities. The discovery of genetic variation has focused recent studies on the biological activity associated with individual variants derived from single field viral isolates (Stiles and Himmerich. 1998; Munoz et al., 1999; Hodgson et al. (in press)). In some instances, cloned variants have shown no significant difference in virulence, such as those from Spodoptera exigua N P V (SeNPV) (Munoz et al., 1999), while in others virulence was variable, e.g. Lymantria dispar N P V (LdMNPV) (Shapiro and Robertson, 1991). Stiles and Himmerich (1998) found that several A c M N P V variants exhibited lower L C 5 o (concentration lethal to 50% of subjects in the study) and L T 5 0 (time required to reach 50% mortality in test subjects ) values than control variants (most prevalent variants) in both permissive (Heliocoverpa virescens) and semi-permissive (Helicoverpa zed) hosts. More recently, Hodgson et al. (in press) examined whether differences in pine beauty moth (Panolis flammed) N P V (PfNPV) genotypes translated to different phenotypes and whether different fitness values could be associated with each phenotype. As many as 24 genotypic variants of PfNPV have been found in a single pine beauty moth larva. Hodgson et al. (in press) compared four of these genetic variants. Genotypes were found to differ in three traits predicted to be important components of viral fitness: pathogenicity (defined here as the capacity to kill the host, generally described by the LD50), speed of kill (time between initial infection and death of the host) and yield (number of OBs produced upon death of the host). The most striking differences were in pathogenicity. A 7-fold difference was found between the highest and lowest LD50. Differences were also noted in the slope of the dose-response curve. Changes in 17 speed of kill were less dramatic, with differences in the mean speed of kill being less than 36 hours between the fastest and slowest genotype. Yield of occlusion bodies was also found to vary with genotype. Subtle phenotypic differences were associated with the different M c p l N P V virus stocks used in this study. Pathogenicity, measured here as LD50, was lowest in both assays for the Montague virus stock. The Montague host population is thought to be a sink population for moths from other areas on Galiano Island or nearby Salt Spring Island (Myers, 2000). It is likely that virus came with immigrant females from high density source populations in 1995 and 1996. Sink populations are areas where local reproductive success is less than mortality and are characterized by low host densities or extinction during years with no immigration. This is the case for the Montague population where tent caterpillars go extinct between outbreaks (Figure 2.1). Virus from immigrating females is likely to be vertically transmitted from immigrating females to larvae within egg masses. Vertical transmission (passage of virus parent to offspring) is more likely with a less pathogenic virus to cause sublethal infection and may be an important mechanism for the maintenance of virus at low host densities (Myers and Rothman, 1995). Evidence for the movement of virus between sites is supported by R E N analysis. Similar genetic variants occurred on Galiano (Montague Provincial Park) and Mandarte Islands and Saturna and Mandarte Islands (Table 2.3). In 2000, a similar genetic variant occurred on both Saturna Island and the mainland (Westham). The Westham host populations were uninfected from 1997-1999, but in 2000 and 2001 there were a very small number of infected individuals (Myers 2000, personal observation). Pathogenicity, as measured here, provides only limited information about viral phenotypes, their associated fitness trade-offs and their importance in maintaining genetic variability in field populations. The study of other viral life history traits such as speed of kill and yield, may provide a more complete measure of the effects of genetic variability in field populations. A n examination of all viral life history traits in host populations at different densities could help explain differences in local population dynamics. Selection pressures acting on both the virus and host population should promote selection for viral genotypes with the highest fitness. Trade-offs between viral traits such as speed of kill and yield or pathogenicity could allow the virus to maximize fitness through changing host densities. Host densities that favor genotypes expressing a faster speed of kill are likely to experience high mortality, whereas populations favoring reduced pathogenicity will experience low mortality. 18 The maintenance of genetic variation in wild baculovirus populations suggests that heterogeneity is a common aspect of viral replication that is either not selected against or is important for virus survival. Mechanisms known to produce variation, including recombination, D N A insertions and deletions, base pair mutations and acquisition of host D N A have been well documented, but the factors that maintain variability in natural populations have not. Long-term field studies have not only shown that genetic heterogeneity is maintained, but that the relative proportion of each genotype in wild type populations seems to remain stable throughout successive generations (Gelernter and Federici, 1990, Munoz et a l , 1998). For example, the same R E N variants have been found in natural populations of S e M N P V surveyed from 1992 through 1998 (Caballero et al., 1992). If selection of the most fit genotypes is constantly occurring, we would expect to see a decrease in genetic variability. Godfray et al. (1997) suggest that at the host level, parasitic interactions can help maintain variation. When host and pathogen densities are high, hosts ingest a variety of genetic variants. Multiple virus genomes can be packaged within one O B leading to the transmission of several genotypes when an O B is ingested. In addition, multiple sites of infection can occur in the insect midgut so that many genotypes can infect separately without being co-occluded. The R E N results from our bioassays support this theory (Table 2.6). In all bioassays, more than one variant was produced among insects challenged with mixed genotype virus stock. Additionally, the presence of submolar fragments in R E N profiles of field isolates suggests that more than one genotype is being expressed in more than one individual (Figure 2.8). Levels of genetic variation can be augmented at the host population level with movement of virus between host populations. Hodgson et al. (in press) suggest that if the fitness of viral genotypes differs with varying ecological conditions, then mixing of differentially selected virus subpopulations can occur by natural movement of virus to new habitats. Both of these factors can result in high levels of local genetic variation. The mixed genotype virus stocks used in the M. c. pluviale bioassays demonstrate that local variation can be maintained with the mixing of virus subpopulations. As many as six different variants were expressed in larvae fed a mixed virus stock. This number would likely increase if more larvae were sampled. The importance of mixed infections is not well defined. Bioassays designed to test the interactions between two and more viral genotypes can better address the functional importance of mixed genotype infections. 19 In this study, we found both genetic and phenotypic differences associated with the different virus populations. The lack of sequence information available on the M c p l N P V genome makes it difficult to determine whether the specific genetic changes found here are responsible for the observed variation in pathogenicity. It is interesting to note that the £ c o R I digestion patterns found in this study appear to place Jen and Saturna into one group and Montague and Mandarte into another, similar to the pattern observed with the LDsosCFigure 2.7). Sequence information could provide some indication as to which regions of the genome are responsible for changes in LD 5 o. Understanding the role that genetic variants have in wild populations may provide insight into the evolution of baculoviruses and their hosts. It may also aid in the development of more realistic population dynamics models and more effective virus strains for biological control. Whether selection occurs on the genetically diverse virus as host density changes will be investigated over the next cycle of tent caterpillar populations. 20 Table 2.1. Incidence of NPV infection in western tent caterpillar populations in south western British Columbia in 1998. Population # infected tents Total tents # infected Total individuals (%) individuals (%) Galiano Island -Jen 6 20 (30) 49 (13) 377 -Montague 5 10 (50) 20(11) 187 -Roadside (N) 4 12 (33) 30(13) 225 Saturna Island 6 25 (24) 47 (10) 488 Mandarte Island 3 22 (14) 25 (6) 414 Victoria 3 28 (11) 6(1) 573 Westham o 37 (0) 0(0) 726 Cypress Mountain 0 23 (0) 0(0) 356 Table 2.2. Estimated sizes (kb) of restriction fragments of the McplNPV genome. Fragment sizes and genome sizes are the mean size of 10 restriction endonuclease profiles (viral isolates) from two Band HindIII EcoRl XhoI A 18.9 18.5 18.8 B 14.5 18.0 18.1 C 13.2 14.4 16.7 D 12.1 12.9 14.4 E 10.7 11.0 10.9 F 7.7 9.6 9.0 G 6.1 8.2 8.5 H 5.7 7.8 8.2 I 4.6 7.4 6.0 J 4.3 6.0 3.7 K 4.1 5.5 1.7 L 3.2 5.2 0.94 M 3.2 4.3 0.92 N 3.0 3.2 O 2.8 3.0 P 2.4 1.6 Q 1.0 1.5 Total size 114.9+/-0.12 122.5+/-0.13 118.0a a size estimates based on R E N profiles from less than five viral isolates. 22 Table 2.3. Frequencies of HindIII genetic variants from the 1998 McplNPV field collected viral isolates. Variant frequencies were used to assess viral populations for structure. Collection site # families from which data was collected Family: REN variants (a) Saturna Island 5 A:V1(1), V10(l) B:V10(3) C:V10(1) D:V9(1) E:V1(1), V10(l) Mandarte Island 2 A:V8(4) B:V9(1) Galiano Island -Montague Provincial Park 2 A:V5(1) B:V6(1), V8( l ) , V13(4) -Jen 3 A:V7(3), V I 1(2), V12(4) B:V12(1) C:V13(2) -Roadside 3 A:V1(1),V2(1), V3(1),V4(1) B:V3(2) C:V4(1) a: the number of isolates sharing the same profile 23 Table 2.4. F statistics and associated P values associated with each of the scenarios used to assess the McplNPV populations in southwestern British Columbia for population substructure. Source of Variation F statistic P value Scenario 1 Among populations: comparison of variant frequencies among populations (all variants within region sampled) 0.334 (F S T ) <0.001 Scenario 2 Among populations: comparison of variant frequencies among populations (all variants within region sampled) 0.246 (F C T ) <0.001 Among families within populations: comparison of variant frequencies among family groups within each population 0.262 ( F s c ) <0.025 Within families: comparison of variant frequencies among family groups within the same population 0.444 (F S T ) <0.001 Scenario 3a Among families within a population: comparison of variant frequencies among family groups within the same population; choosing only one individual from those families represented by only one genetic variant 0.162 (F S T ) <0.001 Scenario 3b Among families within a population: comparison of variant frequencies among family groups within the same population, randomly choosing only one variant per family 0.090-0.438 (F S T ) 24 Table 2.5. LD 5 0s for the Jen, Saturna, Mandarte, and Montague virus types from the bioassays conducted in 1999 and 2000. Virus Type Instar Log LD50 (95% confidence intervals) LD50 (#OBs) (95% confidence intervals) 1999 Jen 3 r d 3.02 (2.91,3.14) 1047 (812,1380) Sat 3 r d 3.20 (2.99, 3.43) 1585 (977,2692) Man 3 r d 3.33 (3.12,3.56) 2138 (1318,3631) Mont 3 r d 3.85 b 7079 b 2000 Jen 4 t h /5 t h 4.18 (3.85,4.52) 15136 (7079, 33113) Sat 4 t h /5 t h 4.22 (3.97, 4.48) 16596 (9705, 30200) Man 4 t h /5 t h 4.64 (4.00, 5.38) 43652 (10000, 239883.) Mont 4 t h /5 t h 4.71 (4.22, 5.25) 51286 (16596, 177828) b 95% confidence intervals could not be estimated Table 2.6. The REN variants collected from insects challenged with mixed genotype virus stocks Virus stock REN variants a Saturna Island Vl(6) , V10(6) Mandarte Island V4(5), new variant (2), new variant (2) Montague Prov Park V4(l) , V(5), V(6), V6(6), V12(4) Jen V10(l) , V12(10), new variant (3) a the number of isolates sharing the same profile 25 10000 c u Si .fi B 3 Z 1000 A Westham Montague Saturna cc o\ o c-j _ OS o\ o © o os ON O\ © o © ts c-i c-l Figure 2.1. Population densities (number of tents (log) per year) of western tent caterpillar populations in southwestern British Columbia. Populations at most sites persist even at low density with the exception of Montague, which disappears following the population crash. Galiano, Mandarte, and Saturna Islands (closed squares, closed circles, and open triangles) are believed to be source populations for Westham (open diamonds) and Montague (open circles). The Mandarte Island population (closed circles) is a persistent population that is likely to receive new individuals from Sydney, British Columbia during peak densities (Myers...). 26 -123" km 0 5 10 Figure 2.2. M. c. pluviale collection sites in south-western British Columbia. Molecular data was collected from Galiano (Roadside, Jen, Montague Prov. Park), Saturna, and Mandarte Islands, as well as Westham. 27 VI V2 V3 V4 V5 V6 V7 V8 V9 V10 VII V12 V13 L (3) (2) (3) (1) (1) (1) (3) (5) (2) (5) (4) (5) (4) Figure 2.3. Restriction digests (Hindlll) of 39 of thel998 field isolates. The fragments are labeled (A-N) on the left and side of Variant 1 (VI). New fragments are indicated by arrows and assigned new letters. Numbers in brackets represent the number of isolates sharing the same REN profile. L is the X-Hindlll molecular marker. 28 Figure 2.4. Hindlll restriction digests of Mandarte Island isolates from 1998,1999 and 2000 collections. Families 1 and 2 are represented by V8 and V9. The 1999 isolates are a new variant, V14. This is similar to V13 from Montague Provincial Park with the addition of the 5.5 Kb G band. The 2000 isolates are similar to V3 from the roadside site on Galiano Island. 29 23130 fegl B,C-94E6 ^ E " F -G 6557 H -4361 I_ J -M -2322 mm* N -2027 W * o -Man Cral-1 West Gal-2 (}al-3 Figure 2.5. HindIII restriction digests of the viral isolates collected during the spring o f 2000. The Mandarte isolates (Man) produced a new variant, V14. The Galiano Island isolates (Gal-1,2,3) were identical to V3 from the 1998 samples (Figure 1). Whestham Island (West) was represented by only one isolate which was identical to V10 from the 1998 Saturna Isalnd isolates. 30 Figure 2.6. Eco RI restriction digests of wild-type isolates of McplNPV from Mandarte Island (Vel), Saturna Island (Ve2), and Galiano Island (Montague Provincial Park (Ve3,4,5), Roadside (Ve6) and Jen (Ve7). Fragments are labeled (A-Q) along the left side of Vel. L is the X-Hindlll molecular marker. 31 a ro C o E o 5 4 3 2 1 0 -1 -2 -3 -4 -5 H -6 o Jen • Sat A Man o Mont Linear (Jen) • - - - - - Linear (Sat) • Linear (Man) — - - - Linear (Mont) 2.5 3 log dose 3.5 o E o 5 4 3 2 H -1 -2 -3 -4 •5 •6 2.5 Jen Saturna Mandarte Montague Linear (Jen) - - - - Linear (Saturna) — — L i n e a r • A 0 - Linear (Montanuet 3.5 4 4.5 log dose 5.5 Figure 2.7 a,b. Levels of mortality experienced by western tent caterpillars exposed to a range of doses and four different McplNPV virus stocks (Jen, Sat, Man, Mont) at the third instar (a) and 4 t h/5 t h instar (b) stage. At the 3 r d instar stage (a), diamonds (•) describe actual mortality and the solid line the fitted mortality due to Jen virus stock (logit(mortality)=-4.4835+1.5116(logi0dose)), squares ( ) and small dashed line Saturna virus stock (logit(mortality)=-12.458+3.7002(logiodose)), triangles (A) and large dashed line Mandarte virus stock (logit(mortality)=-7.8904+2.3634(logiodose)), and circles (o) and mixed large and small dashed line Montague virus stock (logit(mortality)=-7.46+1.7009(Iogi0dose)). Note that mortalities greater than 50% were not achieved with the Montague virus type at the third instar stage. At the 4 t h/5 t h instar stage (b), diamonds and solid line describes actual mortality and fitted mortality due to Jen stock (logit (mortality)=-11.722+2.718(logi0dose)), squares and small dashed line Saturna stock (logit(mortality)=-11.196+2.6072(logi0dose)), triangles and large dashed line Mandarte stock (logit(mortality)=-13.73+3.0384(logi0dose)), circles and mixed large and small dashed line Montague stock (logit(mortality)=-7.7351+1.5935(logi0dose)). 32 Stock virus Progeny virus Figure 2.8. HindIII restriction digests of genetic variants produced from larvae fed virus stocks in 2000. * represent new variants. Vl,4,5,6,10, and 12 are identical to variants from 1998 field samples (Figure 3). 33 3.0 E V I D E N C E F O R T H E A C T I V A T I O N O F L A T E N T O R S U B L E T H A L N P V I N F E C T I O N IN F I E L D P O P U L A T I O N S O F T H E F O R E S T T E N T C A T E R P I L L A R , MALACOSOMA DISSTRIA (HUBNER) ABSTRACT A field population of Malacosoma disstria larvae is suspected of harboring a latent or sublethal infection. The latent or sublethal virus was activated during a series of bioassays designed to test the cross-infective potential of Malacosoma californicum pluviale nucleopolyhedrovirus (NPV) (McplNPV) and Malacosoma disstria N P V (MadiNPV) on both host species. M. disstria larvae were hatched from egg masses collected from high density field populations. Restriction endonuclease (REN) analysis of virus produced from the cross infection of M. disstria with M c p l N P V indicated that only six of 50 larval mortalities were the result of the M c p l N P V used as the inoculum. The majority of mortalities were caused by a virus similar to, if not identical to, M. disstria N P V (MadiNPV). P C R was used to rule out MadiNPV contamination of the M c p l N P V stock. External contamination and laboratory contamination were unlikely because no mortalities were observed in the controls. We suggest that virus may have been present as either a latent or sublethal infection and was likely vertically transmitted from females to larvae within the eggs themselves. The discovery of latent or sublethal infections in natural populations may provide insight into how virus is maintained throughout host population cycles. Whether latent or sublethal infections are present in other Malacosoma populations at similar densities will be examined in all caterpillar life stages over the next population cycle. 34 3.1 INTRODUCTION Fluctuating populations of many forest Lepidoptera are associated with epizootics of viral disease (Myers, 1998). Viruses, such as baculoviruses, are considered potentially major factors affecting the dynamics of insect populations because they are associated with high host numbers (Anderson and May, 1981). The Baculoviridae is a family of double stranded D N A viruses composed of two genera, Nucleopolyhedrovirus (NPV) and Granulovirus (GV). NPVs form large polyhedral occlusion bodies (OBs) containing one or more enveloped nucleocapsids (virions). G V s form smaller occlusion bodies, each containing only a single viral particle within a matrix called granulin (Miller, 1997). Baculovirus infections are generally initiated following the ingestion of OBs, which dissolve in the alkaline p H of the insect midgut. Dissolution releases virions, which then infect cells lining the midgut epithelium (Washburn et al., 1998). Infection spreads from these cells to the tracheolar cells of the host respiratory system (Miller, 1997). Ultimately, nearly all host tissues become infected. Hosts that succumb to infection can often be characterized by a ruptured cuticle, through which millions of infectious OBs are released into the environment where they can be transmitted to new hosts. Epizootics of baculovirus disease can kill a large number of hosts within weeks, making them attractive agents for biological control (Anderson and May, 1981). They are cited as having a narrow host range, especially when compared to chemical pesticides. However, host range within the Baculoviridae is known to be variable (Doyle et al., 1990). Major commercial targets for baculoviruses are generally pests of crops such as cotton and vegetables, both of which are attacked by groups of species, where a baculovirus with a broad host range would be preferred (Cory et al., 2000). Baculoviruses have also been the subject of numerous genetic modifications aimed at improving viral efficacy (Tomalski and Miller, 1991; Cory et al., 1997; Burden et al., 2000). Host range testing is therefore motivated by the development of genetically modified baculoviruses and the need to assess their environmental impacts. The risks involved with releasing insect pathogens, such as baculoviruses, are two-fold. Firstly, will there be detrimental effects on non-target species, and secondly, will the gene that has been inserted into the genetically modified virus be transferred to another virus resulting in detrimental effects to non-target species (Cory, 2000). Furthermore, baculoviruses cannot be relied upon to remain within the area in which they are released (Shapiro et al., 1991). Host range determination can 35 provide some indication of which species might be at risk to the wide scale release of baculoviruses. It is well established that baculoviruses do not infect vertebrates or plants and that host range is often limited to the insect order from which they were isolated (Cory et al., 2000). However, studies have shown there to be considerable variation in the number of hosts baculoviruses can infect (Cory et al., 1997). Some appear to be species-specific (and therefore are very unlikely to be a hazard when genetically modified), whereas others infect a range of hosts. Lymantriid NPVs, such as those from Lymantria dispar, Orygia antigua, and Euproctis chrysorrhoea, have an apparently narrow host range with N P V s able to infect only a single species (Barber et al., 1993; Cory et al., 1997; Cory et a l , 2000). Barber et al. (1993) found that a total of 43 Lepidoptera species, one dipteran, and one hymenoptera were not permissive to gypsy moth Lymantria dispar N P V . Similarly, Richards et al (1999) found 23 species of Lepidoptera not permissive to the rusty tussock moth, Orygia antigua N P V , though no other Lymantriidae were included in their study. Detailed host range testing of another lymantriid, the browntail moth, Euprotis chrysorrhoea N P V , revealed 73 species of Lepidoptera from 14 different families not permissive for E. chrysorrhoea N P V (Cory et al., 2000). This list includes the closely related Euproctis similis. The narrow host ranges of lymantriid N P V s contrasts the more variable and broad host range of noctuid NPVs. The NPVs isolated from noctuid moths can infect dozens of species including those from more than one lepidopteran family (Doyle et al., 1990). Autographa californica N P V , the type species for the baculoviridae, is found to be infective for 43 species, covering 11 lepidopteran families, though progeny virus was not confirmed in all cases (Payne et al. 1996). The host range for Anagrapha falcifera N P V is similar to that of A. californica N P V . A. falicera N P V is shown to be infective for 30 out of 38 species tested, covering 10 families of lepioptera (Hostetter and Puttier, 1991). Again, progeny virus was not examined. Most other host range studies have examined more restricted groups of hosts but still find noctuid NPVs to infect more than one species (Carner et a l , 1979; Harper et al., 1976; Allaway and Payne, 1984). This trend does not extend across the Noctuidae. Viruses from other species, such as Spodoptera exigua (Hiibner), also appear to be species-specific (Gelernter and Federici, 1986). Currently, there are few published host studies in which the same methodology has been applied to a number of different species and where the identity of progeny virus has been determined. This makes it difficult to ascertain any trends useful for predicting impacts on non-36 target organisms (Cory et al. 2000). In addition, several studies have shown that baculoviruses produced during cross infections are not always the same as the inoculum used, either as a result of contamination or the presence of a vertically transmitted virus in the test species (Jurcovicova 1979; McKinley et al., 1981; Cory et a l , 2000). The presence of latent and sublethal infections in natural populations is an area which we know little about, thus it is important to identify progeny virus in all cross infection studies, particularly those studies involving field infected individuals. This study compares the NPVs from two closely related host species, the western and forest tent caterpillar, M. californicum pluviale (Dyar) and M. disstria (Hiibner). Both species occur in north western North America where they demonstrate cyclic population dynamics and periodically defoliate deciduous tress. Tent caterpillar populations cycle through peak densities every 6 t o l l years (Myers 1988). Population cycles are characterized by declining and low host densities for three to four years, followed by a period of increase for two to three years, and high densities for two-three years (Myers 1988). Baculovirus infections, particularly N P V , are often associated with high density populations of both tent caterpillar species (Wellington, 1962; Myers, 1988; Kukan and Myers, 1997; Myers, 2000). Tent caterpillar populations in British Columbia are of particular interest because they are closely related host species and are known to have historically overlapping home ranges (Kukan, 1996). Baculoviruses from different host species within the same geographic region have been shown to be more similar than baculoviruses from the same host species in geographically separate populations (Cory et al., 1997). Host species closely related to target hosts may be at risk to the release of baculoviruses, particularly if there is potential for home range overlap. In the following study we used restriction endonuclease analysis (REN) to characterize the genomes of M. c. pluviale N P V (McplNPV) and M. disstria N P V (MadiNPV) and used bioassays to determine the cross-infective potential of each virus. Progeny virus was examined using R E N analysis to confirm that true cross-infection had taken place. Analysis of progeny virus from M. disstria larvae cross infected with M c p l N P V indicated that only 12% of larval mortalities were the result of the M c p l N P V inoculum. The remaining moralities resulted from a virus identical to MadiNPV. P C R was used to rule out MadiNPV contamination of the M c p l N P V stock. External contamination and laboratory contamination were also unlikely. We suggest that virus was present within larvae as either a latent or sublethal infection. 37 3.2 METHODS Insects The larvae used in each bioassay were hatched from egg masses collected from the field during the spring of 1999. M. c. pluviale egg masses were collected from Pender and Galiano Islands in the southern Gulf Islands of British Columbia. Caterpillar populations on these islands reached peak densities in 1996 and declined through 2000. M. disstria has not been reported to occur in this area but has been seen occasionally at very low densities in Victoria and on the south western mainland (Myers, pers. comm). M. disstria egg masses were collected from various sites around Prince George, British Columbia. Populations in this area had been at high densities for approximately three years (Myers, pers. comm). M. c. pluviale is rarely observed in the Prince George region. Virus M. c. pluviale N P V (McplNPV) and M. disstria N P V (MadiNPV) were collected from insects naturally infected in the field in 1998. Virus stocks used for each bioassay were a mixture of OBs collected from five different infected individuals. OBs were separated from the individuals by first suspending infected cadavers in 500 ul of dH 2 0 then grinding them with plastic pestles. A series of low speed centrifuge spins (1000 rpm for 35 sees) and washes with dH 2 0 were used to separate OBs from insect debris. OBs were then pelleted with a high speed centrifuge spin (14,000 rpm for 20 minutes). The pellet was washed two times with dH 2 0 and re-suspended in 1000 ixl of dH 2 0. Equal amounts of O B solution from each individual were combined to make the representative virus stock. Stock concentrations were estimated using a Neubauer Haemocyctometer and light microscope (450X). Serial dilutions were made using dH 2 0. Bioassays Each egg mass was surface sterilized with 10% bleach (60 seconds) and rinsed with dH 2 0 (60 seconds) to remove any virus or bacteria that may have been present as surface contaminants. Egg masses were left to hatch at room temperature. Once larvae began to emerge they were placed as family groups in I L paper cups where they were fed fresh alder leaves and monitored daily. Alder leaves were washed with 10% bleach and rinsed with dH 2 0 prior to feeding. 38 Individuals were removed immediately following the respective molts to ensure that all larvae in any given assay were approximately the same age. Larvae were starved for 24 hours then individually fed N P V contaminated alder leaf discs (50 mm 2). Doses ranged from 120-5800 OBs/larvae for M. c. pluviale and 130-26000 OBs/larvae for M. disstria (Table 1). Control larvae were fed leaf discs treated with dH^O. Individuals were given 24 hours to consume a minimum of 95% of the leaf disc. Those not consuming 95% were removed from the study. Larvae from the same dose were reared in groups of 10-15 in 0.5L paper cups. Each bioassay was monitored for 14 days. The numbers of larvae alive as well as the number of larvae dead from virus or other causes were recorded daily. N P V infection was diagnosed by physical characteristics such as a fragile integument that could easily be broken to release liquid. The presence of occlusion bodies was verified using a light microscope. A l l dead larvae were stored at - 2 0 ° C in 1.7 ml Eppendorf tubes. A l l bioassays using M. disstria larvae were conducted using larvae at the 3 r d, 4 t h and 5 t h instar stages. Only 3 r d instar M. c. pluviale larvae were used because low host population densities resulted in a limited number of egg masses. Virus purification and restriction endonuclease analysis Viral D N A was extracted from both M c p l N P V and M a d i N P V by pelleting OBs extracted from larvae as described above and re-suspending the pellet in 1000 ju.1 of dP^O. This solution was heated at 6 5 ° C for 30 minutes to denature any insect DNases present. Virions were released from OBs by treatment with an alkali lysis solution ( I M N a 2 C o 3 , 150 m M NaCl , 0.1 m M E D T A , p H 10.8) at 3 7 ° C for 60 minutes followed by a high speed centrifuge spin (14,000 rpm for 30 minutes). D N A was released from the virions by re-suspending the pellet in 500 uJ of proteinase K buffer ( lOmM Tris (pH 7.4), l O m M E D T A , 150mM NaCl , 0.4% SDS) and 20 jil of proteinase K enzyme (20 mg/ml) overnight at 3 7 ° C . Proteins were removed using a protein precipitation solution (Puregene™). One hundred percent ethanol and sodium acetate (3M, p H 8.0) were used to precipitate D N A ( - 2 0 ° C for three hours). D N A was pelleted with a high speed spin (14,000 rpm for 15 minutes), followed by a wash with 70% ethanol for 1.5 hours. The D N A was then spun at 14,000 rpm for 10 minutes and re-suspended in 65(il of T E buffer at room temperature. Al l D N A was stored at 4 ° C . 39 Genetic comparisons between the M c p l N P V and M d N P V genomes were made using EcoRl, HindUI and Xhol (GIBCO B R L ) . Viral D N A (~ 2ug) was digested for 12 hours at 3 7 ° C . Digests were run on 0.7% agarose gels at 65V for 24 hours. Gel and running buffer contained ethidium bromide at a final concentration of 0.5%. D N A fragment size was estimated by comparison with X-HindlII molecular marker (NEB Tech). Genome sizes are the mean of up to 15 isolates from two to six different gels. Sizes were estimated using all three enzymes. HindUI R E N profiles were collected from the progeny virus produced from each bioassay to confirm that mortalities were the result of the virus inoculum used. PCR amplification of MdNPV specific sequences P C R primers (designed by D . Theilmann) were synthesized as 20-mers consisting of D N A complimentary to the 5' (5' - C G C C G A C A A T G A T A C A T T T A-3' ; nt 5'156-175-3 and 3' ( 5 ' - T G T T G G C G A T T C T C T T G A T G - 3 ' ; nt 5'-579-598-3') regions of the 0.9 kb Pol region of MadiNPV (Neilson et al., in prep). Amplification reactions used total viral genomic D N A isolated from infected larvae and both the M c p l N P V and M a d i N P V stocks. Viral D N A was heated for 15 minutes at 6 5 ° C prior to its addition to each 25 ul reaction. Approximately 50 ng of D N A was added to a P C R mixture that contained a final concentration of 4 m M MgCl2, 2.5 ul of 10X P C R buffer, 0.2 m M of each dNTP, 0.5 u M of each primer, together with 1.25 U of Taq polymerase (Gibco B R L ) . Each reaction consisted of one cycle at 9 5 ° C (1.5 min); three cycles at 9 5 ° C 1.5 min, 4 9 ° C (45 sec), and 7 4 ° C (1.0 min); 32 cycles at 9 5 ° C (1.5 min), 5 2 . 5 ° C (45 sec), 7 4 ° C (1.0 min); and a final extension at 7 4 ° C for 5 min. These primers consistently detected 0.001 ng of MadiNPV D N A . P C R inhibitors were often a problem with M c p l N P V D N A samples, therefore all reactions used primers designed to target a 1.2 Kb region of the D N A polymerase gene (Nielsen et al., in prep) as a control for false negatives. Positive controls consisted of MadiNPV collected from M. disstria larvae. Negative controls used all reagents except template D N A . Statistical Analysis The equations of the dose-mortality curves were computed using Proc Logistic in SAS (1990). Slopes and intercepts were compared as described by Collet (1991). This involved fitting three models; the first fitted model constrains all lines to have the same slopes and intercepts, the second fits a common slope but allows the intercepts to vary, the third model 40 allows both the slope and intercept to vary. The difference between the deviance of the third and second model has a chi-square distribution under the null hypothesis that the slopes are identical. The degrees of freedom are given by the number of parameters estimated in the third model minus the second one. A l l tests of significance were done at a=0.05. LDso's and associated confidence intervals were determined using the Trimmed Spearman-Karber method (Hamilton et al. 1977) and U S E P A software (Ecological Monitoring Research Division 1992b). 3.3 RESULTS Genetic comparison of McplNPV and MadiNPV R E N profiles of M c p l N P V genomic D N A were different from and M a d i N P V genomic D N A with all three enzymes used (Figure 3.1). No common bands were shared between the two virus types. Restriction fragments were labeled alphabetically and sized by comparison with X-HindUI molecular marker. Approximate sizes for each restriction fragment are listed in Table 2. Virus isolates from both MadiNPV and M c p l N P V were genetically variable with each restriction enzyme used (Appendix 2); therefore, fragment and genome sizes aregiven from the mean size from up to 15 viral isolates. Total genome sizes were taken to be the largest genome size from the R E N profiles. M c p l N P V was found to be 122.5 +/- 0.13Kb and MadiNPV was found to be 129.4 +/- 0.4 kb. It should be noted that these approximations are under-estimates because fragments smaller than 1 kb were difficult to detect. Bioassays M. disstria larvae: MadiNPV challenge M. disstria larvae were apparently susceptible to MadiNPV at 3 r d, 4 t h and 5 t h instar stages. Mortality increased with increasing virus dose when 3 r d instars were challenged with MadiNPV; however, no significant relationship was found between dose and mortality for 4 t h instar larvae (Figure 3.2). A l l mortalities for 4 t h instars challenged with M a d i N P V were above 50%. An increased mortality was associated with increased dose in the 5 t h instar bioassay, though this relationship was not significant (Figure 3.3). High overall mortality was also associated with 41 MadiNPV at the 5 t h instar stage. Mortality did not decrease with older instars when larvae were challenged with MadiNPV. No mortalities were observed in controls. Cross infection: M . disstria larvae: McplNPV challenge M. disstria larvae were apparently susceptible to M c p l N P V at 3 r d, 4 t h, and 5 t h instar stages but there was no obvious relationship between dose and mortality with 3 r d and 4 t h instar larvae (Figure 3.2). Mortality did increase with increasing virus dose when 5 t h instar larvae were challenged, however, this relationship was not significant (Figure 3.3). L D 5 0 S could not be accurately estimated for the 4 t h and 5 t h instars challenged with MadiNPV or for any cross infections because mortalities did not adequately cover the range of the dose response curve (Figures 3.2 and 3.3). The LD50 for 3 r d instar M. disstria larvae challenged with MadiNPV was 1318 (512-3890) OBs/larvae. No mortalities were observed in controls. M. californicum pluviale larvae: McplNPV challenge M. c. pluviale larvae were susceptible to McplNPV. Mortality increased with increasing virus dose (Figure 3.4). The L D 5 0 for larvae challenged with M c p l N P V was 2818 (1949-4168) OBs/larvae. No mortalities were observed in controls. Cross infection M . c. pluviale larvae: MadiNPV challenge M. c. pluviale larvae at the 3 r d instar were not obviously susceptible to cross-infection with MadiNPV (Figure 3.4). Mortalities that did occur in the cross infection bioassay did not produce distinctive signs of viral infection. These larvae were lethargic and stopped feeding prior to death. A small number of larvae died from unknown causes in all bioassays and this number was higher, but not significantly higher in cross infected individuals. Bioassays: examination of progeny virus R E N analysis was used to verify that all mortalities were caused by the virus inoculum used to initiate infection. Restriction profiles collected from 50 cross-infected M. disstria individuals demonstrated that only six individuals died from M c p l N P V infection used to initiate infection. The remaining 44 were infected with a virus identical to MadiNPV (Figure 3.5). Of the six mortalities due to McplNPV, two R E N variants were apparent (Figure 3.5). Deaths 42 occurred only with 4 t h and 5 t h instar larvae and were not dose dependent (Table 3.3). Al l 44 MadiNPV R E N digests were identical. A sample of 23 M. disstria larvae challenged with M a d i N P V were also examined using R E N analysis. A l l individuals examined died from a virus identical to the M a d i N P V found in the cross infection bioassays. This virus, while clearly MadiNPV, differed from the MadiNPV inoculum used to challenge the larvae (Figure 3.5). M. c. pluviale mortalities following inoculation with M c l p N P V were also examined using R E N analysis. M c p l N P V was confirmed in the 30 larvae examined. R E N profiles of M c p l N P V produced during the cross infections were genetically variable but identical to the M c p l N P V stock inoculum used (Figure 3.5). P C R examination of M c p l N P V stock To confirm that the initial M c p l N P V stock was not contaminated with MadiNPV, the virus stock was tested using PCR. No MadiNPV was detected in the M c p l N P V stock (Figure 3.6). 3.4 DISCUSSION The original goals of this study were to use R E N analysis to characterize the NPVs from two closely related tent caterpillar hosts, M. c. pluviale and M. disstria, and to assess the cross-infective potential of both viruses. We show that both viruses are genetically distinct. However, properly addressing the cross-infections was complicated by the activation of what appeared to be a latent or sublethal infection in M. disstria larvae. The infection was activated when M. disstria larvae were cross infected with McplNPV. Only 12% of M. disstria larvae succumbed to M c p l N P V infection. The remaining 88% of larvae succumbed to infection by MadiNPV, a virus not provided in the initial viral challenge. In the reciprocal cross infection, our results suggest that MadiNPV does not produce an overt infection in 3 r d instar M. c. pluviale larvae. However, there were several individuals in all bioassays that died from unknown causes, and this number was slightly, but not significantly, higher in cross-infected individuals. These larvae were often lethargic and stopped feeding prior to death. The cause for these unknown mortalities was not determined, but this may indicate that M. c. pluviale are partially susceptible to early stages of infection. Mortality may result from the insect's efforts to clear the virus from its system by the sloughing of midgut cells (Hoover et al., 2000) or some form of damage to the midgut. 43 Three possible explanations for the replication of a virus identical to MadiNPV in insects fed M c p l N P V are (1) we had low levels of MadiNPV contamination in the M c p l N P V inoculum, (2) we had low levels of contamination within the laboratory, or (3) we activated a latent or sublethal infection in M. disstria larvae. We have several lines of evidence to suggest the latter. First, feeding M. disstria larvae with both MadiNPV and M c p l N P V produced the same MadiNPV R E N profile, which differed from the MadiNPV genotype of the inoculum. The production of a single MadiNPV profile in individuals cross-infected with M c p l N P V is consistent with low levels of MadiNPV contamination; however, the MadiNPV isolated from the cross infected individuals was different from the MadiNPV innoculum used. The MadiNPV stock used to infect M. disstria larvae consisted of a collection of OBs from five different field-infected individuals. R E N analysis of virus isolated from field infected M. disstria larvae has shown MadiNPV to be genetically heterogeneous both within and between sampling sites (Appendix 2). Previous experience with mixed genotype bioassays has demonstrated that different genotypes are often expressed in different individuals (Cooper and Myers, in prep). Therefore we would not expect that all mortalities, from either a low level of contamination in the M c p l N P V stock or from the MadiNPV stock to be the result of the same genetic variant of MadiNPV. Secondly, P C R examination of the M c p l N P V stock failed to find evidence that MadiNPV D N A was present. External contamination is unlikely because egg mass surfaces were decontaminated before larvae began to hatch. Incomplete decontamination, laboratory contamination, and contamination of foliage are also unlikely because no mortalities were seen in controls. Lastly, M. disstria populations had been in outbreak densities for approximately three years and viral infection has been observed in field collected individuals (Myers, pers comm.). Similar studies have reported the detection of unexpected or covert infections in larvae from egg masses collected in the field (McKinley et al. 1981). Larvae hatching from egg masses can become infected by either surface contamination or transovarial transmission (passage of virus from mother to egg within the ovary). Surface contamination involves transfer of virus from females to egg mass surfaces during oviposition and larvae become infected by consuming virus as they emerge from the egg mass (Kukan, 1999). Transmission studies of both laboratory and field insects have demonstrated that adults surviving infection can pass low levels of infection to progeny (Neelgund and Mathad, 1978; Abul-Nasr et al., 1979; Young and Yearian, 1982; Smith and Vlak, 1988, Fuxa and Richter, 1991). Gypsy moth adults (Lymantria dispar) 44 externally contaminated with virus have been shown to contaminate egg masses during oviposition (Murray and Elkinton, 1990). Similarly, contaminated adults have been shown to transmit virus to uncontaminated mates. Corn earworm males (Heliothis zea) fed virus as adults were found to transmit virus to untreated females during mating, and these females were sufficiently contaminated to transmit virus to some of their progeny (Ham and Young, 1974). In a few of these cases, sterilization of egg mass surfaces could eliminate transmission (Murray and Elkinton, 1990) though this was often not the case. Kukan (1999) reviewed four studies that examined progeny from field collected egg masses, and in each case, surface sterilization was found to reduce the incidence of virus but not eliminate it, suggesting that virus may also be transmitted within the eggs themselves. This form of transmission helps to explain low levels of infection in field collected egg masses and provides an explanation for the inability to consistently rear healthy lab stocks of larvae such as the cabbage looper (Trichoplusia ni) (McEwen and Hervery, 1960; Guy et al. 1985; Fuxa et al. 1999). N P V infections have been observed in T. ni colonies not treated with virus, with higher prevalence in insects under stress than in control insects. More recently, Fuxa et al. (1999) demonstrated transovarial transmission of both N P V and C P V (cytoplasmic polyhedrosis virus) in laboratory colonies of T. ni larvae after decontamination of egg mass surfaces. The presence of viral infections in field studies and lab colonies suggests that this form of transmission may be common in some baculovirus systems. It has been suggested that sporadic baculovirus outbreaks in natural lepidopteran populations result from the activation of latent or sublethal infections (Longworth and Cunningham, 1968, Evans and Harrap, 1982). Latency is used to describe the state of a virus that does not produce obvious signs of infection but can be transmitted between host generations (Hale and Margam, 1988). This requires the pathogen to remain non-replicating and non-infective (Fuxa et al., 1992) until an appropriate stressor or stimulus can activate the infective form. Several mechanisms for latent viruses have been suggested. For example, Herpes simplex virus is maintained within the host nucleus as independent viral genetic material (Mellerick and Fraser, 1987), while Hepatitis B virus can be integrated directly into the host genome (Howard, 1986). In contrast, the measles virus is maintained as a low-level persistent infection (Catteneo et al., 1988). Information concerning the possible forms of latent baculovirus infection is limited because the location and form of the virus was not determined in most studies. A few studies 45 describe the transmission of empty polyhedra and non-occluded virions (not embedded within polyhedra) in cell lines (Smith and Arnott, 1969; Godwin and Adams, 1980; Vaughn et al., 1991) and adults surviving infection (Fuxa et al., 1992; Fuxa et al., 1999). Fuxa et al. (1992) found these particles to be non-infectious to first instar larvae and concluded that formation of non-infectious virus may be a viral adaptation that allows adults to disperse, oviposit, and transmit virus to new areas. Fuxa et al. (1999) suggest that non-occluded virions may be the mechanism by which latent infections are maintained within the host. Since latent or low level persistent infections are believed to be present in small amounts, detection of such infections requires more sensitive molecular techniques. Hughes et al. (1997) detected a low-level persistent infection in a laboratory culture of the cabbage moth, Mamestra brassicae. M. brassicae N P V ( M b M N P V ) was activated when insects were fed the closely related Panolis flammea N P V (PfNPV) and the more distantly related Autographa californica N P V (AcMNPV) . Transcripts for early (ie-1), late (p6.9) and very late (polh) gene expression were found in fat-body cells of larvae. Detection of the appropriate transcripts allowed the authors to suggest that the latent infection was maintained as a persistent infection, allowing the continuous expression of viral proteins at low levels. This is analogous to the situation proposed to explain persistent measles infections (Catteneo et al., 1988). Hughes et al. (1997) suggest that virus maintained as D N A alone would not be able to initiate an infection via the insect mid gut. A similar low level persistent infection was detected in the Indian meal moth, Plodia interpunctella (Burden et al., in press). Here, R N A transcripts for the granulin protein of the Indian meal moth granulo virus, Plodia interpunctella(GY) were present in a high proportion of P. interpunctella larvae surviving virus challenge. Granulin is a late expressed gene that is only transcribed after viral genome replication (Griffiths et al., in press). In addition, R N A transcripts were detected in 60 to 80% of second generation larvae derived from the surviving adults demonstrating that covert infections can be vertically transmitted to offspring. The presence of latent or sublethal infections in field populations may explain viral persistence at low host densities. The traditional route for baculovirus infection is thought to be mediated by OBs surviving in the environment, with only limited evidence for vertical transmission in some systems. Given that many insect systems do not reach outbreak densities, the vertical transmission of sublethal infections may be crucial for long-term virus persistence (Kukan, 1999; Griffiths et al., in press). Latency and sublethal infection not only provide a mechanism for the transfer of virus from one generation to the next, they also provide a 46 mechanism for the movement of virus to new areas via host dispersal. The infectious form of the virus must be subsequently activated to account for epizootics. Stressors such as crowding and poor diet are known to activate what are believed to be latent baculovirus infections (Smith, 1963; Longworth and Cunningham, 1968, Fuxa et al., 1992; Fuxa et al., 1999). Experiments designed to detect and identify low levels of infection will provide clues about the factors that mediate the shift from a virus that kills its host to one that can cause a covert infection and vice versa. These data have broad implications for understanding the role baculoviruses may play in insect population dynamics and further research is needed to ascertain the prevalence of sublethal infections in other field populations. Finally, the presence of MadiNPV in the larvae used in this study makes interpretation of this bioassay data difficult. Results from various studies have shown that recombination between virus types can increase virus host range (Maeda et al., 1993; Kondo and Maeda, 1991; Bideshi and Frederici, 2000). Baculovirus helicases are considered to be a key host range determinant. Maeda et al. (1993) demonstrated that recombination between A c M N P V and B m N P V resulted in A c M N P V acquiring a fragment of D N A encoding part of the putative D N A helicase gene from B m N P V . This allowed A c M N P V to infect a previously non-permissive B. mori cell line. Although HindIII digests of M c p l N P V progeny virus resulting from 12% of cross infections of M. disstria larvae with M c p l N P V in this study did not differ from the stock McplNPV, we cannot rule out the possibility of recombination or other synergistic viral effects between innoculum and latent virus. 47 Table 3.1. Virus type and dosage, in OBs, used to challenge M. disstria and M. c. pluviale larvae in the 1999 bioassays. Numbers in parenthesis represent the number insects challenged at each dose. Virus • MadiNPV Host Instar Dose (#insects challenged) M. disstria 3rd 130(13), 660(17), 1300(20), 2300(21), 6600(22) 4 t h 130(8), 660(12), 1300(10), 8800(12), 13000(12) 5 t h 1300(3 groups;7,8,9), 8800(3 groups;10,10,7), 13000(3 groups; 10,10,9), 26000(3 groups; 10,10,9) M. c. pluviale 3rd 130(20), 1300(20), 2200(21), 6600(20) Virus M c p l N P V Host Instar Dose M . disstria 3 r d 120(20), 700(13), 1200(16), 2300(12), 5800(20) 4th 700(22), 1200(11), 7000(20), 12000(19) 5 t h 1200(3 groups; 10,10,9), 7000(3 groups;ll,10,9), 12000(3 groups;9,9,7) 23000(3 groups;9,9,10) M. c. pluviale 3rd 120(15), 700(21), 2300(16), 5800(24) 48 Table 3.2. Estimated REN fragment sizes (kb) of M. disstria NPV and M. californicum pluviale NPV genomes. Letters represent fragments produced by digestion with each enzyme. Total genome and fragment sizes are mean estimates 15 restriction profiles (virus isolates) from six gels. Enzyme Xhol EcoRI HindUI MadiNPV McplNPV MadiNPV McplNPV MadiNPV McplNPV Letter Fragment size A 17.9 18.8 19.9 18.5 22.8 18.9 B 15.9 18.1 18.5 18.0 21.7 14.5 C 13.2 16.7 14.1 14.4 20.6 13.2 D 11.5 14.4 13.2 12.9 13.2 12.1 E 10.1 10.9 10.5 11.0 10.9 10.7 F 9.5 9.0 7.7 9.6 5.4 7.7 G 9.2 8.5 6.4 8.2 4.0 6.1 H 8.4 8.2 5.9 7.8 2.1 5.7 I 7.7 6.0 5.4 7.4 2.0 4.6 J 6.7 3.7 5.2 6.0 1.7 4.3 K 6.2 1.7 4.0 5.5 4.1 L 5.6 0.94 3.9 5.2 3.2 M 5.2 0.92 3.9 4.3 3.2 N 4.3 3.7 3.2 3.0 O 2.4 3.5 3.0 2.8 P 2.4 2.5 1.6 2.4 Q 0.9 1-4 1.5 R 1.3 S 1.1 T 1.0 U 0.9 Total 129.4+/-0.4 118.0a 127.5" 122.5+/0.13 106.1+/0.11 114.9+/0.12 hragment sizes b a s e ( j o n REN profiles from less than eight viral isolates 49 Table 3.3. Dose and instar of the six M. disstria larvae succumbing to McplNPV infection. Host Virus type Instar Dose (# OBs) M. disstria M c p l N P V 4 m 700 M. disstria M c p l N P V 5 t h 1200 M. disstria M c p l N P V 5 t h 12000 M. disstria M c p l N P V 5 t h 23000 M. disstria M c p l N P V 5 t h 23000 M. disstria M c p l N P V 5 t h 23000 Figure 3.1. Xhol (A), EcoRl (B), Hindlll (C) REN digests of MadiNPV (1) and McplNPV(2). Fragments are labeled alphabetically and correspond to the sizes in Table 2. * represent submolar bands. 51 3rd instar "55 O E o 2 1 ^ 0 -1 ^ -2 -3 2.5 © . © 3 3.5 4 log dose O McplNPV o MadiNPV Linear 4.5 (McplNPV)| Linear (MadiNFV)l i 5 4th instar >. o E •*-« "5> o 3 2 1 -I 0 -1 -I -2 -3 ® © O McplNPV o MadiNPV Linear (McplNFV) Linear (MadiNPV) 2.5 3 3.5 4 log dose 4.5 Figure 3.2. Mortality experienced by 3 r d and 4 t h instar (a and b respectively) M. disstria larvae challenged with MadiNPV (open circles) and McplNPV (open diamonds). Circles and diamonds represent actual mortality while solid and dashed lines represent fitted logit mortality. Log dose and actual mortality/samples size follow the equation of the dose-mortality curve. 3 r d instar MadiNPV challenge: logit (mortality)=-6.0923+1.887(log10dose), 2.11(1/12), 2.82(5/12), 3.11(6/14), 3.36(8/13), 3.82(9/13). McplNPV challenge: logit (mortality)=-0.4568+0.407(log10dose),2.08(7/13), 2.84(2/11), 3.08(4/12), 3.36(0/12), 3.76(4/16). 4 t h instar MadiNPV challenge: logit (mortality)=-4.4598+0.1481 (logi.dose), 2.85(7/15), 2.08(6/15), 3.85(8/12), 4.08(11/18). McplNPV challenge logit (mortality)=-2.6815+0.4925(log,„dose),2.84(6/15), 3.11(6/15), 3.94(8/12), 4.11(11/18). 52 5th instar 3 -, £ 1 10 I -1 -I -2 -3 O § O O O O O O o McplNPV O MadiNPV Linear (McplNPV )| Linear (MadiNPV)l 2.5 3.5 4 log dose 4.5 Figure 3.3. Mortality experienced by 5 t h instar MadiNPV challenge with MadiNPV(circles) and McplNPV (open diamonds). Circles and diamonds represent actual mortality while solid and dashed lines represent fitted logit mortality. Log dose and mortality/sample size follow the equation of the dose-mortality curve. MadiNPV challenge logit (mortality)=-3.4383+1.1332(log10dose), 3.11(2/8, 6/9,7/9), 3.94(7/10,6/10, 6/8), 4.11(8/10, 7/9, 6/10), 4.41(8/9, 9/10,10/10). McplNPV challenge: Iogit(mortality)=-8.8794+1.4774(log10dose), 3.08(0/10, 0/9, 0/10), 3.85(2/9, 0/10, 0/9), 4.08(2/7,1/8,3/4), 4.36(0/9, 0/9, 3/7). 53 3rd instar 3 - i 2 - o 1 O McplNPV rtal 0 o MadiNPV o -1 O o o Linear logit i -2 -3 -4 (McplNPV) -5 o o © © I I I 2 2.5 3 3.5 4 4.5 log dose Figure 3.4. Mortality experienced by 3 r d instar M. californicum pluviale larvae challenged with McplNPV (open diamonds) and MadiNPV coloured circles). Log dose and mortality/sample size follow the equation of the dose-mortality curve. MadiNPV: logit(mortality)=-4.968+1.508(log10dose), 2.08(5/21), 2.84(4/24), 3.36(6/26), 3.76(22/24)) and MadiNPV (open circles). No viral mortalities were observed from the cross infection of M. c. pluviale NPV with MadiNPV. 54 Cross infection Progeny Virus StockVirus Figure 3.5. HindIII REN digests of progeny virus from 11 M. disstria larvae succumbing to viral infection during McplNPV cross infections. McplNPV was the virus used as the inoculum. MadiNPV was found in 44/50 cross infected individuals. L is the X-Hindlll marker. 55 Figure 3.6. Examination of McplNPV stock for MadiNPV contamination using PCR. Primers were designed to distinguish between MadiNPV and McplNPV by targeting a 443 bp region of the MadiNPV DNA pol gene. MadiNPV isolated from field collected M. disstria larvae served as positive controls (lanes 1-4). No product is produced from McplNPV DNA (lanes 8 and 9). No MadiNPV product was detected in McplNPV stock (lanes 6 and 7). Primers could detected 0.001 ng DNA. All PCRs used dH20 as negative control (lane 12). Primers designed to target a 1.2 Kb region of McplNPV genome were used as a control for false negatives (lane 11) (Nielsen et al., in prep). 56 4.0 GENERAL CONCLUSIONS Viral diseases are an important feature of fluctuating lepidopteran populations. In this thesis I examine two aspects of baculovirus ecology. First, I examine the genetic variability present in natural N P V populations. Second, I determine the host range of two NPVs infecting two closely related host species, Malacosoma californicum pluviale (Dyar) and Malacosoma disstria (Hiibner). M y results may apply more generally to other lepidopteran-NPV systems that exhibit cyclic population dynamics. 4.1 SUMMARY OF FINDINGS In chapter 2,1 demonstrate that natural M c p l N P V populations in southwestern British Columbia are genetically heterogeneous. Fourteen genetic variants were found among six host sites, and the distribution of this genetic variation was found to be structured at the population level. Certain genetic variants were detected in more than one host site and on more than one island, suggesting that virus can be transported between sites, either by the dispersal of adults or the movement of virus by wind or birds. Subtle changes in pathogenicity were found to be associated with different M c p l N P V populations indicating that phenotypic changes may be associated with genetic variation. Genetic population structure, combined with the subtle changes in pathogenicity associated with the different virus populations, implies that local genetic variability in N P V populations can influence local host dynamics. Chapter 3 demonstrates that the NPVs infecting two closely related host species, M. disstria and M. c. pluviale, are genetically distinct and that the N P V infecting M. disstria can be vertically transmitted. Cross infection experiments indicated that 3 r d instar M. c. pluviale larvae were not susceptible to MadiNPV. In the reciprocal cross infection, only small proportions of M. disstria larvae were susceptible to McplNPV. The majority of mortalities resulted from an unsuspected MadiNPV infection. The cross infection of M. disstria larvae with M c p l N P V appears to have activated the latent or sublethal MadiNPV infection. Contamination of the M c p l N P V stock used to challenge M. disstria larvae, contamination of egg mass surfaces, and laboratory contamination were all unlikely suggesting the MadiNPV infection was present within the larvae themselves. Detection of a latent or sublethal infection suggests that virus can be vertically transmitted in field populations and provides a mechanism by which virus can be 57 maintained within host populations during low host densities. It also provides a route for the transmission of virus to new sites via adult dispersal. Overall, my results suggest that genetic variability present in virus populations may be important for virus survival. Phenotypic changes have been shown to be associated with genetic variation in other baculovirus systems (Hodgson et al., in press), however, the dynamics of mixed genotype (genetic variant) infections are poorly understood. The existence of phenotypic variation between sympatric genetic variants has important implications for the maintenance of genetic diversity in baculovirus populations. Although selection pressures acting on the virus and host population will promote dominance of the most fit genetic variant, the coexistence of genetic variants may be promoted by differential selection of genetic variants in different environments. These processes are not mutually exclusive and deserve further study. The examination of other life history traits important for viral fitness, such as speed of kill and O B yield, may provide important information about the dynamics of mixed genotype infections and how such infections alter virus-host interactions. The least understood stage in the dynamics of viral infection is the persistence of virus between host generations. The persistence of virus through changing host densities has important implications for the ability of virus to regulate host populations. The detection of latent or sublethal infections in field populations provides a key mechanism by which virus can be maintained during low host densities. Covert infections, vertically transmitted from parents to offspring, may be maintained within larvae providing a viral reservoir that can later be activated by the stresses of high host densities. Vertical transmission also has important implications for the long term effectiveness of pathogens in biological control programs. In addition, latent and sublethal infections provide a mechanism by which virus can be moved between sites with dispersing adult moths. 4.2 FUTURE RESEARCH The results of my research have generated many additional questions for future studies: (1) How do the traits important for virus survival, such as speed of kill, O B production and pathogenicity, vary between individual genetic variants? What are the trade-offs associated with each life history trait? (2) What regions of the genome are important for viral life history traits such as the LD50? 58 (3) What are the effects of mixed genotype bioassays? How do interactions between two and more genetic variants affect virus-host interactions? (4) Do viral phenotypes change over the course of the tent caterpillar cycle? (5) What is the biological relevance of recombination in natural systems? How does recombination affect the production of new variants? What role does recombination play in host range expansion in field populations? (6) How prevalent are latent or sublethal infections? Are they associated with other lepidopteran field populations at similar host densities? (7) How prevalent is vertical transmission in field populations? How prevalent is vertical transmission in migrating and dispersing species? (8) What are the factors that trigger latent and sublethal infections? 59 REFERENCES Abul-Nasr, S.E. , Ammar, E . D . , and Abul-Ela, S . M . 1979. Effects of nuclear polyhedrosis virus on various developmental stages of the cotton leafworm Spodoptera littoralis (Boisd.). J.Appl. Entomol. 88:181-187. Allaway, G.P. , and Payne, C . C . , 1983. A biochemical and biological comparison of three European isolates of nuclear polyhedrosis virus. Arch. Virol. 15 A3 Anderson, R . M . and May R . M 1980. Infectious diseases and population cycles of forest insects. Science. 210: 658-661 Andrews, R . E . , Spence, K . D . and Miller, L . K . 1980. Virulence of cloned variants of Autographa californica nuclear polyhedrosis virus. Appl. Environ. Micorbiol. 39:932-933. Barber,K.N., Kaupp, W.J. , and Holmes, S.B. 1993. Specificity testing f the nuclear polyhedrosis of the gypsy moth, Lymantria dispar (L.) (Lepidoptera: Lymantriidae). Can Entomol. 125:1055 Bell, R .A . Owens, C D . , Shapiro, and M . , Tardif, J.R. 1981. Mass rearing and virus production: Development of mass-rearing technology. In 'The gypsy moth; research toward integrated pest management" (C.C. Doane, and M . L . McManus, Eds) Forest Service U S D A , Tech. Bull. 1584. Berryman, A . A . 1996. What causes population cycles of forest Lepidoptera? Trends Ecol.Evol. 11:28-32. Bideshi, D . K . and Federici, B . A . 2000. The Trichoplusia ni granulovirus helicase is unable to support replication of Autographa californica multicapsid nucleopolyhedrovirus in cells and larvae of T.ni. J. Gen Virol. 81:1593-1599. Burden, J.P., Griffiths, C . M . , Cory, J.S., Smith, P., Smith, S., and Steven, M . 2001. Vertical transmission of sublethal granulovirus infection in the Indian meal moth, Plodia interpunctella. In press. Burden, J.P., Hails, R.S., Windass, J.D., Suner, M . - M . , and Cory J.S. 2000. Infectivity, speed of kill and productivity of a baculovirus expressing the itch mite toxin Txp-1 in second and fourth instar larvae of Trichoplusia ni. J. Invert. Pathol. 75:226-236. Carner, G.R., Hudson, J.S., and Barnett, O.W. 1979. The infectivity of a nuclear polyhedrosis virus of the velvet bean caterpillar for eight noctuid hosts. J. Invertebr. Pathol. 33:211-216. Cabellero, P., Aldebis, H . K . , Vargas-Osuna, E . , and Santiago-Alvarez, C . 1992. Epizootics caused by a nuclear polyhedrosisvirus in populations of Spodoptera exigua in southern Spain. Biocontrol Sci. Tech. 2: 35-38. Catteneo, R., Schmid, A . , Eschle, D . , Baszko, K. , ter Muelen, V . , and Billiter, M . A . 1998. Biased hypermutation and other genetic changes in defective measles viruses in human brain infections. Cell. 55:255-265. Cherry, C . L . , and Summers, M . D . , 1985. Genotypic variation among wild isolates of two nuclear polyhedrosis viruses isolated from Spodoptera littoralis. J. Invertebr Pathol. 46: 289-295 Collett, D . 1991. Modeling binary data. Chapman and Hall Inc., London. Cory, J.S. 2000. Assessing the risk f releasing genetically modified virus insecticides: progress to date. Crop Protec. 19:779-785. Cory, J.S., Hails, R.S., and Sait, S .M. 1997. Baculovirus ecology. In: L .K.Mi l l er (ed.) The Baculoviruses. pp. 301-339. Plenum, New York. Cory, J.S., Hirst, M . , Williams, T . , Hails, R., Goulson, D. , Green, B. , Carty, T. , Possee, R., Cayley, P.J., and Bishop, D . H . L . 1994. Field trial of genetically improved baculovirus insecticide. Nature (London). 370(6485): 138-140. 61 Cory, J.S., Hirst, M . L . , Sterling, P .H. , and Speight, M . R . 2000. Narrow host range nucleopolyhedorvirus for control of the browntail moth (Lepidoptera: Lymantridae). Biol Contr. 23(3):661-667. Crozier, G . , Crozier, L . , Quiot, J . M . , and Lereclus, D . 1988. Recombination of Autographa californica and Rachiplusia ou nuclear polyhedrosis viruses in Galleria mellonellah. J. Gen. Virol. 69:179-185. Crozier, G . , and Ribeiro, H . C . T . 1992. Recombination as a possible major cause of genetic heterogeneity in Anticarcia gemmatalis nuclear polyhedrosis virus wild populations. VirRes. 26:183-196 Doane, C . C . 1967. Bioassay of nuclear-polyhedrosis virus against larval instars of the gypsy moth. J. Invertebr. Pathol. 9:376-386. Doyle, C.J . , Hirst, M . L . , Cory, J.S., and Entwhistle, P.F. 1990. Risk assessment studies: Detailed host range testing of wild-type cabbage moth, Mamestra brassicae (Lepidoptera:Noctuidae) nuclear ployhedrosis virus. Appl. Environ. Microbiol. 56:2704. Evans, H.F . and Harrap, K . A . 1987. Persistence of insect viruses. In "Virus persistence: Symposium 33." Society for General Microbiology Ltd., Cambridge Univ. Press. Fine, P . E . M . 1984. Vertical transmission of pathogens of invertebrates. In "Comparative pathobiology: Pathogens of invertebrates" (T.C. Cheng, Ed.) V o l 7: pp 205-241. Fuxa, R.J. and Richter,A.R. 1991. Selection for an increased rate of vertical transmission of Spodoptera frugiperda (Lepidoptera:Noctuidae) nuclear polyhedrosis virus. Environ. Entomol. 20:603-609. Fuxa, J.R., Weidner, E . H . , and Richter, A.R. 1992. Polyhedra without virions in a vertically transmitted nuclear polyhedrosis virus. / . Invertebr. Pathol. 60:53-58. Fuxa, J.R., Sun, J.-Z. , Weidner, E . H . , and LaMotte, L . R . 1999. Stressors and rearing diseases of Trichoplusia ni: Evidence of vertical transmission of N P V and C P V . / . Invertebr Pathology. 74:149-155. Gelernter, W . D . , and Federici, B . A . 1990. Virus epizootics in Californian populations of Spodoptera exigua: Dominance of a single viral genotype. Biochem. System. Ecol. 18:461-4663 Gettig, R . G . , and McCarthy, W.J. 1982 Genotypic variation among wild isolates of Heliothis spp. nuclear polyhedrosis viruses from different geographic regions. J. Virol. 43: 174-181. Godfray, H . C . J . , O'Reilly, D.R. and Briggs, C.J. 1997. A model of nucleopolyhedrovirus (NPV) population genetics applied to co-occlusion and the spread of the few polyhedra (FP) phenotype. Proc. R. Soc. Lond. B. 264:315-322. Guy, R . H . , Leppla, N . C . , Rye, J.R., Green, C .W. , Barrette, and S.L. Hollien, K . A . 1985. Trichoplusia ni. In "Handbook of insect rearing" (P.Singh and R . F Moore, Eds.), Vol.11, pp487-494. Hamilton, M . A . , Russo, R . C . , and Thurston, R . V . 1977. Trimmed Spearman-Karber method for estimating median lethal concentrations. Environ. Sci. Tech. 11(7):714-719. Harper, J.D. 1976. Cross-infectivity of six plusiine nuclear polyhedrosis virus isolates to plusiine hosts. J. Invertebr. Pathol. 27:275-277'. Hodgson, D.J. , Vanderbergen, A . J . , Watt, A . D . , Hails, R.S., and Cory, J.S. 2001 Phenotypic variation between naturally coexisting genotypes of a Lepidopteran baculovirus. In press Hostetter, D . L . and Puttier, B . 1991. A new broad host spectrum nculear polyhedrosis virus isolated from a celery looper, Anagrapha falcifera (Kirby), (Lepidoptera:Noctuidae). Environ. Entomol. 20:1480-1488. Howard, C R . 1986. The biology of hepadnaviruses. J. Gen Virol. 67:1215-1235. Hughes, D.S. , Possee, R .D. , and King, L . A . 1993. Activation and detection of a latent baculovirus resembling Mamestra brassicae nulcear polyhedrosis virus in M. brassicae insects. Virol. 194:608-615. Hughes, D.S. , Possee, R .D. , and King, L . A . 1997. Evidence for the presence of a low-level, persistent baculovirus infection of Mamestra brassicae insects. J. Gen. Virol. 78:1801-1805. Jurcovicova, M . 1979. Activation of latnet infections in larvae of Adoxophyes orana (Lepidoptera: Torticidae) and Baratha brassicae (Lepidoptera: Noctuidae) by foreign polyhedra. J. Invertebr. Pathol. 34:213-223. Keddie, A . and Erlandson, M . 1995. Characterization of a nuclear polyhedrosis virus of the forest tent caterpillar, Malacosoma disstria. J. Invertebr. Path. 65(l):43-47. Kislev, N . and Edelman, M . 1982. D N A restriction pattern differences from geographic isolates of Spodoptera littoralis nuclear polyhedrosis virus. Virol. 119:219-222. Knell, J.D., and Summers, M . D . , 1981. Investigation of genetic heterogeneity in wild isolates of Spodoptera frugiperda nuclear polyhedrosis virus by restriction endonuclease analysis of plaque-purified variants. Virol. 112:190-197. Kondo, A . , Maeda, D. 1991. Host range expansion by recombination of the baculoviruses Bombyx mori nuclear polyhedrosis virus and Autographa californica nuclear polyhedrosis virus. J. virol. 65:3625-3632. Kukan, B . 1996. "The occurrence and persistence of nuclear polyhedrosis virus in fluctuating populations of tent caterpillars." Ph.D. dissertation. Univ of British Columbia, Vancouver. Kukan, B . 1999. Vertical transmission of nucleopolyhedrovirus in insects. J. Invert. Pathol. 74:103-111. Lautenshlager, R . A . , Podgewaite, J.D. and Watson, D . E . 1980 Natural occurrence of the nuclear polyhedrosis of of the gypsy moth, Lymantira dispar (Lep.:Lymantriidae) in wild birds and mammals. Entomophaga 25:260-265. Lee, H . H . and Miller, L K . , 1978. Isolation of genotypic variants of Autographa californica nuclear polyhedrosis virus. / . Virol. 27:754-767. Longworth, J.F. and Cunningham, J.C. 1968. The activation of occult nuclear-polyhedrosis viruses by foreign nuclear polyhdra. J. Invertebr. Pathol. 10:361-367. Maeda, S., Kamita, S.G. , and Kondo, A . 1993. Host range expansion of Autographa californica nuclear polyhedrosis virus (NPV) following recombination of a 0.6-Kilobase-pair D N A fragment originating from Bombyx mori N P V . J. Virol. 67(10): 6234-6238. Maeda, S., Mukohara, Y . , and Kondo, A . 1990. Characteristically distinct isolates of the nuclear polyhedrosis virus from Spodoptera litura. J.Gen. Virol. 71:2631. Martin, D .W. , and Weber, P.C. 1997. D N A replication promotes high-frequency homologous recombination during Autographa californica multiple nuclear polyhedrosis virus infection. Virol. 32:300-309. Maruniak, J. E . Garcia-Maruniak, A . Souza, M . L . Zanotto, and P. M . A . Moscardi, F. 1999. Physical maps and virulence of Anticarsia gemmatalis nucleopolyhedrovirus genomic variants. Arch, of Virol. 144: 10. McEwen, F . L . and Hervey, G .E .R . 1960. Mass-rearing the cabbage looper, Trichoplusia ni, with notes on its biology in the laboratory. Ann. Entomol. Soc. Am. 53:229-234. McKinley, D.J . , Brown, D . A . , Payne, C C , and Harrap, K . A . 1981. Cross infectivity and activation studies with four baculoviruses. Entomophaga 26:79-90. Mellerick, D . M . and Fraser, N.W. 1987. Physical state of the latent herpes simplex virus genome in a mouse model system: Evidence supporting an episomal state. Virology 158:265-275. Milks, M . 1997. Comparative biology and susceptibility of Cabbage Looper (Lepidoptera: Noctuidae) lines to a nuclear polyhedrosis virus. Environ. Entomol. 24(4):839-846. Murray, K . D . Elkington, J.S. 1989. Environmental contamination of egg masses as a major component of trans-generational transmission of gypsy moth nuclear polyhedrosis virus (LdMNPV) . J. Invert. Pathol. 53:324-334. Myers, J .H. 1988. Can a general hypothesis explain population cycles of forest Lepidoptera? Adv. Ecol. Resear. 18:179-242. Myers, J .H. 2000. Population fluctuation of the western tent caterpillar in southwestern British Columbia. Pop Ecol. 42:231-241. Myers, J .H. 1990 Population cycles of western tent caterpillars: Experimental introduction and synchrony of fluctuations. Ecology. 71:986-995 Myers, J .H . , and Kukan, B. , 1995. Changes in the fecundity of tent caterpillars: A correlated character of disease resistance or sublethal effects of disease? Oecologia 103:475-480 Myers J .H. , and Rothman, L . E . 1995. Virulence and transmission of infectious diseases in human and insects: evolutionary and demographic patterns. TREE. 10(5): 194-198 Munoz, D. , and Cabellero, D . 2000. Persistence of parasitic genotypes in a mixed population of SeNPV. Biol Con. 19:259-264. Munoz, D . , Murillo, R., Krell, P.J., Vlak, J . M . , and Caballero, P. 1999. Four genotypic 66 variants of a nucleopolyhedrovirus (Se-SP2) are distinguishable by a hypervariable genomic region. Vir Res. 59:61-74. Nielsen, C , Short, S .M. , Cooper.D.M., and Suttle,C. 2001. D N A polymersase sequence suggests low divergence between western and forest tent caterpillar NPVs. (in prep) Neeglund, Y . F . and Mathad, S.B. 1978. Transmission of nuclear polyhedrosis virus in laboratory population of the army worm, Mythimna (Pseudaletia) separata. J. Invert. Pathol. 31:143-147. O'Reilly, D.R. , Hails, R.S. and Kelly, T .J . 1998. The impact of host developmental status on baculovirus replication. J. Invert Pathol. 72:269-275. Pair, S.D., Wiseman, B.R. , and Sparks, A . N . 1986 Influence of four corn cultivars on fall armyworm (Lepidoptera: Noctuidae) establishment and parasitization. Flor. Entomol. 69(3):566-570. Reeson, A . F . , Wilson, K . , Gunn, A . , Hails, R.S., and Goulson, D . 1998. Baculovirus resistance in the noctuid Spodoptera exempta is phenotypically plastic and responds to population density. Proc. R. Soc. Lond. B 265: 1787-1791. Rothman and Myers, J .H. 1996. Is fecundity correlated with resistance to viral disease in the western tent caterpillar? Ecol. Ent. 21:396. Shapiro, D.I., Fuxa, J.R., Braymer, H . D . , and Pashley, D.P. , 1991 D N A restriction polymorphism in wild isolates of Spodoptera frugiperda nuclear polyhedrosis virus. J. Invertebr. Pathol. 58: 96-105. Smith, K . M . 1963. The cytoplasmic virus diseases. In "Insect Pathology: A n advanced Treatise" (E.A. Steinhaus, Ed.), Vol . 1:457-497. Smith, G . E . and Summers, M . 1978 Analysis of baculovirus genomes with restriction endonucleases. Virol. 89:517-527. Stiles, S., and Himmerich, B . 1998. Autographa californica N P V isolates: restriction endonulcease analysis and comparative biological activity. J Invertebr. Pathol. 72:174-177. Smits, P .H. , and Vlak, J . M . 1988. Biological activity of Spodoptera exigua nuclear polyherosis virus against S. exigua larvae. J. Invertebr. Pathol. 51:107-114. Tomalski, M . D . and Miller, L . K . 1991. Insect paralysis by baculovirus-mediated expression of a mite neurotoxin gene. Nature. 352:82-85. Washburn, J.O., Kirkpatrick, B . A . , Haas-Stapleton, E . , and Volkman, L . E . 1998. Evidence that the stilbene-derived optical brightner M 2 R enhances Autographa californica M nucleopolyhedrovirus infection of Trichoplusia ni and Heliothis virescens by preventing sloughing of infected midgut epithelial cells. Biol. Cont. 11:58-69. Wellington, W. 1960. Qualitative changes in populations in unstable environments. Can. Entomol. 96:436-451. Wellington, W. 1964. Qualitative changes in populations in unstable environments. Can. Entomol. 96:436-451. Young ,S.Y. and Yerian, W . C . 1982. Nuclear polyhedrosis virus infection of Pseudoplusia includens (Lep.:Noctuidae) larvae: Effect on post larval stages and transmission. Entomophaga. 27:61-66. APPENDIX 1.0 T H E INFLUENCE OF HOST P L A N T ON BACULOVIRUS E F F I C A C Y Despite abilities to genetically enhance the efficacy of baculoviruses, their use as biopesticides is hampered by resistance of insects to fatal infection under different circumstances. There is strong evidence to suggest that sloughing of mid-gut cells is responsible for developmental resistance in larvae. Englhard and Volkman (1995) were able to establish the presence and location of an A c M N P V infection 48 hours after the inoculation of both Trichoplusia ni and Heliothis virescens which allowed them to document an age dependent rate of establishing and/or sloughing of infected mid-gut cells. The virus therefore has a decreasing window of opportunity to establish systemic infections in progressively older insects (Hoover et al., 2000). Several studies have shown that rapid movements of A c M N P V into the tracheal epidermis is crucial to the outcome of infection because hosts can clear primary infection by sloughing infected mid-gut cells (Washburn et al., 1995, 1998). Tracheal epidermal cells cannot be sloughed, therefore successful infection of the tracheal system inevitable results in death. Once systemic infection has established, replacement of infected cells with new cells benefits the virus by providing additional host biomass for viral amplification (Hoover et al., 2000). Decreased susceptibility to polyhedrosis disease is known to result from interactions among the host insect, the virus and the chemistry of the food plant (Granados and Federici, 1986; Young et al., 1995). These have all been postulated to cause accelerated sloughing rates (Hoover et al., 2000). For example, Heliothis virescens is markedly more susceptible to virus-induced mortality when inoculum was fed on lettuce or tomato than on cotton (Hoover et al., 1998). Here, foliar peroxidase activity, a plant oxidative enzyme was negatively correlated with A c M N P V induced mortality. Similarly, Forschler et al. (1992) found that percent lethal infection of H. zea was lowest in larvae feeding on cotton and sorghum compared to soybean and tomato. Cotton and sorghum are plants high in hydrolizable tannins. Young et al. (1995) established that leaf phenolics may present a barrier to viral infection in insect herbivores. Phenolics such as tannins are polyphenolic compunds, capable of binding to protein, reducing the activity of many enzymes. In addition, tannins as secondary metabolites, are potent prtein denatruants and consequently may act as enzyme inhibitors, antisporulants and or mild antibiotocs. If host plants contain tannins and other plant chemicals that provide antiviral properties, the prevalence of disease may be somewhat dependent on host plants. 69 In this study I examined the effect of host plant on virus-induced mortality by challenging forest tent caterpillars (Malacosoma disstria (Hiibner)) with M a d i N P V on it's host plant, trembling aspen (Populus tremuloides) or the alternate, alder (Alnus rubra). If host plants confer resistance to viral disease then we would expect t see a decreased mortality and an increased LD50 in those larvae fed aspen foliage. Larvae used in this study were 3 r d and 4 t h instars hatched from egg masses collected from the field during the spring of 1998. Egg masses were surface sterilized as described in Chapter 2. Larvae were fed and monitored daily. Third and fourth instars were removed immediately following molt and starved for 24 hours. Larvae were divided into two groups: aspen or alder. Virus used as inoculum was the MadiNPV stock described in Chapter 2. Viral doses ranged from 130-8800 for 3 r d instars and 880-13000 for 4 t h instars for both food plant experiments (Table A . l ) . Larvae were fed individually and given 24 hours to consume a minimum of 95% of an O B contaminated leaf disc (50 mm 2). Those not consuming the required amount were not included in the study. Caterpillars were then placed as groups of 10-15 in (0.5 L) paper coffee cups where they were fed either aspen or alder foliage for 14 days. Viral mortality was recorded daily. No mortalities were experienced in controls. A general linear model was used to analyse bioassay data. Larval mortality was transformed using a logit transformation. Food plant was the categorical variable and log dose was the covariate. There was significant dose effect for both bioassays at the 3 r d and 4 t h instar stage (Tables A. 1.2 and A . 1.3) however, the effect of dose was different between the different instars. A n increasing mortality was associated with increasing dose for 3 r d instars where as mortality was highest when dose was lowest for 4 t h instars (Figure A . 1.3). There was a significant host plant effect detected at the third instar stage (Table A . 1.2). In contrast, no host plant effect was detected at the 4 t h instar stage (Table A. 1.3). LD50S could not be calculated for 3 r d instars fed Aspen because mortalities were not 50% or greater, and could not be calculated for either 4 t h instar bioassay because all mortalities were above 50%. The LD50 for 3 r d instars fed Alder was 3235 (1033, 10127). A significant reduction in viral induced mortality for larvae challenged with MadiNPV on its host plant Aspen, was seen only at the 3 r d instar stage. Although I believe this trend to be real, high over-all mortalities experienced by 4 t h instar larvae suggest that contamination may have been a factor contributing to background mortality in the bioassays. Larvae used in this particular study came from the same collection of egg masses that were shown to carry a latent 70 or sublethal infection in chapter 2. It is likely that levels of mortality may have been inflated by an infection already present within larvae themselves. There is no indication that this was a factor with 3 r d instar larvae however, without molecular confirmation of the virus produced from these infections it is difficult to rule out sublethal or latent contamination. Results from other studies suggest that future experiments should re-visit the effects of host plants on viral infection and should include other measurements of other viral life history traits such as speed of kill. Questions concerning the nature and timing of antiviral effects can be addressed by providing only the initial viral challenge on a variety of host plants. Larvae can then be reared on similar diets. 71 Table A. 1.1 Doses and number of larvae fed for Aspen/Alder food Instar Log dose (# OBs) # larvae fed (Aspen/Alder) 3 r d 2.11 (130) 15/13 3 r d 2.94 (880) 15/12 3 r d 3.34 (2200) 12/13 3 r d 3.94 (8800) 14/12 4 t h 2.94 (880) 9/10 4 t h 2.94 (880) 9/10 4 t h 3.94 (8800) 8/9 4 t h 3.94 (8800) 9/8 4 t h 2.94 (880) 8/9 4 t h 3.11 (1300) 9/10 4* 3.11 (1300) 9/9 4 t h 3.11 (1300) 8/9 4 t h 3.94 (8800) 8/7 4 t h 4.11 (13000) 7/9 4 t h 4.11 (13000) 7/8 4* 4.11 (13000) 7/9 72 Table A.1.2. ANCOVA of the effects of food plant on virus-induced mortality experienced by 3r instar M. disstria. Source df Sums of Squares F ratio Prob>F Food plant 1 1.1627944 12.3585 0.0246 Log dose 1 2.6359938 28.0161 0.0061 Plant*log dose 1 0.1376545 1.4630 0.2930 Table A. 1.3. ANCOVA of the effects of food plant on virus-induced mortality experienced by 4 instar M. disstria. Source df Sums of squares F ratio Prob>F Food plant 1 0.5744395 1.4484 0.2428 Log dose 1 5.4108887 13.6433 0.0014 Plant*log dose 1 1.207419 3.0503 0.0961 3rd instar 3 -, o Aspen 2 -mortality 1 -0 -' ' ' o o Alder O » Linear ogit -1 -Q - -. - - O . . . .— — < ? — " (Aspen) -2 -o - - - - Linear (Alder) -J 1 1 1 1 2 2.5 3 3.5 4 log dose 4th instar re r o E o 3 2H 1 0 -1 H -2 -3 2.5 O O 3.5 4 log dose 4.5 o Aspen o Alder Linear (Aspen) I • Linear (Alder) Figure A. l . Mortality experienced by 3 r d and 4 l h instar M. disstria larvae challenged with MadiNPV on either Aspen (•••) and alder (o). Diamonds and circles represent actual mortality and solid and dashed lines represent Fitted logit mortality (3rd instar: Aspen logit(mortality)=-3.1959+0.6662(log dose), alder logit (mortality)=-3.6499+1.0608(log dose) and 4 t h instar: Aspen logit (mortality)=5.3826-1.379(log dose), alder logit(mortality)=1.9484-0.4936(logdose)) 74 APPENDIX 2.0 M O L E C U L A R C H A R A C T E R I Z A T I O N OF T H E NUCLEOPOLYHEDROVIRUS INFECTING T H E FOREST TENT C A T E R P I L L A R , MALACOSOMA DISSTRIA (HUBNER), IN PRINCE G E O R G E BRITISH COLUMBIA. In the springs of 1998 and 2000, natural MadiNPV populations infecting the forest tent caterpillar, Malacosoma disstria (Hiibner), were examined for genetic variability using restriction endonuclease analysis (REN). M. disstria larvae were collected from five sites around Prince George, British Columbia (Table A.2.1). Larvae were reared as described in Chapter 2. MadiNPV D N A purification and restriction analysis was performed as described in the methods sections of chapters 2 and 3. Genetic variation was examined using £ c o R I , HindIII and XhoI. One dominant genetic variant was assigned to each viral isolate as described in Chapter 2.. MadiNPV was genetically variable with each enzyme used (Figures 1-3). MadiNPV isolates collected in 1998 were examined with both EcoRl and XhoI. A l l samples collected in 2000 were examined with HindIII only. Genetic variation was found to occur among the isolates sampled in both 1998 and 2000. Variation was found both within and between host sites. Genetic variation with XhoI XhoI digestion patterns for ten MadiNPV isolates collected in 1998 are shown in Figure A.2.1. Three R E N variants were found among all the isolates examined ( V x l , Vx2, Vx3). Twelve fragments, B , C , D , F , G , H , I , J , M , N , 0 , and P were common all isolates. Variations occurred in fragments E (10.5 kb), K (6.2 kb), L (5.6), and Q (0.9 kb). Genetic variation with EcoRl EcoRl digestion patterns were the most genetically variable (Figure A.2.2). Eight R E N variants were found from eight MadiNPV isolates collected in 1998. Twenty one fragments, A , B , C , D , E , F , G , H , J , K , L , M , M , 0 , P , Q , R , S , T , U , and V were common to all isolates. Variations occurred in fragments I (5.4 kb), W (), X (), and Y(). Variations also occurred with the addition of fragments labeled D+l,D+2,D+3, O+l , and 0+2 Genetic variation with HindIII 75 A total of 30 isolates from five sites were digested with HindIII. Ten fragments, A , B , C , D , E , F , G , H , I , and J were common to all isolates. Variations occurred with the addition of fragments E + l , E+2, and G + l . A total of four genetic variants were identified among all the host sites sampled (Figure A.2.3). A l l genetic variants were found in all of the sites sampled, with the exception of the West Lake isolates, which were represented by V2 and V 4 only (Table A.2.1). The two most common genetic variants were V4 found in 12 isolates, and V2 found in 12 isolates (Figure A.2.1). 76 Table A.2.1 Frequencies of HindlU genetic variants from MadiNPV isolates collected from five M. disstria populations around Prince George, British Columbia. Host Site REN variant (a) Truck Vl(2) , V2(2), V3(l) , V4(4) Johnson Road V3( l ) Airport V4( l ) University Vl(2) , V2(4), V3(2), V4(5) West Lake V2(6), V4(2) a the number of isolates sharing the same R E N profile. Vx Vx Vx Vx Vx Vx Vx Vx Vx Vx Figure A.2.1. Xhol restrcition disgests of 1998 MadiNPV isolates. A total of three variants were found from 15 isolates sampled. The last profile on the left is the X-Hindlll molecular marker. 78 Figure A.2.2. Restriction digests (EcoRl) of MadiNPV isolates collected from field populations of M. disstria near Prince George British Columbia. Fragments are labelled to the right of Variant 3 (Ve3) and the left of variants 4 (Ve4). A total of eight genetic variants were foundamong the eight isolates sampled. 79 Figure A.2.3. HindUI restriction digests of MadiNPV isolates collected from five M. disstria sites around Prince George British Columbia. Fragments are labeled (A-J) on the left and side of variant I (VI). New fragments are indicated by a letter/number combination. The numbers in brackets represent the number of isolates sharing the same REN profile. A total of four genetic variants were found among the 30 isolates sampled. All genetic variants were found in each of the host sites sampled. L is the X-Hindlll molecular marker. 80 

Cite

Citation Scheme:

        

Citations by CSL (citeproc-js)

Usage Statistics

Share

Embed

Customize your widget with the following options, then copy and paste the code below into the HTML of your page to embed this item in your website.
                        
                            <div id="ubcOpenCollectionsWidgetDisplay">
                            <script id="ubcOpenCollectionsWidget"
                            src="{[{embed.src}]}"
                            data-item="{[{embed.item}]}"
                            data-collection="{[{embed.collection}]}"
                            data-metadata="{[{embed.showMetadata}]}"
                            data-width="{[{embed.width}]}"
                            async >
                            </script>
                            </div>
                        
                    
IIIF logo Our image viewer uses the IIIF 2.0 standard. To load this item in other compatible viewers, use this url:
http://iiif.library.ubc.ca/presentation/dsp.831.1-0089978/manifest

Comment

Related Items