SYSTEMATICS OF THE ENTOMOPATHOGENIC BACTERIA BACILLUS POPILLIAE, BACILLUS LENTIMORBUS, AND BACILLUS SPHAERICUS

Karen Rippere Lampe

Dissertation submitted to the Faculty of the Virginia Polytechnic Institute and State University in partial fulfillment of the requirements for the degree of

DOCTOR OF PHILOSOPHY in Biology

Allan A. Yousten, Chair Noel R. Krieg Khidir W. Hilu Eric A. Wong David L. Popham

September 11, 1998 Blacksburg, Virginia

Keywords: Bacillus popilliae, Bacillus lentimorbus, Bacillus sphaericus, DNA reassociation, RAPD, vancomycin resistance SYSTEMATICS OF THE ENTOMOPATHOGENIC BACTERIA BACILLUS POPILLIAE, BACILLUS LENTIMORBUS, AND BACILLUS SPHAERICUS

Karen Rippere Lampe A. A. Yousten, Chairman Department of Biology (ABSTRACT)

Bacillus popilliae and B. lentimorbus, causative agents of milky disease in Japanese beetles and related scarab larvae, have been differentiated based upon a small number of phenotypic characteristics, but they have not previously been examined at the molecular level. Thirty- four isolates of these bacteria were examined for DNA similarity. Three distinct but related similarity groups were identified; the first contained strains of B. popilliae, the second contained strains of B. lentimorbus, and the third contained two strains distinct from but related to B. popilliae. Some strains received as B. popilliae were found to be most closely related to B. lentimorbus and some received as B. lentimorbus were found to be most closely related to B. popilliae.

Geographically distinct strains of B. popilliae and B. lentimorbus were analyzed using RAPD. Eight decamer primers were tested against nineteen new and seventeen isolates previously described by randomly amplified polymorphic DNA (RAPD) analysis (M. Tran). Of the new isolates, ten were found to be B. popilliae while nine isolates were more related to the B. lentimorbus species. Paraspore formation, believed to be a characteristic unique to B. popilliae, was found to occur among a subgroup of B. lentimorbus strains.

Using a combination of two PCR primer pairs, the cry18Aa1 gene was detected in 31 of 35 B. popilliae isolates and in 1 of 18 B. lentimorbus isolates. When hemolymph smears were examined microscopically, a parasporal crystal was seen in three of the four B. popilliae strains where the PCR primers could not amplify the paraspore gene. The fourth strain was not tested due to the unavailability of infected hemolymph. A paraspore was also detected by microscopic examination in a subgroup of 14 B. lentimorbus strains. In combination, the primer pairs CryBp1 and CryBp2 are effective at detecting the paraspore gene in B. popilliae isolates, but not in the B. lentimorbus isolates. Growth in media supplemented with 2% NaCl was found to be less reliable in distinguishing the species than was vancomycin resistance, the latter present only in B. popilliae.

The basis for vancomycin resistance in all isolates was investigated using a polymerase chain reaction assay designed to amplify the vanB gene in enterococci. An amplicon was identified and sequenced. The amplified portion of the putative ligase gene in B. popilliae had 77% and 68-69% nucleotide identity to the sequences of the vanA gene and the vanB genes, respectively. There was 75% and 69-70% identity between the deduced amino acid sequence of the putative ligase gene in B. popilliae and the deduced amino acid sequence of the vanA gene and the vanB genes,

iii respectively. It has been determined that the vanE gene is located either on a plasmid greater than 16 kb in size or on the chromosome. The gene in B. popilliae may have had an ancestral gene in common with vancomycin resistance genes in enterococci. Bacillus sphaericus strains isolated on the basis of pathogenicity for mosquito larvae and strains isolated on the basis of a reaction with a B. sphaericus DNA homology group IIA 16S rRNA probe were analyzed for DNA similarity. All of the pathogens belonged to homology group IIA, but this group also contained nonpathogens. It appears inappropriate to designate this homology group a species based solely upon pathogenicity.

iv ACKNOWLEDGEMENTS

First and foremost, I would like to thank my advisor Dr. Allan Yousten for his guidance and wisdom throughout this project. Without his input, I would never have made it though this program. I would also like to thank my committee members, Dr. N. R. Krieg, Dr. K. W. Hilu, Dr. E. Wong and Dr. D. L. Popham for their help, advice and support. The late Dr. John L. Johnson enabled me to get started on this project and we miss him very much. To my family, thanks for believing in me and supporting me both emotionally and financially. Finally, to all the other graduate students on the hall; without you it wouldn't have been as much fun.

v TABLE OF CONTENTS

Page ABSTRACT...... ii ACKNOWLEDGEMENTS...... v LIST OF FIGURES...... ix LIST OF TABLES...... xi INTRODUCTION...... 1 I. REVIEW OF THE LITERATURE Use of Bacillus popilliae and Bacillus lentimorbus as biological control agents...... 4 Pathology of Bacillus popilliae and Bacillus lentimorbus...... 6 Physiology of Bacillus popilliae and Bacillus lentimorbus...... 7 Genetics of Bacillus popilliae...... 10 Taxonomy of Bacillus popilliae and Bacillus lentimorbus...... 12 DNA-DNA similarities...... 15 RAPD analysis...... 16 Vancomycin resistance...... 18 References...... 24 II. MATERIALS AND METHODS Media and Reagents...... 33 Bacterial strains and growth conditions...... 35 Isolation of bacteria from dried beetle hemolymph...... 38 DNA isolation for DNA-DNA reassociation...... 38 DNA sample preparation...... 39 DNA labeling...... 40 S1 Nuclease assay...... 42 DNA isolation for RAPD experiments...... 43

vi Determination of DNA concentration...... 44 RAPD analysis...... 45 Isolation, amplification and digoxygenin labeling of individual RAPD bands...... 47 Estimation of probe yield...... 49 Southern transfer and hybridization...... 50 RAPD band analysis...... 52 Data analysis...... 52 Multiplex PCR-RFLP for detection of the van ligase....53 Paraspore gene detection using PCR...... 54 PCR product sequencing...... 55 Labeling of the vanE PCR product...... 56 Determination of vanE location in B. popilliae...... 56 References...... 56 III. BACILLUS POPILLIAE AND BACILLUS LENTIMORBUS, BACTERIA CAUSING MILKY DISEASE IN JAPANESE BEETLES AND RELATED SCARAB LARVAE Abstract...... 58 Results DNA similarity...... 60 Growth in 2% NaCl or vancomycin...... 63 Discussion...... 63 References...... 65 IV. RANDOMLY AMPLIFIED POLYMORPHIC DNA ANALYSIS OF GEOGRAPHICALLY DISTINCT ISOLATES OF BACILLUS POPILLIAE AND BACILLUS LENTIMORBUS Abstract...... 68 Results RAPD analysis...... 68 Growth in 2% NaCl or vancomycin...... 75 Discussion...... 78 References...... 80

vii V. IDENTIFICATION AND DETECTION OF THE CRY GENE IN STRAINS OF BACILLUS POPILLIAE AND BACILLUS LENTIMORBUS Abstract...... 82 Results Detection of the cry operon...... 82 Discussion...... 90 References...... 92 VI. DNA SEQUENCE RESEMBLING VANA AND VANB IN THE VANCOMYCIN- RESISTANT BIOPESTICIDE BACILLUS POPILLIAE Abstract...... 93 Results...... 95 Discussion...... 101 References...... 102 VII. DNA SIMILARITIES AMONG MOSQUITO-PATHOGENIC AND NONPATHOGENIC STRAINS OF BACILLUS SPHAERICUS Abstract...... 105 Bacteria and DNA isolation...... 106 DNA similarities...... 107 Results and Discussion...... 108 References...... 109 SUMMARY...... 111 CONCLUSIONS...... 114 CURRICULUM VITAE...... 116

viii LIST OF FIGURES

Page CHAPTER THREE Figure 1. Distance dendogram of B. popilliae and B. lentimorbus strains generated from DNA similarity analysis...... 62

CHAPTER FOUR Figure 1. RAPD banding patterns of B. popilliae and B. lentimorbus isolates using primer OPA-03...... 70

Figure 2. RAPD banding patterns of B. popilliae and B. lentimorbus isolates using primer OPA-03...... 71

Figure 3. RAPD banding patterns of B. popilliae and B. lentimorbus isolates using primer OPA-15...... 72

Figure 4. RAPD banding patterns of B. popilliae and B. lentimorbus isolates using primer OPA-15...... 73

Figure 5. Dendogram showing the relationships between strains of B. popilliae and B. lentimorbus generated from RAPD analysis...... 74

CHAPTER FIVE Figure 1. Structure of the Bacillus popilliae cry operon...... 83 Figure 2. ATCC 14706 and NRRL B-4081 PCR products using primer pair CryBp2...... 86

Figure 3. B. popilliae cry18Aa1 gene sequences...... 87

Figure 4. Deduced amino acid sequence comparison of B. popilliae cry genes...... 89

CHAPTER SIX Figure 1. Multiplex PCR-RFLP of enterococcal isolates carrying the vanA and vanB ligase genes and B. popilliae ATCC 14706...... 96

ix Figure 2. Sequence comparisons of the putative ligase genes in B. popilliae isolates...... 97

Figure 3. Comparison of the translation of the putative ligase gene in B. popilliae ATCC 14706 to the translations of four previously characterized vanB genes(isolates 55, 94, 45, and 91) and one vanA gene (isolate)...... 98

Figure 4. Southern blot of digested and undigested B. popilliae chromosomal DNA probed with the vanE PCR product...... 100

x LIST OF TABLES

Page

CHAPTER TWO Table 1. B. popilliae and B. lentimorbus strains used in DNA-DNA reassociation...... 36

Table 2. B. popilliae and B. lentimorbus strains used in RAPD analysis from diverse host insects and geographic regions...... 36

Table 3. RAPD primer sequences...... 45

Table 4. Dilution series for probe estimation...... 49

Table 5. Primer sequences used in multiplex PCR-RFLP reaction for detection of van ligase genes in enterococci...... 53

Table 6. Primer sequences used for detection of cry genes in B. popilliae and B. lentimorbus...... 55

CHAPTER THREE Table 1. Levels of DNA similarity between B. popilliae and B. lentimorbus as determined by the S1 nuclease method...... 61

Table 2. Characteristics of B. popilliae and B. lentimorbus strains used in DNA similarity studies...... 63

CHAPTER FOUR Table 1. Characteristics of B. popilliae and B. lentimorbus isolates from diverse host insects and geographical regions...... 77

CHAPTER FIVE Table 1. Detection of the paraspore crystal in strains of B. popilliae and B. lentimorbus by visualization and PCR...... 84

CHAPTER SEVEN Table 1. Bacillus sphaericus strains studied using

xi DNA-DNA similarity analysis...... 107

Table 2. Levels of DNA similarity among strains of B. sphaericus...... 109

xii INTRODUCTION

Classification schemes illustrating the relationships among living organisms have been documented since the writings of Aristotle, and the classification of bacteria since the different morphological forms were described by Muller in 1773. Bacterial classification began by organization into taxonomic units based solely on the morphological characteristics of the organisms and has progressed to the use of a wide variety of characteristics including the physiology, biochemistry and genetic material of the bacteria. Today, the sheer number of bacterial species that have been identified and the wide diversity among them make classification of these organisms into discreet arrangements both difficult and necessary.

Classification can be described as having three major purposes. The arrangement of organisms into discrete groups provides a way to summarize and catalog information about them. The classification takes the form of a database in which information about an organism can be stored and retrieved by the use of a particular name. The classification can be used to predict the properties of a group of organisms so that members may be recognized by their defining characteristics. The organization of organisms into groups by classifying them must be accomplished before an identification system can be created which will recognize new isolates. Finally, classification systems can provide insights into the evolutionary origins and relationships among organisms. To fulfill these purposes, classifications should contain as much information as possible, be stable and should be based on empirical evidence.

1 The species concept is less rigorously defined for bacteria than for other organisms. Bergey's Manual defines the bacterial species as "a collection of strains that share many features in common and differ considerably from other strains." It goes on to say that "a species consists of the type strain and all other strains that are considered to be sufficiently similar to it as to warrant inclusion with it in the species." A more uniform definition of the bacterial species is desireable and can possibly be obtained through the use of genetic relatedness among bacteria.

Microbiologists typically use two different types of classifications, phenetic classifications and special purpose classifications. Phylogenetic classifications are beginning to be developed with the information provided by macromolecule sequencing but have only been applied to select bacterial groups. Phenetic classifications encompass all bacteria and are useful to all microbiologists, regardless of their specific discipline. They are organized using affinities based on the phenotype and genotype of organisms as they exist in the present, with no regard for evolutionary context. Special purpose classifications are designed for a particular discipline. These systems are often based on a single feature which is thought to be sufficient and necessary for the placement of an organism within a group. A disadvantage of these systems is that they are based on very little information and therefore tend to be unstable. Due to the lack of information, an unknown organism that is lacking the single essential feature of the classification would be assigned to the wrong taxon.

2 Classification of bacteria allows microbiologists to associate certain characteristics with groups of bacteria. This ability to define discrete groups allows for identification of new isolates and the rapid association of certain properties to them. In addition, classification of bacteria into orderly groups eliminates confusion that could be caused by the large numbers of bacterial species and the diversity among them.

3 CHAPTER ONE Review of the Literature

Use of Bacillus popilliae and Bacillus lentimorbus as biological control agents. Bacillus popilliae and B. lentimorbus are the causative agents of types A and B (respectively) milky disease, a fatal infection of Japanese beetle larvae as well as other members of the family Scarabaeidae (35). Japanese beetles and other scarabaeids feed on more than 257 different plants and cause economic loses through damage to turfgrasses and crops, making their control important to various industries (86). Biological control of these pests using B. popilliae may be easier, less expensive and ecologically safer than use of synthetic chemicals (41). The bacteria are also very specific, targeting only the insect of choice while leaving beneficial insects unharmed (72).

Bacillus popilliae has been used as a biopesticide since 1937 when Dutky artificially added diseased larvae to field plots (32). He successfully established the disease in one location and showed that B. popilliae populations built up and spread in the field. Due to the inability to produce spores in vitro, a process involving the injection of spores into healthy larvae was developed by White and Dutky in order to mass produce milky disease spores (101). A standardized spore powder was developed and used to establish B. popilliae at new field sites (33). Establishment of milky disease in the field appears to be dependent on achievement of larval densities between 180 and 480 larvae per square meter (10). Milky disease organisms may be spread in the environment by birds, insects, skunks, moles, and mice (100).

4 Other methods of producing milky disease spores have been explored, including the use of tissue culture, in vitro culture and sporulation and the use of vegetative cells as disease agents. Limited sporulation of B. popilliae has been achieved in vitro using both solid media and chemostat cultures (19, 81). Spores of B. popilliae produced in these ways are less infective than spores produced in the larvae (48). Bacillus popilliae spores germinate poorly, requiring injection of 105-107 spores into larvae to cause 50-80% infection. In contrast, injection of 102-103 viable vegetative cells causes comparable infection rates in Japanese beetle larvae (88). Splittstoesser et al. (85) reported that germination and outgrowth of B. popilliae spores in cabbage looper hemolymph reached 90% in one hour. The spores had to be heated at 37oC under alkaline conditions with the addition of tyrosine in order to achieve such rates of outgrowth (85). Sharpe et al. (82) developed a microscope slide culture system used to track the germination and outgrowth of B. popilliae B-2309 spores. They found that the vegetative cells emerged in 23-24 hours and 5% of total spores showed outgrowth after 48 hours. However, only 1% of the spores produced visible colonies on a plate, indicating that 80% of germinating spores fail to develop visible colonies. Sharpe (82) suggested that the low rate of germination and outgrowth in vitro may indicate the reason for a low infectivity rate in vivo. In tissue culture consisting of hemocytes of Phyllophaga anxia, Luthy (53) reported growth and sporulation of B. popilliae var. melolonthae and growth without sporulation of B. popilliae var. popilliae. Lyophilized vegetative cells pelleted using tung oil polymer coated with paraffin have been shown to have 93% infectivity when injected into larvae (48).

5 Milky disease bacteria have been reported to persist in the environment for extended periods of time, often longer than twenty-five years, negating the need for reapplication of spore powders to field plots (48). Resistance of the insects to B. popilliae has not been shown to occur in these areas. Investigation of the taxonomy of B. popilliae and related species will assist in the development of this element in an integrated approach to pest management.

Pathology of Bacillus popilliae and B. lentimorbus Bacillus popilliae spores are ingested by the beetle larvae during feeding, and once ingested, enter the larval midgut where the spores germinate. The vegetative cells proliferate and enter the hemocoel where they continue to multiply. Milky disease can be said to occur in four stages. An initial incubation stage (2 days) where few bacterial cells are found in the hemolymph is followed by rapid proliferation of vegetative cells (day 3 to day 5). Stage three is characterized by a change from predominantly vegetative growth to sporulation (days 5-10). Stage four is a sporulation phase terminating in the death of the larvae (day 14 to day 21) (19, 86). Infections caused by B. popilliae var. melolonthae do not follow this pattern, instead increase in vegetative cell numbers and sporulation occur simultaneously (48).

Eventually, the number of spores in the hemolymph reaches numbers as high as 5 ´ 1010 per milliliter of hemolymph. The normally clear insect hemolymph becomes turbid, leading to the name “milky disease”. The B. popilliae spores are released into the soil from the larval cadaver, thus beginning the process again. This accounts for the extended persistence of B.

6 popilliae in the environment. In larvae infected by B. lentimorbus for extended periods of time, there is a build up of blood clots, causing the hemolymph to look brownish in color instead of milky white (19).

Physiology of Bacillus popilliae and B. lentimorbus

Gordon, Haynes and Pang (39) provided phenotypic information on 12 strains of B. popilliae and 5 strains of B. lentimorbus. They reported the vegetative cells to be gram negative and the prespores and sporangia to be gram positive. Dutky originally reported the vegetative cells to be gram positive (35). When examined by electron microscopy the cells exhibit a gram positive cell wall structure (15, 16). Gordon et al. (39) reported that B. popilliae was motile by peritrichous flagella while all the strains of B. lentimorbus tested were nonmotile. Splittstoesser (85) also reported that B. popilliae cells were extremely motile upon germination and outgrowth in hemolymph slide mounts.

Bacillus popilliae and B. lentimorbus are nutritionally fastidious and only grow well on a rich medium containing yeast extract and digests of casein (19, 78). Cells reach stationary phase after 16–20 hours of growth and the maximum number of viable cells at this time is about 6 ´ 108 for B. lentimorbus and 1.2 ´ 109 for B. popilliae (86). After the cultures reach stationary phase there is a rapid decrease in viability. The cause of cell death is not fully understood, but both organisms lack the enzymes peroxidase and catalase, leaving them sensitive to hydrogen peroxide damage (67). It has been hypothesized that the lack of these enzymes may play a role in culture death (26,

7 67, 92). Pepper et al. (67) tested for oxygen evolution when hydrogen peroxide was added to a Warburg flask containing B. popilliae cells and were unable to detect any evolution of oxygen. They also tested for the breakdown of hydrogen peroxide by B. popilliae by iodometric titration and were unable to detect any breakdown of peroxide. Bacillus popilliae was examined for the presence of peroxidase and it was found that while cell extracts rapidly oxidized NADH2, the rate was not enhanced by the addition of hydrogen peroxide (67). St. Julian et al. (86) suggested that hydrogen peroxide toxicity is not the cause of death because its build up in vegetative cells is slight. They also state that death caused by exposure to the superoxide radical is unlikely because of the high levels of superoxide dismutase found in B. popilliae cells (86).

Thiamine and tryptophan have been found to be essential nutrients for B. popilliae, while biotin, myoinositol and niacin are stimulatory for growth (86, 90). Many of the amino acids must be supplied to B. popilliae and B. lentimorbus in some form, including any amino acids in the serine or aromatic families (93). Bacillus popilliae metabolizes sugars including glucose, fructose, mannose, galactose, maltose, sucrose and trehalose, the latter sugar found in the larval hemolymph (19). Products formed by glucose catabolism are lactic acid, acetic acid and carbon dioxide (68). The decrease in culture medium pH has a slight effect on the viability of the culture once it reaches stationary phase. When the culture medium was appropriately buffered, the amount of growth increased and the survival of the cells was slightly enhanced. The Embden- Meyerhof-Parnas pathway and the pentose phosphate pathways are the preferred routes of carbohydrate catabolism in B. popilliae and B. lentimorbus (68, 87). The EMP pathway is the major route

8 of glucose assimilation while the PP pathway functions mostly for formation of biosynthetic intermediates (18). Pepper et al. (68) found no evidence for the presence of either the Entner- Dudoroff or the phosphoketolase pathways in B. popilliae. Using inhibitors specific for the enzymes glyceraldehyde-3-phosphate dehydrogenase and enolase they found that they could inhibit glucose oxidation by 100%. This provided preliminary evidence for the lack of the ED and phosphoketolase pathways and further enzyme assays showed no KDPG (2-keto-3-deoxy-6-P-gluconate) aldolase or phosphoketolase activity (68). The enzymes that breakdown trehalose are expressed constitutively and both respiration and growth rates are higher when the bacteria are grown with trehalose than with glucose. Trehalose is transported into the cell by the PEP phosphotransferase system and the trehalose-6-phosphate is cleaved by a phosphotrehalase into glucose and glucose-6-phosphate (12).

B. popilliae lacks a complete tricarboxylic acid cycle, suggested by some to be the cause of the poor sporulation in vitro (18, 68). McKay et al. (56) were unable to detect a- ketoglutarate dehydrogenase activity in B. popilliae strain NRRL B-2309 and its derivatives. St. Julian et al. (86) suggested that lack of sporulation in vitro is caused by a decrease in protein synthesis and lipid metabolism once the cells reach the stationary phase of growth. B. popilliae and B. lentimorbus do contain cytochromes and are capable of oxygen dependent growth (86).

These characteristics make it difficult to grow and maintain the bacteria in the laboratory. In addition, strains such as RM9 are unable to be grown in the laboratory and can be

9 maintained only in the insects themselves. This makes it difficult to rapidly identify and classify which species of bacteria are present in natural populations in any given area (84). Strains of B. popilliae were shown to vary in virulence and for some strains the virulence and host preference could be modified by repeated passage through the insect host (19). Strains have also been noted to vary widely in their growth characteristics in vitro. Milky disease infections may exhibit some degree of insect host specificity. Bacillus popilliae var. melolonthae was isolated only from the common cockchafer (Melolontha melolontha). Two distinct B. popilliae isolates were identified in New Zealand Costelytra zealandica populations (37, 38). An atypical strain of B. popilliae has been reported to be associated with the northern masked chafer (Cyclocephala borealis) (34). The spores from the diseased larvae had an unusually large paraspore and virtually no cross-infectivity was found between spores from C. borealis and Japanese beetles (Popillia japonica) (19). Klein (48) stated that this lack of cross-infectivity stressed the need for commercial spore preparations intended for use against the Japanese beetle to be produced in Japanese beetle larvae. Milner (60) also found a lack of cross-infectivity for an isolate he called B. popilliae var. rhopaea. He showed that this isolate had virtually no ability to infect Rhopaea morbillosa and Othnonius batesi grubs in Australia but could effectively infect Rhopaea verreauxi larvae. Due to the possible lack of cross infectivity, it is necessary to be able to properly identify which species is needed to control a population of insects so that effective control of the insect is achieved. An understanding of the classification of these bacteria could lead to a means of distinguishing varieties with specific host infectivity ranges.

10 Genetics of B. popilliae

Very little is known about the genetics of B. popilliae and B. lentimorbus. Both species contain plasmids of various sizes and number. Valyasevi and Kyle (96) reported that an isolate of B. popilliae collected from infected larvae in New York contained three plasmids denoted pBP149, pBP082 and pBP043. Plasmid pPB149 showed no homology to pBP082. pBP149 was estimated to be 12 kb, pBP082 was 7.4 kb, and pBP043 was 4.9 kb in size (96). Dingman (29) performed a study on interrelated plasmids in B. popilliae strain KLN4 and B. lentimorbus strain NRRL B-2522, also finding that these isolates contained three plasmids. However, these plasmids differed from those found by Valasevi, at 6.8 kb, 8.8 kb and 9.4 kb in size. These three plasmids were named pBP68, pBP88 and pBP94, respectively ((29). All three plasmids showed contiguous regions of similarity to each other as tested by hybridization of segments of each plasmid to the others, indicating that they are an interrelated family of plasmids. A plasmid designated pBP614 has been characterized from B. popilliae and found to replicate by the rolling circle mechanism (51). This plasmid is 5.6 kb in size and the coding strand of the plasmid is deficient in cytosine (16.1% of the total base composition). Two open reading frames were found on this plasmid, one corresponding to the rep gene and the other to a protein of unknown function (51).

Bacillus popilliae and B. lentimorbus have been shown to contain N6-methyladenine in GATC sequences distinguishing them from all other Bacillus species tested for this characteristic (30). The paraspore gene (cry) has been cloned and sequenced from B. popilliae strain H1, isolated near Heidelberg, Germany. Two open reading frames of a putative operon were sequenced, the

11 first codes for a protein of 175 amino acids while the second has been designated cry18Aa1 and codes for the paraspore protein (706 amino acids) (108). The name cry18Aa1 is in accordance with the nomenclature of the Bacillus thuringiensis toxin genes as revised and summarized by (27). The first open reading frame, designated orf1, shows significant similarity to orf1 of the cry2Aa-cry2Ac operon, orf1 of the cry9Ca operon and p19 of the cry11Aa operon of B. thuringiensis. orf1 and cry18Aa1 are transcribed as an operon and EsE and EsK acting at the same site in the promoter can drive transcription of the operon (109). Cry18Aa1 has significant amino acid similarity to the Cry proteins of B. thuringiensis and hydrophobicity distribution throughout the protein seems to be similar to that found in Cry3A and Cry1A toxins of B. thuringiensis (108). Zhang et al. (108) suggested that the strong similarity between the B. popilliae cry gene and the cry genes of B. thuringiensis indicates a possible role of the paraspore protein in the pathogenesis of the milky disease organism.

Taxonomy of B. popilliae and B. lentimorbus

Isolated and described by Dutky in 1940, B. popilliae and B. lentimorbus were defined as two separate species based on the presence of a refractile parasporal body in B. popilliae and its absence in B. lentimorbus (35). In addition, there are differences in the color of the hemolymph from larvae infected by the two bacteria (18). The mol% G+C of B. popilliae is listed by Bergey’s manual as 41.3% while that for B. lentimorbus is 37.7% (23). Generally, greater than two percent difference in the mol% G+C is considered to be indicative of speciation (23). Serological differences between the two species have been demonstrated, as well as minor differences in the fatty acid

12 composition of the bacteria (47, 50, 52). Both species readily form spores in vivo, but sporulate poorly or not at all in vitro.

Spore morphology has been examined by scanning electron microscopy and it was found that the spores of B. popilliae and B. lentimorbus share a common ridged surface (20, 64, 88). The most widely used characteristic in differentiating B. popilliae from B. lentimorbus is the parasporal inclusion found in B. popilliae. This inclusion has been considered to be absent in B. lentimorbus. However, it has been suggested that the parasporal body is not a stable characteristic and should not be used for species identification (107). The parasporal body is formed at the time of sporulation as it is for other insect pathogens, such as Bacillus thuringiensis and Bacillus sphaericus. In contrast to these latter bacteria, the parasporal body in B. popilliae has not been conclusively shown to play a role in pathogenesis, although recent evidence suggests that it may have a function similar to the Cry toxins of B. thuringiensis (108). Weiner (99) found that solubilized parasporal protein was capable of killing 58 % of larvae in 48 hours when injected into the grubs. Intact parasporal inclusions were able to kill 25 % of larvae. Parasporal protein fed orally to the larvae was nontoxic (99). Zhang et al. (108) proposed a role for the B. popilliae parasporal protein in milky disease, suggesting that once the spores germinate in the larval gut the paraspore protein is activated. Once activated, the protein binds to the brush border membrane and damages the gut wall in some fashion, allowing the vegetative cells to enter the hemolymph and the disease to progress. The shape of the parasporal crystal as well as its size and position in the sporangium differs among strains of B. popilliae (59). Gordon

13 et al. (39) reported that B. popilliae was able to grow in broth supplemented with 2% NaCl while B. lentimorbus was unable to grow under this condition. This finding has been disputed by Milner (59), as all of the B. lentimorbus strains used by Gordon (39) were from the same source and the same insect. They may not have been representative of the non-paraspore forming strains. Distinct strains of B. popilliae have been isolated from several different scarabaeids. These isolates show very little cross infectivity between insect species, suggesting a fundamental difference between the isolates (49, 59).

Presently, the major criteria for establishing two species of milky disease bacteria are the presence or absence of a parasporal body in sporulated cells, the physical appearance of infected larvae, and the ability to grow in broth supplemented with 2% NaCl.

Several classification schemes have been suggested for the milky disease bacteria. After their original isolation and characterization by Dutky (35), many more strains of milky disease bacteria were isolated from different insect species. A strain causing milky disease was isolated from the European cockchafer and named B. melolonthae. A similar strain was isolated in Europe but named B. fribourgensis. These two strains were later shown to be identical, but the individual names carried on for some time. Luthy and Krywienczyc (50, 52) demonstrated that B. popilliae, B. melolonthae and B. lentimorbus shared common antigens and suggested the classification of the milky disease bacteria into two species, B. popilliae (containing three varieties, popilliae, melolonthae and lentimorbus) and B. euloomarahae, an Australian isolate that has not been grown in vitro (11).

14 Milner has described a fourth variety of B. popilliae (var. rhopeae) (61-63). This isolate produces parasporal inclusions, like var. popilliae and melolonthae, but has a larger paraspore than these varieties. This isolate and var. melolonthae will not grow in vitro at 37oC. Milner (59) has suggested classifying the milky disease bacteria on the basis of formation of the parasporal crystal during sporulation and its size and position in the sporangium. His categories include: A1 - large spore, produces parasporal body that is often small and overlaps the spore. Example: B. popilliae var. popilliae. A2 - large spore, produces parasporal body that is often large and separated from the spore. Example: RM12, the only example of this type B1 - large central spore, no paraspore. Example: B. popilliae var. lentimorbus B2 - small spore in a small sporangium, spore often eccentric, no paraspore. Example: B. popilliae var euloomarahae. This method of classification has the advantage of being purely morphological in nature, and milky disease bacteria can be identified when viewed under a microscope. This also allows the identification of isolates unable to be grown in vitro. A disadvantage to this classification scheme is the necessity for infecting larvae to produce the spores.

DNA-DNA Similarities

One method of determining phenetic relationships between bacteria is the study of deoxyribonucleic acid similarity. DNA similarity has been used to differentiate between bacteria at the species level. It has been recently proposed that DNA similarity should be used to examine relationship between

15 closely related strains, while rRNA gene sequence analysis should be used to determine more distant relationships (89). The primary structure of the rRNA gene is highly conserved, and species with more than 70% DNA similarity usually have more than 97% rRNA sequence similarity (94). DNA similarity values of 70% or more are generally considered to be indicative of identical species (46). This demonstrates that rRNA sequence analysis will not differentiate between closely related members of a species because of the high conservation of the sequences.

DNA similarity studies are based on the fact that deoxyribonucleic acid can be denatured and then renatured back into the native molecule. If competitor DNA is introduced after the denaturation of the DNA molecule, to some extent the competitor DNA will renature or hybridize with the original molecule. The amount that it renatures correlates with the amount of similarity between the sequences of the two molecules.

Similarity experiments are performed using a small amount of labeled DNA and a large amount of unlabeled competitor DNA. The labeled DNA does not reassociate appreciably with itself because it is a small amount and the strands are outcompeted by the competitor DNA in solution. Instead, the labeled DNA reassociates to the extent possible with the unlabeled DNA. The reassociated DNA is then treated with S1 nuclease to degrade any single stranded DNA left in the mixture (28). This eliminates the radioactive count from any labeled DNA that did not reassociate with another strand. Sheared native salmon sperm DNA is used as a control to determine the amount of reassociation of the labeled DNA (45). The salmon DNA is highly unrelated to the bacterial DNA and therefore will not reassociate appreciably with the labeled DNA. After S1 nuclease treatment, only the

16 rehybridized labeled DNA will be detected. This allows a determination of the amount of radioactive background caused by reassociation of labeled DNA molecules to be made (45).

RAPD Analysis

A method used to differentiate bacteria, including members of the genus Bacillus, at the strain level is the technique called randomly amplified polymorphic DNA, or RAPD (102). RAPD’s are performed using genomic DNA as a template and arbitrarily chosen PCR primers. The primers are short in length (10 base pairs) and may prime the DNA at none, one or many locations. Polymorphisms in the size of the PCR fragments result from loss or addition of a primer site through point mutations or through deletions and insertions in the chromosome between primer sites (58). This differentiates between strains because any given strain may or may not contain the same site where the primer binds or the same amount of DNA between primer sites. PCR conditions are optimized in order to facilitate the binding of an arbitrary primer (annealing temperature 36oC). The low annealing temperature allows for a certain amount of base pair mismatching between the primer and the template, thereby increasing the number of PCR fragments received from the primer. The bands created by the use of the random primers could produce a unique fingerprint when electrophoresed. This fingerprint is then compared to that of other strains, and each band is considered to be one characteristic. It can then be decided which strains share more characteristics, and their relatedness evaluated based on shared bands.

Originally used as a genetic mapping tool, RAPD analysis has been used extensively to distinguish among strains of

17 bacteria, fungi, plants and animals (7, 58, 102). RAPD strain typing has been shown to be much more sensitive than typing using multi-locus enzyme electrophoresis (MLEE). Wang et al. (98) found that by using RAPD analysis, they could distinguish 74 out of 75 isolates of Escherichia coli, compared to the identification of 15 groups of the same isolates by MLEE.

RAPD has been correlated to restriction enzyme analysis of PCR amplified small-subunit DNA coding for rRNA. This correlation illustrated that RAPD analysis is useful for providing taxonomic information at the species level (8). In a later study, Baleiras Couto et al. (7) compared the usefulness of RAPD analysis in discriminating organisms at the strain level to that of restriction enzyme analysis of the internal transcribed spacer (ITS) and nontranscribed spacer (NTS) regions of Saccharomyces cerevisiae. This study proved that RAPD primers could give rise to recognizable intraspecies patterns, thereby distinguishing between strains of S. cerevisiae isolated from spoiled beer and wine. Both RAPD analysis and restriction enzyme analysis of the ITS and NTS spacer regions of S. cerevisiae were shown to be useful in yeast identification (7). Renders et al. (75) compared RAPD analysis with pulsed field gel electrophoresis (PFGE) of Pseudomonas aeruginosa, showing that RAPD results were very comparable to those obtained from PFGE.

RAPD analysis is technically easier and more straightforward than most of these other molecular typing methods, making it a strain typing method of choice in bacterial systematics and epidemiology (40, 75). Because RAPD’s are PCR based, they require only nanogram amounts of DNA, which does not need to be highly purified or double stranded. This allows RAPD’s to be used in many situations where isolation of DNA is

18 difficult (98). It has been shown to be useful both at identifying bacterial species and bacterial strains with the use of properly selected primers. The formation of the RAPD fingerprint requires no prior genetic knowledge of the organism and is unaffected by DNA modifications such as methylation, making this technique particularly useful for taxonomic purposes (98).

Vancomycin Resistance

Stahly et al. (91) showed that certain strains of Bacillus popilliae and B. lentimorbus are resistant to the antibiotic vancomycin. Vancomycin is a glycopeptide antibiotic that was isolated from Streptomyces orientalis in 1956 (55). The molecular structure of vancomycin is based upon a linear heptapeptide molecule substituted with five aromatic rings. Vancomycin inhibits bacterial growth by halting peptidoglycan synthesis (9). The antibiotic is readily adsorbed onto the cell wall of gram positive bacteria and the UDP-N- acetylmuramylpentapeptide precursors (Chatterjee 1966). Vancomycin binds to the pentapeptide side chain at the terminal D-alanyl-D-alanine residues (70). This binding is accomplished through hydrogen bonds formed between the D-alanyl-D-alanine terminus of the precursor and the heptapeptide backbone of the antibiotic molecule (83). These bonds are strengthened by hydrophobic interactions between the peptide methyl groups and the hydrocarbons of the antibiotic (Williams 1983). Binding of vancomycin to the terminus of the pentapeptide side chain inhibits transglycosylation of the sugar backbone and transpeptidation of the pentapeptide side chain (9).

19 Vancomycin is only effective against Gram positive organisms as it is unable to cross the outer membrane of Gram negative cells (9). Vancomycin is unusual in that it never actually enters the bacterial cell, but is active at the cell surface. This means that cells are unable to use efflux mechanisms or metabolism of the antibiotic to protect themselves, relying mainly on changing the antibiotic target to become resistant.

Resistance to this antibiotic has emerged among several clinically important bacteria, including Enterococcus, Staphylococcus epidermidis, Leuconostoc and Pediococcus (Rubin) (21, 24, 80, 95). Prevalence of vancomycin resistant enterococci (VRE) in the United States has risen from 0.3% of hospital acquired infections in 1989 to 7.9% of hospital acquired infections in 1993 (25). Clonally related isolates of VRE have been obtained from different patients in the same hospital as well as in different cities (22, 65). It is thought that the increase in the number of VRE may be due to increased use of glycopeptide antibiotics as prophylactics and their use in patients sensitive to penicillin. Markopulos et al. (54) showed that glycopeptide resistance could not be developed in a step-wise fashion in enterococci, however, Staphylococcus epidermidis was able to develop increased resistance to glycopeptides due to selection pressure (54). These findings support the idea that increased use of vancomycin and related glycopeptide antibiotics has contributed to the increase in bacterial resistance.

Vancomycin resistance appears to be present in four distinguishable types; A, B, C and D. Type A resistance (VANA phenotype) is characterized by a very high minimum inhibitory

20 concentration (MIC) for vancomycin, as well as a high MIC for the related antibiotic teicoplanin (1). Type A resistance is encoded by a transposon, Tn1546, a member of the Tn3 family, and is usually found on a plasmid (4, 6, 17). Like Tn3, the transposase and resolvase genes are transcribed in opposite directions and the genes for vancomycin resistance are located downstream from the resolvase gene (6). This transposon, when placed in a host deficient in general recombination, is replicative and leads to formation of a conjugative plasmid.

The VANA operon consists of seven genes, five of which are necessary for resistance to vancomycin, and two of which are accessory genes (4). The first two genes in the operon, vanS and vanR, encode a two component regulatory system analogous to the CheY/CheA and OmpR/EnvZ systems (5). VanS shows sequence similarity to the membrane bound histidine kinase sensor proteins while VanR shows response regulator similarity (104). Arthur et al. (5) showed that expression of the downstream genes vanH, vanA and vanX were transcriptionally regulated by VanS and VanR. Wright et al. (104) proved that the cytosolic domain of VanS is phosphorylated at His194 and that phosphorylated VanS readily transferred the phosphate to VanR at Asp53. Phosphorylated VanR binds to DNA at the vanH and putative vanR promoter regions, activating transcription of vanH, vanA and vanX in response to vancomycin or related antibiotics teicoplanin and moenomycin (5, 44). Binding of phosphorylated VanR to the vanR putative promoter region represses transcription of VanR (44). VanS was shown to negatively control promoter activation by VanR in the absence of glycopeptides due to dephosphorylation of VanR by VanS (2).

21 vanH encodes a dehydrogenase which converts pyruvate to D- lactate, providing the substrate for the VanA protein (1, 13, 17). vanA codes for a ligase of altered specificity. The normal cellular ligase (ddl gene product) ligates two D-alanines to provide the D-alanyl-D-alanine precursor used in the synthesis of many bacterial cell walls (97). VanA ligates D- alanine to the D-lactate produced by VanH. When this is incorporated into the pentapeptide precursor and eventually the cell wall, it prevents binding of vancomycin to the peptidoglycan (17). The final required gene product is VanX, a d,d-dipeptidase which hydrolyzes the vancomycin sensitive precursor D-alanyl-D-alanine (106). Digestion of this molecule ensures that only resistant peptidoglycan will be manufactured by the cell (76). These five genes and protein products are required for a cell to exhibit resistance to vancomycin. The accessory proteins VanY and VanZ are also encoded by the VANA operon. VanY is a d,d-carboxypeptidase that cleaves the terminal D-lactate from side chains that have not participated in crosslinking (4, 105). VanZ confers resistance to teicoplanin, a glycopeptide antibiotic structurally related to vancomycin, in an unknown fashion (3).

VANB type resistance is characterized by a variable MIC for vancomycin and sensitivity to teicoplanin (1). The VANB operon consists of seven genes and is located on either a large conjugative chromosomal element or on a plasmid (73, 74, 103). The VANB element has been transferred naturally from enterococci to Streptococcus bovis, giving weight to the fear that vancomycin resistance will be eventually transferred to

Staphylococcus aureus (43, 71). VanRB and VanSB comprise a two component regulatory system that operates in a similar manner to that found in the VANA operon. VanRB and VanSB have a low amino

22 acid similarity to VanR and VanS, 34 and 23 % respectively.

However, VanRB and VanSB do show structural similarity to other two component regulatory system proteins. The C terminal region

of VanSB contains conserved amino acid residues characteristic of

histidine kinase sensor proteins. The N terminal domain of VanRB has conserved lysine and aspartate residues characteristic of response regulators (36). Constitutively expressed, VanRB and

VanSB together trans-activate transcription of downstream genes

vanYB, vanW and vanHB. Preexposing the cells to vancomycin can

induce resistance to teicoplanin. Activation of VanRB and VanSB

seems to be due to functional activation of VanSB by vancomycin.

The VANB operon contains five additional genes; vanHB, vanB,

vanXB, vanW and vanYB. VanHB, VanB and VanXB show very high structural and functional similarity to VanH, VanA and VanX (67,

76 and 74 % respectively) (57). VanY and VanYB share only 30 % amino acid similarity, although both proteins are d,d- carboxypeptidases (36). VanW does not show similarity to any sequence in the databases and the VANB operon does not contain a VanZ homolog, explaining the sensitivity to teicoplanin exhibited by VANB organisms (36).

Type C resistance is considered natural resistance (VANA and VANB are acquired) and is found in organisms such as Leuconostoc, Lactobacillus, and Enterococcus spp. This resistance can be either constitutive, found in Leuconostoc and Lactobacillus, or inducible (found in enterococci) (31, 79). In E. gallinarum a ligase gene responsible for vancomycin resistance was found and designated vanC-1. The protein VanC-1 shows 29 % similarity with VanA and 38 % similarity with the D- alanyl-D-alanine ligases of E. coli (31). However, as opposed to VanA which ligates D-alanine and D-lactate, VanC-1 ligates D-

23 alanine with D-serine, resulting in peptidoglycan with lowered affinity for vancomycin (14, 77).

Two organisms related to E. gallinarum, E. casseliflavus and E. flavescens were examined and shown to posses different vanC ligases designated vanC-2 and vanC-3 respectively. vanC-2 shows high nucleotide and amino acid similarity with vanC-1, 66 and 69 % respectively. vanC-3 differs from vanC-2 by 10 nucleotides, equivalent to 4 amino acid changes (66). Both E. casseliflavus and E. flavescens contain an additional ligase

gene, designated ddlE. Cass. and ddlE. flav. These gene products ligate D-alanine with D-alanine and are related to the ddl genes found in E. coli. The deduced amino acid sequences for the two genes found in E. casseliflavus and E. flavescens are identical (66). These organisms make peptidoglycan that has D-alanyl-D- lactate at the end of the pentapeptide side chain, rather than the sensitive D-alanyl-D-alanine even though the ddl genes are present. Lactobacillus and Leuconostoc have also been shown to synthesize peptidoglycan precursors that terminate in D-lactate in a constitutive manner (14, 42).

VAND has been recently described in Enterococcus faecium by Perichon et al. (69). It is characterized by constitutive, low level resistance to both vancomycin and teicoplanin. The ligase responsible for this phenotype was identified and designated vanD. The deduced amino acid sequence of this gene has 69 % similarity with VanA and VanB and 43 % similarity with VanC. This E. faecium isolate was found to synthesize peptidoglycan precursors that terminate in D-lactate (69).

24 References

1. Arthur, M., and P. Courvalin. 1993. Genetics and mechanisms of glycopeptide resistance in enterococci. Antimicrob. Agent. Chemother. 37(8):1563-1571. 2. Arthur, M., F. Depardieu, G. Gerbaud, M. Galimand, R. Leclercq, and P. Courvalin. 1997. The VanS sensor negatively controls VanR-mediated transcriptional activation of glycopeptide resistance genes of Tn1546 and related elements in the absence of induction. J. Bacteriol. 179(1):97-106. 3. Arthur, M., F. Depardieu, C. Molinas, P. Reynolds, and P. Courvalin. 1995. The vanZ gene of Tn1546 from Enterococcus faecium BM4147 confers resistance to teicoplanin. Gene. 154:87-92. 4. Arthur, M., C. Molinas, and P. Courvalin. 1992. Sequence of the vanY gene required for production of a vancomycin- inducible D,D-carboxypeptidase in Enterococcus faecium BM4147. Gene. 120:111-114. 5. Arthur, M., C. Molinas, and P. Courvalin. 1992. The VanS- VanR two-component regulatory system controls synthesis of depsipeptide peptidoglycan precursors in Enterococcus faecium BM4147. J. Bacteriol. 174(8):2582-2591. 6. Arthur, M., C. Molinas, F. Depardieu, and P. Courvalin. 1993. Characterization of Tn1546, a Tn3-related transposon conferring glycopeptide resistance by synthesis of depsipeptide peptidoglycan precursors in Enterococcus faecium BM4147. J. Bacteriol. 175(1):117-127. 7. Baleiras Couto, M. M., B. Eijsma, H. Hofstra, J. H. J. Huis in't Veld, and J. M. B. M. van der Vossen. 1996. Evaluation of molecular typing techniques to assign genetic diversity among Saccharomyces cerevisiae strains. Appl. Environ. Microbiol. 62(1):41-46. 8. Baleiras Couto, M. M., J. T. W. E. Vogels, H. Hofstra, J. H. J. Huis in't Veld, and J. M. B. M. van der Vossen. 1995. Random amplified polymorphic DNA and restriction enzyme analysis of PCR amplified rDNA in taxonomy: two identification techniques for food-borne yeasts. J. Appl. Bacteriol. 79:525-535. 9. Barna, J. C. J., and D. H. Williams. 1984. The structure and mode of action of glycopeptide antibiotics of the vancomycin group. Ann. Rev. Microbiol. 38:339-357. 10. Beard, R. L. 1945. Studies on the milky disease of Japanese beetle larvae. Conn. Agr. Exp. Sta. Bull. 491:505-583. 11. Beard, R. L. 1956. Two milky diseases of Australian Scarabaeidae. Can. Entomol. 138:640-647.

25 12. Bhumiratana, A., R. L. Anderson, and R. N. Costilow. 1974. Trehalose metabolism by Bacillus popilliae. J. Bacteriol. 119(2):484-493. 13. Billot-Klein, D., L. Gutmann, E. Collatz, and J. van Heijenoort. 1992. Analysis of peptidoglycan precursors in vancomycin-resistant enterococci. Antimicrob. Agent. Chemother. 36(7):1487-1490. 14. Billot-Klein, D., L. Gutmann, S. Sable, E. Guittet, and J. van Heijenoort. 1994. Modification of peptidoglycan precursors is a common feature of the low-level vancomycin- resistant species Lactobacillus casei, Pediococcus pentosaceus, Leuconostoc mesenteroides, and Enterococcus gallinarum. J. Bacteriol. 176(8):2398-2405. 15. Black, S. H. 1968. Cytology of milky disease bacteria. I. Morphogenesis of Bacillus popilliae in vivo. J. Invertebr. Pathol. 12:148-157. 16. Black, S. H. 1968. Cytology of milky disease bacteria. II. Morphogenesis of Bacillus popilliae in vitro. J. Invertebr. Pathol. 12:158-167. 17. Bugg, T. D. H., G. D. Wright, S. Dutka-Malen, M. Arthur, P. Courvalin, and C. T. Walsh. 1991. Molecular basis for vancomycin resistance in Enterococcus faecium BM4147: biosynthesis of a depsipeptide peptidoglycan precursor by vancomycin resistance proteins VanH and VanA. Biochem. 30:10408-10415. 18. Bulla, L. A., Jr., G. A. Bennett, and O. L. Shotwell. 1970. Physiology of sporeforming bacteria associated with insects. II. Lipids of vegetative cells. J. Bacteriol. 104(3):1246-1253. 19. Bulla, L. A., Jr., R. N. Costilow, and E. S. Sharpe. 1978. Biology of Bacillus popilliae. Adv. Appl. Microbiol. 2:1- 18. 20. Bulla, L. A., G. St. Julian, R. A. Rhodes, and C. W. Hesseltine. 1969. Scanning electron and phase contrast microscopy of bacterial spores. Appl. Microbiol. 18(3):490- 495. 21. Buu-Hoi, A., C. Branger, and J. F. Acar. 1985. Vancomycin- resistant streptococci or Leuconostoc sp. Antimicrob. Agent. Chemother. 28(3):458-460. 22. Chow, J. W., A. Kuritza, D. M. Shlaes, M. Green, D. F. Sahm, and M. J. Zervos. 1993. Clonal spread of vancomycin- resistant Enterococcus faecium between patients in three hospitals in two states. J. Clin. Microbiol. 31(6):1609- 1611. 23. Claus, D., and R. C. W. Berkeley. 1986. Genus Bacillus Cohn 1872, p. 1105-1207. In P. H. A. Sneath, N. S. Mair, M. E.

26 Sharpe, and J. G. Holt (ed.), Bergey's Manual of Systematic Bacteriology, vol. 2. Williams and Wilkins, Baltimore. 24. Colman, G., and A. Efstratiou. 1987. Vancomycin-resistant leuconostocs, lactobacilli and now pediococci. J. Hosp. Infect. 10:1-3. 25. Committee, H. I. C. P. A. 1995. Recommendations for preventing the spread of vancomycin resistance Vol. 16 No. 2. The Centers for Disease Control and Prevention. 26. Costilow, R. N., C. J. Sylvester, and R. E. Pepper. 1966. Production and stabilization of cells of Bacillus popilliae and Bacillus lentimorbus. Appl. Microbiol. 14(2):161-169. 27. Crosa, J. H., D. J. Brenner, and S. Falkow. 1973. Use of a single-strand specific nuclease for analysis of bacterial and plasmid deoxyribonucleic acid homo- and heteroduplexes. J. Bacteriol. 115(3):904-911. 28. Dingman, D. W. 1994. Physical properties of three plasmids and the presence of interrelated plasmids in Bacillus popilliae and Bacillus lentimorbus. J. Invertebr. Pathol. 63:235-243. 29. Dingman, D. W. 1990. Presence of N6-methyladenine in GATC sequences of Bacillus popilliae and Bacillus lentimorbus KLN2. J. Bacteriol. 172(10):6156-6159. 30. Dutka-Malen, S., C. Molinas, M. Arthur, and P. Courvalin. 1992. Sequence of the vanC gene of Enterococcus gallinarum BM4174 encoding a D-alanine:D-alanine ligase-related protein necessary for vancomycin resistance. Gene. 112:53- 58. 31. Dutky, S. R. 1937. Investigation of disease of the immature stages of the Japanese beetle. PhD thesis. Rutgers University, New Brunswick, NJ. 32. Dutky, S. R. 1942. Method for the preparation of spore-dust mixtures of type A milky disease of Japanese beetle larvae for field inoculation ET-192. United States Department of Agriculture Bureau of Entomology and Plant Quarantine. 33. Dutky, S. R. 1963. The milky diseases, p. 75-115. In E. A. Steinhaus (ed.), Insect Pathology, an Advanced Treatise, vol. 2. Academic Press, New York. 34. Dutky, S. R. 1940. Two new spore-forming bacteria causing milky diseases of Japanese beetle larvae. J. Agri. Res. 61(1):57-68. 35. Evers, S., and P. Courvalin. 1996. Regulation of VanB-type vancomycin resistance gene expression by the VanSB-VanRB two-component regulatory system in Enterococcus faecalis V583. J. Bacteriol. 178(5):1302-1309. 36. Farrell, J. A. K. 1972. Observations on soil-inhabiting insect populations of improved pasture in Nelson Province,

27 with particular reference to Costelytra zealandica (White) (Col: Scarabaeidae). New Zea. J. Agri. Res. 15:878-892. 37. Fowler, M. 1972. A new milky disease organism from New Zealand. J. Invertebr. Pathol. 19:409-410. 38. Gordon, R. E., W. C. Haynes, and C. H.-P. Pang. 1973. The Genus Bacillus, vol. 427. U. S. Department of Agriculture, Washington, D. C. 39. Gori, A., F. Espinasse, A. Deplano, C. Nonhoff, M. H. Nicolas, and M. J. Struelens. 1996. Comparison of pulsed- field gel electrophoresis and randomly amplified DNA polymorphism analysis for typing extended-spectrum-b- lactamase-producing Klebsiella pneumoniae. J. Clin. Microbiol. 34(10):2448-2453. 40. Hall, I. M. 1964. Use of micro-organisms in biological control, p. 610-628. In D. Bach (ed.), Biological control of insect pests and weeds. Reinhold Publishing Corporation. 41. Handwerger, S., M. J. Pucci, K. J. Volk, J. Liu, and M. S. Lee. 1994. Vancomycin-resistant Leuconostoc mesenteroides and Lactobacillus casei synthesize cytoplasmic peptidoglycan precursors that terminate in lactate. J. Bacteriol. 176(1):260-264. 42. Hayden, M. K., R. N. Picken, and D. F. Sahm. 1997. Heterogeneous expression of glycopeptide resistance in enterococci associated with transfer of vanB. Antimicrob. Agent. Chemother. 41(4):872-874. 43. Holman, T. R., Z. Wu, B. L. Wanner, and C. T. Walsh. 1994. Identification of the DNA-binding site for the phosphorylated VanR protein required for vancomycin resistance in Enterococcus faecium. Biochem. 33:4625-4631. 44. Johnson, J. L. 1994. Similarity analysis of DNA's, p. 655- 682. In P. Gerhardt, R. G. E. Murray, W. A. Wood, and N. R. Krieg (ed.), Methods for General and Molecular Bacteriology, 1st ed. American Society for Microbiology, Washington, D. C. 45. Johnson, J. L. 1973. Use of nucleic-acid homologies in the taxonomy of anaerobic bacteria. Int. J. Syst. Bacteriol. 23:308-315. 46. Kaneda, T. 1969. Fatty acids in Bacillus larvae, Bacillus lentimorbus and Bacillus popilliae. J. Bacteriol. 98(1):143-146. 47. Klein, M. G. 1981. Advances in the Use of Bacillus popilliae for Pest Control, p. 183-192. In H. D. Burges (ed.), Microbial Control of Pests and Plant Diseases 1970- 1980, 1st ed. Academic Press, London. 48. Klein, M. G. 1992. Use of Bacillus popilliae in Japanese beetle control, p. 179-189. In T. Jackson and T. Glare

28 (ed.), Use of Pathogens in Scarab Pest Management. Intercept Limited, Andover. 49. Krywienczyk, J., and P. Luthy. 1974. Serological relationship between three varieties of Bacillus popilliae. J. Invertebr. Pathol. 23:275-279. 50. Longley, M., R. MacDonald, and R. T. M. Poulter. 1997. Characterization of pBP614, a putative rolling circle plasmid from Bacillus popilliae. Plasmid. 37:15-21. 51. Luthy, P., and J. Krywienczyk. 1972. Serological comp. 19(2):163-165. 52. Luthy, P., C. Wyss, and L. Ettlinger. 1970. Behavior of milky disease organisms in a tissue culture system. J. Invertebr. Pathol. 16(3):325-330. 53. Markopulos, E., W. Graninger, and A. Georgopoulos. 1998. In-vitro selection of resistance to vancomycin and teicoplanin in Enterococcus faecium and Enterococcus faecalis compared with Staphylococcus epidermidis. J. Antimicrob. Chemother. 41:43-47. 54. McCormick, M. H., W. M. Stark, G. E. Pittenger, R. C. Pittenger, and J. M. McGuire. 1956. Vancomycin, a new antibiotic. I. Chemical and biological properties, p. 606- 611, Antibiotics Annual 1955-1956. 55. McKay, L. L., A. Bhumiratana, and R. N. Costilow. 1971. Oxidation of acetate by various strains of Bacillus popilliae. Appl. Microbiol. 22(6):1070-1075. 56. Meziane-Cherif, D., M.-A. Badet-Denisot, S. Evers, P. Courvalin, and B. Badet. 1994. Purification and characterization of the VanB ligase associated with type B vancomycin resistance in Enterococcus faecalis V583. FEBS Lett. 354:140-142. 57. Michelmore, R. W., I. Paran, and R. V. Kesseli. 1991. Identification of markers linked to disease-resistance genes by bulked segregant analysis: A rapid method to detect markers in specific genomic regions by using segregating populations. Proc. Nat. Acad. Sci., USA. 88:9828-9832. 58. Milner, R. J. 1981. Identification of the Bacillus popilliae group of Insect Pathogens, p. 45-59. In H. D. Burges (ed.), Microbial Control of Pests and Plant Diseases 1970-1980, 1 ed. Academic Press, London. 59. Milner, R. J. 1976. A laboratory evaluation of the pathogenicity of Bacillus popilliae var. rhopaea, the agent of milky disease in Rhopeae verreauxi (Coleoptera: Scarabaeidae). J. Invertebr. Pathol. 28:185-190. 60. Milner, R. J. 1974. A new variety of milky disease, Bacillus popilliae var. rhopaea, from Rhopaea verreauxi. Aust. J. Bio. Sci. 27:235-247.

29 61. Milner, R. J. 1981. A novel milky disease organism from Australian Scarabaeids: Field occurrence, isolation, and infectivity. J. Invertebr. Pathol. 37:304-309. 62. Milner, R. J., and C. D. Beaton. 1981. A novel milky disease organism from Australian Scarabaeids: Ultrastructure. J. Invertebr. Pathol. 37:310-318. 63. Mitruka, B. M., R. N. Costilow, S. H. Black, and R. E. Pepper. 1967. Comparisons of cells, refractile bodies and spores of Bacillus popilliae. J. Bacteriol. 94(3):759-765. 64. Murray, B. E. 1995. What can we do about vancomycin- resistant enterococci. Clin. Infect. Dis. 20:1134-1136. 65. Navarro, F., and P. Courvalin. 1994. Analysis of genes encoding D-alanine-D-alanine ligase-related enzymes in Enterococcus casseliflavus and Enterococcus flavescens. Antimicrob. Agent. Chemother. 38(8):1788-1793. 66. Pepper, R. E., and R. N. Costilow. 1965. Electron transport in Bacillus popilliae. J. Bacteriol. 89(2):271-276. 67. Pepper, R. E., and R. N. Costilow. 1964. Glucose catabolism by Bacillus popilliae and Bacillus lentimorbus. J. Bacteriol. 87(2):303-310. 68. Perichon, B., P. Reynolds, and P. Courvalin. 1997. VanD- type glycopeptide-resistant Enterococcus faecium BM4339. Antimicrob. Agent. Chemother. 41(9):2016-2018. 69. Perkins, H. R. 1969. Specificity of combination between mucopeptide precursors and vancomycin or ristocetin. Biochem. J. 111:195-205. 70. Poyart, C., C. Pierre, G. Quesne, B. Pron, P. Berche, and P. Trieu-Cuot. 1997. Emergence of vancomycin resistance in the genus Streptococcus: Characterization of a vanB transferable determinant in Streptococcus bovis. Antimicrob. Agent. Chemother. 41(1):24-29. 71. Priest, F. G., D. A. Kaji, and M. Aquino de Muro. 1994. Systematics of insect pathogens: Uses in strain identification and isolation of novel pathogens, p. 275- 296. In F. G. Priest, A. Ramos-Cormenzana, and B. Tindall (ed.), Bacterial Diversity and Systematics. Plenum Press, New York. 72. Quintiliani, R., Jr., and P. Courvalin. 1994. Conjugal transfer of the vancomycin resistance determinant vanB between enterococci involves the movement of large genetic elements from chromosome to chromosome. FEMS Microbiol. Lett. 119:359-364. 73. Quintiliani, R., Jr., S. Evers, and P. Courvalin. 1993. The vanB gene confers various levels of self-transferable resistance to vancomycin in enterococci. J. Infect. Dis. 167:1220-1223.

30 74. Renders, N., U. Romling, H. Verbrugh, and A. van Belkum. 1996. Comparative typing of Pseudomonas aeruginosa by random amplification of polymorphic DNA or pulsed-field gel electrophoresis of DNA macrorestriction fragments. J. Clin. Microbiol. 34(12):3190-3195. 75. Reynolds, P. E., F. Depardieu, S. Dutka-Malen, M. Arthur, and P. Courvalin. 1994. Glycopeptide resistance mediated by enterococcal transposon Tn1546 requires production of VanX for hydrolysis of D-alanyl-D-alanine. Molec. Microbiol. 13(6):1065-1070. 76. Reynolds, P. E., H. A. Snaith, A. J. Maguire, S. Dutka- Malen, and P. Courvalin. 1994. Analysis of peptidoglycan precursors in vancomycin-resistant Enterococcus gallinarum BM4174. Biochem. J. 301:5-8. 77. Rhodes, R. A. 1965. Symposium on microbial insecticides. II. Milky disease of the Japanese beetle. Bacteriol. Rev. 29(3):373-381. 78. Sahm, D. F., L. Free, and S. Handwerger. 1995. Inducible and constitutive expression of vanC-1-encoded resistance to vancomycin in Enterococcus gallinarum. Antimicrob. Agent. Chemother. 39(7):1480-1484. 79. Schwalbe, R. S., J. T. Stapleton, and P. H. Gilligan. 1987. Emergence of vancomycin resistance in coagulase-negative staphylococci. New Engl. J. Med. 316(15):927-931. 80. Sharpe, E. S., and L. A. Bulla, Jr. 1978. Characteristics of the constituent substrains of Bacillus popilliae growing in batch and continuous cultures. Appl. Environ. Microbiol. 35(3):601-609. 81. Sharpe, E. S., and L. A. Bulla, Jr. 1977. Germination and outgrowth of Bacillus popilliae spores in microscope slide culture. J. Invertebr. Pathol. 30:242-248. 82. Sheldrick, G. M., P. G. Jones, O. Kennard, D. H. Williams, and G. A. Smith. 1978. Structure of vancomycin and its complex with acetyl-D-alanyl-D-alanine. Nature. 271:223- 225. 83. Splittstoesser, C. M., and C. Y. Kawanishi. 1981. Insect diseases caused by bacilli without toxin mediated pathologies, p. 190-199. In E. W. Davidson (ed.), Pathogenesis of Invertebrate Microbial Diseases. Allanheld, Osmun and Company. 84. Splittstoesser, C. M., C. Y. Kawanishi, and H. Tashiro. 1975. Germination and outgrowth of Bacillus popilliae in hemolymph slide mounts. J. Invertebr. Pathol. 25:371-374. 85. St. Julian, G., and L. A. Bulla, Jr. 1973. Milky disease, p. 57-87. In T. C. Cheng (ed.), Current Topics in Comparative Pathobiology. Academic Press, New York.

31 86. St. Julian, G., L. A. Bulla, Jr., and R. S. Hanson. 1975. Physiology of sporeforming bacteria associated with insects: Metabolism of Bacillus popilliae grown in third- instar Popillia japonica Newman larvae. Appl. Microbiol. 30(1):20-25. 87. St. Julian, G., T. G. Pridham, and H. H. Hall. 1967. Preparation and characterization of intact and free spores of Bacillus popilliae Dutky. Can. J. Microbiol. 13:279-285. 88. Stackebrandt, E., and B. M. Goebel. 1994. Taxonomic Note: A Place for DNA-DNA Reassociation and 16S rRNA Sequence Analysis in the Present Species Definition in Bacteriology. Int. J. Syst. Bacteriol. 44(4):846-849. 89. Stahly, D. P., R. E. Andrews, and A. A. Yousten. 1992. The genus Bacillus-Insect pathogens, p. 1029-2140. In A. Balows, H. G. Truper, M. Dworkin, W. Harder, and K.-H. Schleifer (ed.), The Prokaryotes, 2 ed, vol. 2. Springer- Verlag, New York. 90. Stahly, D. P., D. M. Takefman, C. A. Livasy, and D. W. Dingman. 1992. Selective medium for quantitation of Bacillus popilliae in soil and in commercial spore powders. Appl. Environ. Microbiol. 58(2):740-743. 91. Steinkraus, K. H. 1957. Studies on the milky disease organisms. II Saprophytic growth of Bacillus popilliae. J. Bacteriol. 74:625-632. 92. Sylvester, C. J., and R. N. Costilow. 1964. Nutritional requirements of Bacillus popilliae. J. Bacteriol. 87(1):114-119. 93. Ursing, J. B., R. A. Rossello-Mora, E. Garcia-Valdes, and J. Lalucat. 1995. Taxonomic note: A pragmatic approach to the nomenclature of phenotypically similar genomic groups. Int. J. Syst. Bacteriol. 45(3):604. 94. Uttley, A. H. C., C. H. Collins, J. Naidoo, and R. C. George. 1988. Vancomycin-resistant enterococci. Lancet. i:57-58. 95. Valyasevi, R., M. M. Kyle, P. J. Christie, and K. H. Steinkraus. 1990. Plasmids in Bacillus popilliae. J. Invertebr. Pathol. 56:286-288. 96. Walsh, C. T. 1989. Enzymes in the D-alanine branch of bacterial cell wall peptidoglycan assembly. J. Biol. Chem. 264(5):2393-2396. 97. Wang, G., T. S. Whittam, C. M. Berg, and D. E. Berg. 1993. RAPD (arbitrary primer) PCR is more sensitive than multilocus enzyme electrophoresis for distinguishing related bacterial strains. Nucleic Acids Res. 21(25):5930- 5933.

32 98. Weiner, B. A. 1978. Isolation and partial characterization of the parasporal body of Bacillus popilliae. Can. J. Microbiol. 24:1557-1561. 99. Wheeler, E. H. 1943. Effect of the milky disease on 1942-43 Japanese beetle grubs. Farm Res. 9:8. 100. White, R. T., and S. R. Dutky. 1940. Effect of the introduction of milky diseases on populations of Japanese beetle larvae. J. Econ. Entomol. 33(2):306-309. 101. Woodburn, M. A., A. A. Yousten, and K. H. Hilu. 1995. Random amplified polymorphic DNA fingerprinting of mosquito pathogenic and nonpathogenic strains of Bacillus sphaericus. Int. J. Syst. Bacteriol. 45(2):212-217. 102. Woodford, N., B. L. Jones, Z. Baccus, H. A. Ludlam, and D. F. J. Brown. 1995. Linkage of vancomycin and high-level gentamicin resistance genes on the same plasmid in a clinical isolate of Enterococcus faecalis. J. Antimicrob. Chemother. 35:179-184. 103. Wright, G. D., T. R. Holman, and C. T. Walsh. 1993. Purification and characterization of VanR and the cytosolic domain of VanS: A two-component regulatory system required for vancomycin resistance in Enterococcus faecium BM4147. Biochem. 32:5057-5063. 104. Wright, G. D., C. Molinas, M. Arthur, P. Courvalin, and C. T. Walsh. 1992. Characterization of VanY, a DD- carboxypeptidase from vancomycin-resistant Enterococcus faecium BM4147. Antimicrob. Agent. Chemother. 36(7):1514- 1518. 105. Wu, Z., G. D. Wright, and C. T. Walsh. 1995. Overexpression, purification, and characterization of VanX, a D-, D-dipeptidase which is essential for vancomycin resistance in Enterococcus faecium BM4147. Biochem. 34:2455-2463. 106. Zang, S., Y. Wan, and B. Wang. 1988. Study on the morphological change of milky bacteria in grubs. Acta. Agric. Boreali-Sinica. 3:52-57. 107. Zhang, J., T. C. Hodgman, L. Krieger, W. Schnetter, and H. U. Schairer. 1997. Cloning and analysis of the first cry gene from Bacillus popilliae. J. Bacteriol. 179(13):4336- 4341. 108. Zhang, J., H. U. Schairer, W. Schnetter, D. Lereclus, and H. Agaisse. 1998. Bacillus popilliae cry18Aa operon is transcribed by sE and sK forms of RNA polymerase from a single initiation site. Nucleic Acids Res. 26(5):1288-1293.

33 CHAPTER TWO

Materials and Methods Media and Reagents MYPGP broth 1.5% yeast extract, 1.0% Mueller-

Hinton broth, 0.3% K2HPO4, 0.1% sodium pyruvate, 0.2% glucose MYPGP agar MYPGP broth plus 2.0% agar Cell Suspension buffer 10 mM Tris-HCl (pH 8.0), 1 mM disodium EDTA, 0.35 M sucrose 2X Lysing buffer 100 mM Tris-HCl (pH 8.0), 20 mM disodium EDTA, 0.3 M NaCl, 2% (w/v) SDS, 2% (v/v) b-mercaptoethanol, 100 mg/ml proteinaseK RNase 1 mg/ml RNase A dissolved in 0.15 M NaCl

(pH 5.0), 4,000 U/ml T1 RNase TE buffer 10 mM Tris-HCL (pH 8.0), 1 mM EDTA

Iodination buffer 7.2 M NaClO4, 0.02 mM KI in 80 mM glacial acetic acid (pH 4.8)

TlCl3 catalyst 1.0 mg/ml TlCl3 dissolved in 100 mM acetic acid (pH 4.8)

Sodium phosphate buffer 0.5 M NaH2PO4.H2O, 0.5 M Na2HPO4 (pH 6.8) Stop reaction buffer 0.5 M sodium phosphate buffer (pH 7.0) HA buffer 0.14 M sodium phosphate buffer, 0.5% SDS Tris buffer 1.0 M Tris-HCl (pH 8.0) TE-SDS buffer TE buffer plus 0.1% SDS Salmon sperm DNA salmon sperm DNA dissolved in TE, then sheared to 400-600bp sodium acetate buffer 3.0 M sodium acetate

S1 Nuclease buffer 0.3 M NaCl, 0.05 M acetic acid, 0.5mM

ZnCl2, pH 4.6

HCl buffer 1 M HCl, 1% Na4P2O7.10H2O, 1% NaH2PO4.H2O Acid wash buffer 1:5 dilution of HCl buffer

S1 storage buffer 20 mM Tris, pH 7.5, 50 mM NaCl, 0.1 mM

ZnCl2, 50% glycerol High Salt buffer 13.2X SSC, 5 mM HEPES, pH 7.0 50X TAE 2.0 M Tris base, 57.1 ml/L glacial acetic acid, 100 ml/L 0.5 M EDTA (pH 8.0)

34 Gel loading dye 30% sucrose in TE with bromophenol blue 10X TNE 100 mM Tris-HCl, 10 mM EDTA, 2.0 M NaCl, pH 7.4 10X buffer (Promega) 500 mM KCl, 100mM Tris-HCl (pH 9.0), 1% Triton X-100 dNTP mixture 2.5 mM each dATP, dCTP, dTTP, dGTP

MgCl2 25 mM MgCl2 Depurinating solution 250 mM HCl Denaturing solution 0.5 M NaOH, 1.5 M NaCl Neutralization solution 1.0 M Tris-HCl (pH 8.0), 1.5 M NaCl 20X SSC 3.0 M NaCl, 300 mM sodium citrate (pH 7.5) Prehybridization buffer 5X SSC, 1% (w/v) Blocking reagent (Boehringer Mannheim), 0.1% N- lauroylsarcosine, 0.2% SDS 2X wash 2X SSC, 0.1% SDS 0.5X wash 0.5X SSC, 0.1% SDS 10X Maleic acid buffer 100 mM maleic acid, 150 mM NaCl, pH 7.5 Blocking solution 1% Blocking reagent (Boehringer Mannheim) dissolved in 1X maleic acid buffer Detection buffer 100 mM Tris-HCl, 100 mM NaCl, pH 9.5 Color developing solution 45 ml 4-nitroblue tetrazolium chloride (NBT, Boehringer Mannheim) 35ml 5-bromo-4-chloro-3-indoyl-phosphate (X- phosphate, Boehringer Mannheim) dissolved in 10 ml detection buffer

Bacterial strains and growth conditions Bacterial strains used in this study are listed in Tables 1 and 2. All strains were grown in MYPGP broth or on MYPGP plates (1). Bacteria used to inoculate flasks for DNA isolation were grown overnight in 5 ml of MYPGP broth with shaking at 30oC. Two, two-liter erlenmeyer flasks containing 500 ml of MYPGP broth each were inoculated with 5 ml of culture and incubated for approximately 16 h in a New Brunswick G25 shaker at 30oC with shaking (175 rpm). One- liter erlenmeyer flasks containing 250 ml of media were inoculated

35 for RAPD DNA isolation. Cells were harvested by centrifugation (12,000 x g for 15 min) and the cell pellet was stored at -20oC.

Phenotypic testing was performed on MYPGP plates containing either 150 mg/ml vancomycin (Sigma) or 2% NaCl. The plates were streaked from an MYPGP plate grown overnight and then incubated for 1 to 2 days at 30oC. Growth was determined by visual examination of the plates. Bacterial tolerance of 2% NaCl was also tested using MYPGP broth supplemented with 2% NaCl. Klett tubes containing 5 ml media were inoculated and incubated at 30oC on a New Brunswick model TC-5 roller drum shaker (23 rpm). Growth was determined as greater than a doubling in absorbance.

Table 1. B. popilliae and B. lentimorbus strains used in DNA-DNA reassociation

Refer to Rippere et al. 1998. Molecular systematics of Bacillus popilliae and Bacillus lentimorbus, bacteria causing milky disease in Japanese beetles and related scarab larvae. Int. J. Syst. Bacteriol. 48:395- 402.

Table 2. B. popilliae and B. lentimorbus strains from diverse host insects and geographical regions used in RAPD analysis

Strain Host Insect Source ATCC 14706+ Popillia japonica USA1 ATCC 14707* Popillia japonica USA1 BlPj1+ Popillia japonica USA6 Bp1* Papuana woodlarkiana Papua New Guinea2 Bp6+ Popillia japonica USA2 Bp9+ Ataenius spretulatus USA, NY2 Bp10+ Anomala flavipennis USA, NC2 Bp11* USA2 Bp12+ Holotichia oblita China2 Bp13+ Popillia japonica Australia2 Bp14+ Cyclocephala hirta USA, CA2

36 Bp15* Cyclocephala lurida USA, TX2 Bp16* Polyphyla comes USA, NC2 Bp17* Phyllophaga crinita USA, TX2 Bp18* Anomala diversa Japan2 Bp19* Rhopaea morbillosa Australia2 Bp21* USA, TN2 Bp22+ Phyllophaga sp. Panama2 Bp23++ Popillia japonica USA2 Bp25* Cyclocephala hirta USA, NY2 Bp26* Cyclocephala parallela USA, FL2 BpF+ Europe5 BpCb1* Cyclocephala borealis USA6 BpPa1* Phyllophaga anxia USA6 BpPj1+ Popillia japonica USA6 DNG 2+ Popillia japonica USA6 DNG 11+ Anomala orientalis USA6 DNG 4+ Anomala orientalis USA6 KLN1+ Popillia japonica USA6 KLN3+ Popillia japonica USA6 NRRL B-2524+ Popillia japonica USA4 NRRL B-4081+ Melolontha melolonthae Europe4 NRRL B-4145+ USA4 NRRL B-4154+ Odontria (strain Odontria) USA4 RM23+ Anoplognathus porosus Australia3 RM29+ Lepidiota picticollis Australia3 1ATCC, 2Klein, 3Milner, 4Nakamura 5Schnetter, 6Stahley *B. lentimorbus +B. popilliae ++B. popilliae Dutky

Stock cultures were made by adding 900 ml of an overnight culture grown in MYPGP broth to 100 ml sterile glycerol to make a final glycerol concentration of 10%. The cultures were mixed and stored at -80oC.

Isolation of bacteria from dried beetle hemolymph

37 Hemolymph samples were received as a dried film coating a microscope slide. Ten microliters of sterile distilled water were added to a spot on the slide to allow resuspension of the dried spores. The water was lifted off of the slide using an Eppendorf pipettor and added to 90 ml sterile distilled water. The spore suspension was mixed and incubated in a 60oC waterbath for 20 min. A dilution series was performed and the 10-6 and 10-7 dilutions were plated on MYPGP agar. The plates were incubated at 30oC for approximately one week, during which time any possible B. popilliae colonies were restreaked on MYPGP agar. These cultures were checked for purity by restreaking, microscopic examination and the catalase test (B. popilliae and B. lentimorbus are catalase negative).

DNA isolation for DNA-DNA reassociation The cell pellet was taken out of the freezer and fully thawed. DNA was isolated following a variation of the Marmur procedure (2). Five milliliters of cell suspension buffer were added to the pellet, and the pellet was resuspended using a sterile 5 ml glass pipet. Fifteen ml suspension buffer were added to the cells along with 1 mg/ml lysozyme. The suspension was transferred to a 125 ml glass stoppered erlenmeyer flask and incubated at 37oC for 3 h, followed by the addition of 20 ml 2X lysing solution and 10 ml 5 M sodium perchlorate. Following incubation at 55oC for 2 h, 15 ml of phenol:chloroform:isoamyl alcohol (25:24:1) were added to the cells, which were briefly shaken vigorously by hand to homogenize the mixture, followed by vigorous shaking for 20 min on a platform shaker. The mixture was centrifuged at 17,000 x g for 10 min to separate the aqueous layer from the phenol layer. The aqueous layer was removed from the centrifuge tube with an inverted 5 ml glass pipet and placed in the erlenmeyer flask. The phenol:chloroform extractions were repeated until the aqueous layer was clear. After the final phenol:chloroform extraction, the aqueous layer was transferred to a clean erlenmeyer flask and 0.6 volume isopropanol was added to precipitate the nucleic acids. The nucleic acids were clotted by gentle swirling of the flask, and the clot held back with a sterile 5 ml glass pipet while the alcohol was poured off. The nucleic acids were washed with cold 80% ethanol for 15 min with

38 occasional swirling. The ethanol was poured off in the same manner as the isopropanol, and the nucleic acids were allowed to air dry. Once dry, the DNA was resuspended in sterile TE buffer and refrigerated at 4oC overnight. The next morning, 250 ml RNase mix were added to the nucleic acids and incubated at 37oC for 1 h to degrade any RNA present. Five milliliters chloroform:isoamyl alcohol (24:1) were added to the DNA, shaken vigorously to homogenize the mixture, and then shaken for 20 min. The DNA was centrifuged at 17,000 x g for 10 min. The aqueous layer was removed and placed in a sterile 100 ml beaker, to which was added 0.1 volume of 3 M sodium acetate (2 ml). The DNA was precipitated by the addition of 2 volumes 95% ethanol. The precipitated DNA was spooled on a glass rod, washed with cold 80% ethanol and allowed to air dry. The dry DNA was resuspended in 3 ml TE buffer, quantified at 260 nm and stored at -20oC.

DNA sample preparation The samples to be used for DNA-DNA reassociation experiments were diluted to a concentration of 0.4 mg/ml in a final volume of 4-5 ml. The samples were passed through a French Pressure Cell (American Instrument Co.) at 16,000 lb/in2 and fragment sizes were determined by electrophoresis on a 0.7% agarose gel. Any sample that had fragment sizes larger than 800 bp was passed through the pressure cell again. After shearing, the DNA samples were heated in a boiling water bath for 5 min, cooled rapidly on ice for 5 min, and centrifuged at 17,000 x g for 10 min at 4oC (2). The samples were stored at -20oC.

DNA labeling Five micrograms (12.5 ml) of the DNA to be labeled were placed in a glass autoinjection vial (Chemical Research Suppliers) and 0.1 volume of 3.0 M sodium acetate (pH 6.0) was added. The samples were mixed well, followed by the addition of 2.0 volumes of cold 95% ethanol. The samples were again mixed well and incubated at -20oC for 1 h, followed by centrifugation at 12,000 x g for 15 min. The supernatant was decanted, cold 80% ethanol added to desalt the pellet, and centrifuged again for 15 min. The supernatant was decanted and the pellet dried at 37oC. The vials were covered with

39 parafilm and stored at -20oC until the labeling reaction could be performed.

Fifteen minutes prior to the beginning of the labeling reaction, 23 ml of reaction buffer were added to each dried sample. Once the samples were resuspended, 1.0 ml (100 mCi) of sodium iodide (125I, Dupont New England Nuclear) was placed on the side of the vial, followed by 6.0 ml of TlCl3 catalyst on the opposite side of the vial. 125I was used because it can be chemically linked to cytosine residues in the presence of thallium chloride, thereby eliminating the need to grow the bacteria with a radioactive isotope. A serum bottle cap was crimped onto each reaction vial, the contents mixed and incubated in a 70oC waterbath for 20 min. While the samples were incubating, NAP- 25 sepharose (Pharmacia) columns were equilibrated by washing three times with HA buffer, and a tuberculin syringe was loaded with stop reaction buffer. The reaction tubes were removed from the waterbath, allowed to cool for 2 min and 0.1 ml of stop reaction buffer was injected into each vial. The contents of the vials were mixed and incubated in a 70oC waterbath for 20 min. During this incubation period, hydroxyapatite was added to Pasteur pipets plugged with glass wool and kept moist by plugging the bottom of the pipet. The columns were placed in a glass culture tube in the 70oC waterbath. For each DNA sample, one tuberculin syringe was loaded with 0.15 ml HA buffer and 50 ml salmon sperm DNA (denatured, 0.4 mg/ml) were added to a screw cap tube. The NAP-25 columns were placed in the fume hood and allowed to drain and air dry.

The vials were removed from the waterbath and cooled for 2 min. Using the prepared syringes, HA buffer (0.15 ml) was injected into the bottom of each vial, and the contents were drawn back up into the syringe. The reactions were loaded directly onto the top of the NAP- 25 columns and allowed to drain into the columns. HA buffer (2.2 ml) was added to the column, moving the DNA into the bottom portion of the column. The collection tube containing the salmon sperm DNA was placed under the column, 1.8 ml HA buffer added to the column, and

40 the eluent containing labeled DNA was collected. The DNA was denatured again by heating for 5 min in a boiling water bath.

The labeled DNA samples were loaded onto dried HA columns, and movement of the DNA through the columns was monitored with a survey meter. Once the DNA moved into the bottom of the columns and started to elute from the bottom, the columns were moved to new collection tubes to begin collecting the labeled samples. The HA columns were washed with 0.5 ml HA buffer and the wash eluent was collected in the same tubes as the labeled DNA. NAP-25 columns equilibrated with three changes of TE + 0.1% SDS were drained until the surface was dry. A disposable serological pipet was used to draw up the labeled DNA recovered from the HA column and the volume recovered was recorded. The labeled DNA was loaded onto the NAP-25 column and the eluent was allowed to drain. An additional volume of TE + 0.1% SDS was added to the column to make the total volume of DNA up to 2.5 ml and allowed to drain. A screw capped culture tube was placed under the column, 3.5 ml TE + 0.1% SDS were added to the column, and the eluent was collected in the tube. Ten microliters of the labeled DNA were transferred to a scintillation vial for gamma counting to determine the strength of the label. Once counted,the labeled DNA was diluted to an activity of 30,000 cpm/ml and stored at -20oC.

S1 Nuclease assay The labeled and unlabeled DNA samples were thawed and then heated in a 65oC waterbath for five minutes. Using an Eppendorf repeating pipettor, 10 ml of labeled DNA were added to the bottom of each reaction tube (200ml polypropylene tubes, Robbins Scientific). Fifty microliters of test DNA were added to each tube. Four tubes contained sheared, native salmon sperm DNA (0.4 mg/ml), four tubes contained DNA homologous to the labeled DNA, and each heterologous DNA was done in duplicate. Following addition of the DNA samples to the reaction tubes, 50 ml high salt buffer were added, the tubes were closed and vortexed eight times. The tubes were transferred to stainless steel racks, a cover placed over the rack, and the entire rack incubated in a 65oC waterbath for 24 h. Following incubation the

41 reactions were stored at -20oC until the rest of the experiment could be performed.

The reaction contents were thawed and allowed to come to room temperature. For each reaction, 1 ml of S1 buffer was added to a plastic digestion tube, followed by 50 ml of denatured salmon sperm DNA (0.4 mg/ml). The contents of each 200 ml reaction tube were quantitatively transferred to the digestion tube, and the reaction tube was washed twice with 100 ml S1 buffer. The washes were also added to the digestion tube. Ten microliters S1 nuclease (10 U/ml) were added to each digestion tube, the contents vortexed three times, and incubated for 1 h in a 50oC water bath. Following the incubation period, 50 ml 1.2 mg/ml native salmon sperm DNA were added to each tube to serve as a precipitation matrix for hybridized DNA. To each tube, 500 ml cold 1 M HCl were added, followed by an incubation at 4oC for 1 h.

After the precipitation was complete, the reactions were filtered through Whatman glass filter strips ( No. 1825 915 GF/F). Each reaction tube was rinsed twice with HCl wash buffer and the rinses were filtered on the same strips. The filter strips were dried under a heat lamp for at least 1 h and once dry, the circles where the DNA was collected were removed with forceps. The circles were placed in the bottom of scintillation vials and counted for 2 min each with a Beckman gamma counter (2).

DNA isolation for RAPD experiments Thawed cell pellets were resuspended in 8 ml cell suspension buffer and transferred to a 125 ml glass stoppered erlenmeyer flask. Dry lysozyme (final concentration 1 mg/ml) was added to the contents of the flask, mixed and incubated at 37oC for 3 h. After the

incubation, 8 ml of 2X lysing solution (55oC) and 4 ml 5 M NaClO4 were added to the mixture. The flasks were incubated at 55oC for 2 h to lyse the cells. Following this incubation, 8 ml phenol:chloroform:isoamyl alcohol (25:24:1) were added to each flask,

42 shaken vigorously to homogenize the mixture and placed on the shaker for 20 min. The mixtures were poured into centrifuge tubes and centrifuged at 17,000 x g and 4oC for 10 min. The aqueous layer was removed from the tube, placed in the flask and the extraction was repeated until no protein layer was present in the centrifuged sample.

After the last extraction, the aqueous layer was placed in a clean flask and 0.6 volumes 100% isopropanol were added to precipitate the nucleic acids. The flask was swirled to clot the nucleic acids, and the alcohol was poured off. Cold 80% ethanol was added to the flask and incubated for 10 to 15 min with occasional swirling to wash the samples. The ethanol was poured off, the nucleic acids stuck to bottom of the flask, and the flask was turned upside down to dry.

The nucleic acids were rehydrated in 8 ml TE buffer and 125 ml RNase mix were added to the flask. Following an incubation at 37oC for 1 h, 2 ml chloroform:isoamyl alcohol were added to each flask. The flasks were shaken vigorously to homogenize the mixture and placed on the shaker for 20 min. The contents of each flask were poured into centrifuge tubes and centrifuged at 17,000 x g and 4oC for 10 min. The aqueous layer was removed, placed in a 100-ml beaker and 800 ml of 3 M sodium acetate were added. The sample was overlayed with two volumes of 95% ethanol and the DNA was collected on a glass rod. The DNA was washed in cold 80% ethanol, the glass rod was inverted and placed in the beaker to let the DNA dry. Once dry, the DNA was resuspended in 1 ml warm TE and stored at -20oC (2).

Determination of DNA concentration The DNA was quantified using a fluorometer (Hoefer TKO-100). As a reference, 830 mg/ml standard l strain CI85757 DNA (USB) was diluted to 250 ng/ml in sterile TE buffer. The fluorometer was allowed to warm up and blanked using 2 ml assay solution in a glass cuvette. The assay solution contained 1X TNE and 0.1 mg/ml Hoechst 33258 (Hoefer Scientific). The fluorometer was standardized by

43 adding 4 ml of l DNA standard to the cuvette and adjusting the machine to read 250 ng/ml. Measurements of DNA concentration were made by adding 4 ml of sample to 2 ml assay solution. Samples were diluted to give a final working concentration of 5 ng/ml and stored at -20oC.

RAPD analysis The sequences of the primers (Operon Technologies) used in this study are given in Table 3. The primers were rehydrated in sterile, milli-Q filtered water to a final concentration of 0.125 mg/ml and stored at -20oC. The working solution of dNTP’s was prepared by diluting 100mM dTTP, dATP, dCTP and dGTP together in sterile, milli-Q filtered water and stored at -20oC.

Table 3. RAPD primer sequences Primer Sequence Name OPA-03 5’-AGTCAGCCAC-3’ OPA-04 5’-AATCGGGCTG-3’ OPA-05 5’-AGGGGTCTTG-3’ OPA-07 5’-GAACGGGGTG-3’ OPA-08 5’-GTGACGTAGG-3’ OPA-10 5’-GTGATCGCAG-3’ OPA-11 5’-CAATCGCCGT-3’ OPA-15 5’-TTCCCGACCC-3’

For at least thirty minutes prior to use, milli-Q filtered water, 50% glycerol, mineral oil, microcentrifuge tubes and rack, gloves, aerosol resistant pipet tips and pipettors were exposed to UV light in a laminar flow hood. Primers, Promega 10X buffer, dNTP’s,

MgCl2 and the DNA samples were thawed at room temperature. The Taq DNA polymerase (5000 U/ml, Promega) was stored in Buffer A (Promega) at -20oC until used.

44 All RAPD reactions were prepared in a laminar flow hood after exposure of the contents of the hood to UV light for 30 minutes. The amount of primer used to obtain a final concentration of 0.6 mM varied due to the different molecular weights of the primers. The reagents were added together to make a “master mix” and aliquots were dispensed into the reaction tubes. Each reaction tube contained 0.5 ml 50% glycerol, 2.5 ml 10X buffer, 1.0 ml dNTP’s (100 mM), 3.0 ml

MgCl2 (3 mM), 0.3 ml Taq polymerase (1.5 U), 0.6 mM primer, 3.0 ml DNA template (15 ng) and the appropriate amount of milli-Q filtered water to make up a final volume of 25 ml. Samples were overlayed with two drops sterile mineral oil. Negative controls in which template DNA was replaced with 3.0 ml milli-Q filtered water were also prepared for each primer.

The RAPD reaction tubes were placed in a PTC-100 thermalcycler (MJ Research) with 1 drop of mineral oil per well. The following temperature profile was programmed: 95oC for 5 min followed by 75 cycles of 94oC for 20 sec, 36oC for 20 sec, and 72oC for 2 min. Upon cycle completion, samples were maintained at 4oC until electrophoresis (6).

A 1.7% (w/v) gel composed of 1.0% Synergel (Diversified Biotech) and 0.7% agarose was poured in preparation for electrophoresis. Synergel and agarose were mixed in a slurry with 15 ml 95% ethanol. TAE buffer (1X) was slowly added to the slurry to a final volume of 300 ml and the flask was weighed. The mixture was heated to melt the Synergel-agarose mixture, the ethanol was evaporated off, and water was added (by weight) to the flask to replace the amount which had evaporated during heating. The mixture allowed to cool slightly before pouring the gel. The PCR amplification product was removed from the tube by inserting a pipet tip below the mineral oil layer, expelling an air bubble from the tip, and immediate withdrawal of a 10 ml volume. The sample was mixed with 3 ml loading buffer on parafilm and loaded onto the gel. The gel was electrophoresed at 3.2

45 V/cm in recirculating 1X TAE buffer, stained in 0.5 mg/ml ethidium bromide for 2 h and photographed.

Isolation, amplification and digoxygenin labeling of individual RAPD bands RAPD reactions were prepared as before with the desired primer to isolate single RAPD bands, which were labeled with digoxygenin for use as probes. Reactions were subjected to the thermal cycling conditions described above. Ten microliters of the RAPD reaction were loaded onto a 1.7% low-melt agarose gel prepared with 1X TAE and electrophoresed at 4oC. The gel was stained and photographed as previously described. Using a sterile razor blade, the RAPD band of interest was cut out of the gel and placed in a microcentrifuge tube.

The microcentrifuge tube was placed in a 65oC waterbath for 10 min to melt the agarose. The DNA was purified from the agarose using Wizard PCR Preps (Promega). One milliliter of PCR preps resin was added to the gel slice, vortexed briefly and then incubated for 1 min with occasional vortexing. The DNA/resin mixture was added to a 3 ml disposable syringe attached to a PCR preps minicolumn and then dispensed into the column. The column was washed with 2 ml 80% isopropanol, centrifuged for 20 sec at 12,000 x g and placed on a new microcentrifuge tube. 50 ml warm (50-60oC) TE were added directly to the column and incubated for 1 min, followed by centrifugation for 20 sec at 12,000 x g. The eluted purified DNA was stored at -20oC.

To generate a higher concentration of DNA for storage and labeling, the purified DNA product was amplified by PCR using the following reagents. Fifty microliter reactions were prepared containing 1.0 ml 50% glycerol, 5.0 ml 10X buffer, 4.0 ml dNTP’s

(200mM), 6.0 ml MgCl2 (3mM), 0.8 ml Taq polymerase (2 U), 2 mM primer, 1.0 ml DNA template and the appropriate amount of sterile milli-Q filtered water to make up the final volume. Each reaction was overlaid with 2 drops sterile mineral oil.

The reactions were placed in the thermalcycler with 1 drop mineral oil in each well. The temperature profile was as follows:

46 95oC for 5 min followed by 30 cycles of 94oC for 30 sec, 36oC for 30 sec, and 72oC for 2 min. Upon completion, the reaction tubes were held at 4oC.

The PCR product was removed from the tube and purified using Wizard PCR Preps (Promega). The reaction product was added to 100 ml direct purification buffer and vortexed briefly to mix. One milliliter of PCR preps resin was added to the sample and vortexed. After a 1 min incubation with occasional mixing, the DNA/resin mixture was added to a 3 ml disposable syringe attached to a PCR preps minicolumn and dispensed into the column. The column was washed with 2 ml 80% isopropanol, centrifuged for 20 sec at 12,000 x g and placed on a new microcentrifuge tube. Fifty microliters of warm TE (50-60oC) were added to the column and after a 1 min incubation, the column was centrifuged for 20 sec at 12,000 x g. The eluted product was stored at -20oC.

To label the RAPD band, 25 ml RAPD reactions were prepared as previously described. The working solution of dNTP’s was replaced with a 10X concentrated Boehringer Mannheim dig-DNA labeling mixture (1 mM dATP, 1 mM dCTP, 1 mM dGTP, 0.65 mM dTTP, 0.35 mM dig-dUTP). Each reaction contained sterile water to a final volume of 25 ml, 0.5 ml 50% glycerol, 2.5 ml 10X buffer, 2.5 ml dig-dNTP mix (100 mM), 3.0 ml MgCl2 (3 mM), 0.4 ml Taq polymerase (2 U) and 1.0 ml template DNA. The reactions were overlaid with 2 drops sterile mineral oil.

The amplification reactions were placed in the thermalcycler with 1 drop of mineral oil per well. The following temperature profile was entered: 95oC for 5 min followed by 30 cycles of 94oC for 30 sec, 36oC for 30 sec and 72oC for 2 min. Upon completion of the reaction, the products were held at 4oC. The PCR product was removed and purified as previously described. The purified probe was stored at -20oC (6).

Estimation of probe yield

47 To estimate the probe yield, serial dilutions of the purified PCR products were prepared in DNA dilution buffer (Boehringer Mannheim) and compared to control DNA (Boehringer Mannheim). The dilutions are described in Table 4.

Table 4. Dilution series for probe estimation

Dilution Final Total concentration dilution A. 2 ml DNA/8 ml buffer 1 ng/ml 1:5 B. 2 ml A/18 ml buffer 100 pg/ml 1:50 C. 2 ml B/18 ml buffer 10 pg/ml 1:500 D. 2 ml C/18 ml buffer 1 pg/ml 1:5,000 E. 2 ml D/18 ml buffer 0.1 pg/ml 1:50,000

One microliter of each dilution of the PCR labeled probe and 1 ml of labeled control DNA were spotted onto a positively charged nylon membrane (Boehringer Mannheim). The membrane was placed on a damp paper towel and UV crosslinked (FB UVXL-1000, Fisher) using the optimal setting. The membrane was placed in a glass petri dish, wetted with 1 ml maleic acid buffer and incubated at room temperature for 5 min with enough Blocking solution to cover the membrane. The Blocking solution was discarded and new Blocking solution containing a 1:5000 dilution of anti-DIG-alkaline phosphatase was added to the membrane. The membrane was incubated with gentle shaking at room temperature for 10 min. The membrane was then washed twice with maleic acid buffer, 5 min per wash, and incubated in detection buffer for 2 min. The detection buffer was discarded and the membrane was placed in a heat sealable bag. Color developing solution was added to the bag, the bag sealed and placed in the dark for 30-60 min until adequate color was developed. The reaction was stopped by the addition of TE. Estimation of yield was done by visual comparison of probe intensity to that of the controls.

Southern transfer and DNA hybridization

48 The protocol given by Boehringer Mannheim for Southern transfer, prehybridization, hybridization and colorimetric detection of hybridized probe was followed with a few modifications. RAPD DNA to be transferred was electrophoresed, stained and photographed as previously described. The gel was shaken gently at room temperature for 10 min each in depurinating solution and denaturing solution, then soaked twice at room temperature for 20 min each in neutralization solution.

The DNA was transferred overnight to a positively charged nylon membrane (Boehringer Mannheim) by capillary action in 10X SSC. After transfer the membrane was placed on a damp paper towel and UV crosslinked using the optimal setting.

The membrane was placed in a heat-sealable bag and incubated in 20 ml/100 cm2 standard prehybridization buffer for 2 h in a 65oC water bath. After prehybridization, the solution in the bag was replaced with an equal amount of prehybridization buffer. One and one-half nanogram/100 cm2 of labeled probe was also added to the bag, the bag sealed and incubated overnight in a 65oC waterbath. After hybridization, the membrane was removed from the bag, placed in a glass baking dish and washed twice in 2X wash solution for 5 min each. Then the membrane was washed twice in 0.5X wash solution for 15 min each.

To start color development, the membrane was incubated at room temperature in Blocking solution with gentle shaking for 1 h. After the intial incubation, the blocking solution was discarded and anti- DIG alkaline phosphatase diluted 1:5000 in blocking solution (20 ml) was added to the membrane. The membrane was incubated with the antibody at room temperature with gentle agitation for 30 min. After the antibody had bound, the membrane was washed twice in 15 ml 1X maleic acid buffer for 15 min each and washed once in 20 ml detection buffer for 2 min. The membrane was placed in a heat sealable bag and the color developing solution was added. Color development was allowed to proceed in the dark at room temperature until sufficient color had been deposited on the membrane. The color development was

49 stopped by the addition of TE buffer to the membrane and the membrane was stored at 4oC in the dark until photographed.

RAPD band analysis The presence (1) or absence (0) of each RAPD band among the strains was determined for each primer by visual examination of the gel photographs. The results for each strain were recorded in an ASCII format as a rectangular matrix consisting of total bands. All detectable RAPD bands present in the strains were analyzed and scored.

Data analysis The percent DNA similarity data were analyzed with the average taxonomic distance algorithm (3, 5). The distance coefficient was utilized in this case because the data were all quantitative real variables without a range of variation, and as such could be treated as points in space. The coefficient calculates the distance between the points and this value is converted into a dissimilarity value. The matrix obtained was subjected to clustering by the unweighted pair group method with arithmetric averages (UPGMA)(5). Cophenetic matrices for the clusters were computed and the correlation between these coefficients and their corresponding rectangular matrix was computed by using normalized Mantel statistics z (5). This determined how much distortion was present in the phenetic tree. The RAPD data were analyzed using either the Jaccard or Dice similarity coefficients (3, 5). The Jaccard coefficient uses only positive matches in the calculation of similarity. This allows characters which are missing in two or more isolates to be ignored, resulting in a similarity value calculated from characters which are present. The Dice coefficient is also a measure of similarity between OTU's, and does not include characters which are absent in each of the isolates being compared. The matrix of coefficients obtained was subjected to clustering by UPGMA. The NTSYS-pc computer program (version 1.8) was used in the analysis of the data (5).

Distance coefficient: dij =Ö1/nåk(xki + xkj)2

50 Jaccard coefficient: a/(n-d) where a = all positive matches, n = total sample size and d = all negative matches Dice coefficient: 2a/(2a + b + c) where a = all positive matches, b and c = unmatches

Multiplex PCR-RFLP for detection of the van ligase All PCR reactions were set up in a laminar flow hood. The pipettors, tips, gloves, racks and tubes were exposed to ultra-violet light for at least 30 min prior to use. Each reaction contained 10 pmol of each of six primers designed for detection of the van ligase in the enterococci. Primer sequences are shown in Table 5.

Table 5. Primer sequences used in multiplex PCR-RFLP reaction for detection of van ligase genes in enterococci (4). Refrer to Patel et al. 1997. Multiplex polymerase chain reaction detection of vanA, vanB, vanC-1, and vanC-2/3 genes in enterococci. J. Clin. Microbiol. 35:703-707.

In addition to the primers, each reaction consisted of 1.25 U Taq polymerase, 200 mM each dNTP, 50 mM KCL, 10 mM Tris-HCl pH 8.3,

1.5 mM MgCl2 and 5 % glycerol. For the Enterococcus strains, a single 24 h colony was picked off a BHI plate supplemented with 4 mg/ml vancomycin and suspended in a 50 ml PCR reaction. Five nanograms of DNA isolated as for RAPD reactions was used as a PCR template for Bacillus popilliae strain ATCC 14706 and Bacillus lentimorbus strain ATCC 14707. The reactions were overlaid with two drops sterile mineral oil and placed in the thermalcycler. The reaction profile was as follows: 95oC for 10 min to lyse the enterococci and denature the template DNA, followed by 60 cycles of 94oC for 1 min, 56oC for 1 min and 74oC for 1 min. Upon cycle completion the reactions were held at 4oC. One microliter of MspI (10 U/ml) and 5 ml of restriction buffer were added to each tube, followed by centrifugation at 13,200 ´ g for 20 sec to drive the enzyme down into the reaction. The tubes were

51 then incubated overnight at 37oC and the restriction products were electrophoresed on a 4% MetaPhor (FMC Corp. MA) gel in 1X TAE.

Primers specific for the ligase gene found in B. popilliae strain ATCC 14706 were designed from the sequenced ATCC 14706 PCR product. The primer sequences are 5’-GCTGCTTGTTATGCGGAATA-3’ (BPOP-FOR) and 5’-AATTGCTTTCGCCGTCTC-3’ (BPOP-REV). The B. popilliae and B. lentimorbus strains were screened for the presence of the ligase gene using these primers and the above PCR conditions.

Paraspore gene detection using PCR

The paraspore gene (cry18Aa1) sequence from a European B. popilliae isolate has been previously described (7). Using the published nucleotide sequence, two sets of PCR primers were designed to cover both the open reading frame found just prior to the gene and the gene itself. The primer sequences are shown in Table 6.

PCR reactions were set up in the laminar flow hood with all tools subjected to 30 minutes UV exposure prior to use. Reaction mixtures contained 25 ng template DNA (isolated as for RAPD

reactions), 5% glycerol, 1 X buffer, 200 mM each dNTP, 3 mM MgCl2, 25 pmol each primer and 0.5 U Taq polymerase. Reaction mixtures were overlaid with two drops sterile mineral oil and placed in the thermalcycler. Cycling parameters were 95oC for 2 min, followed by 35 cycles of 94oC for 1 min, 54oC for 1 min and 72oC for 2 min. Upon completion of the programmed cycles, the tubes were held at 4oC until electrophoresis on a 1 % agarose gel.

Table 6. Primer sequences used for detection of cry genes in B. popilliae and B. lentimorbus. Primer Sequence Location Expected

52 size cryBP1-F 5’-AGGGAATGGACAGAATGG-3’ 1058 962 bp cryBP1-R 5’-GAAAGCTGAACGCCAATC-3’ 2020 cryBP2-F 5’-AGGATGTTCCTCCGATCCCCATCAC-3’ 441 806 bp cryBP2-R 5’-GTTCCGTGGCTCGTAAAATCTCTTC-3’ 1247

PCR conditions for the second set of primers, cryBP2-F and cryBP2-R were identical except for a primer annealing temperature of 56oC.

PCR product sequencing All DNA sequencing was performed at the Mayo Clinic (Rochester, MN). Six microliters of PCR product, 1 ml of 1 U/ml shrimp alkaline phosphatase and 1 ml of 10 U/ml exonuclease I (United States Biochemicals) were incubated at 37oC for 30 min followed by incubation at 80oC for 15 min. One microliter of sequencing primer (3.2 mM) and 1 ml of dimethyl sulfoxide were added to the mixtures and the DNA sequence was determined in both directions using a Taq dideoxy terminator cycle sequencing kit and a 373 A DNA Sequencer (Applied Biosystems, CA). The sequence data were analyzed using version 8 of the Genetics Computer Group Sequence Analysis software (4).

Labeling of vanE PCR product A portion of the vanE gene was amplified from ATCC 14706 using PCR conditions identical to those used for screening the Bacillus strains for the presence of the ligase gene. The amplified product was cleaned using the Wizard PCR Preps system as described for RAPD band labeling. The PCR reaction for digoxygenin labeling was set up as described for the detection of the vanE gene in B. popilliae, with the replacement of the dNTP mix with a digoxygenin labeled dNTP mix. Upon completion of the cycling program, the product was cleaned as detailed before.

53 The probe concentration was determined following the procedure used for RAPD probe determination. The probe was stored at –20oC.

Determination of vanE location in B. popilliae DNA from both B. popilliae strain ATCC 14706 and B. lentimorbus strain ATCC 14707 was digested with MboII in a 20 ml reaction. The reaction included 1X enzyme buffer, 3 U enzyme and 2 mg BSA, 1 mg DNA and Milli-Q water to the final volume. The digestions were incubated at 37oC for 2 h and then electrophoresed on a 1 % agarose gel. As controls, undigested DNA as well as unlabeled vanE PCR product were also run on the gel. The gel was stained with ethidium bromide and photographed under UV light, followed by a Southern transfer to a positively charged nylon membrane as described for RAPD’s. Hybridization of the probe and colorimetric detection were performed as previously described for RAPD bands.

References 1. Billot-Klein, D., L. Gutmann, S. Sable, E. Guittet, and J. van Heijenoort. 1994. Modification of peptidoglycan precursors is a common feature of the low-level vancomycin-resistant species Lactobacillus casei, Pediococcus pentosaceus, Leuconostoc mesenteroides, and Enterococcus gallinarum. J. Bacteriol. 176(8):2398-2405. 2. Dingman, D. W., and D. P. Stahly. 1983. Medium promoting sporulation of Bacillus larvae and metabolism of medium components. Appl. Environ. Microbiol. 46(4):860-869. 3. Handwerger, S. 1994. Alterations in peptidoglycan precursors and vancomycin susceptibility in Tn917 insertion mutants of Enterococcus faecalis 221. Antimicrob. Agent. Chemother. 38(3):473- 475. 4. Johnson, J. L. 1994. Similarity analysis of DNA's, p. 655-682. In P. Gerhardt, R. G. E. Murray, W. A. Wood, and N. R. Krieg (ed.), Methods for General and Molecular Bacteriology, 1st ed. American Society for Microbiology, Washington, D. C.

54 5. Patel, R., J. R. Uhl, P. Kohner, M. K. Hopkins, and F. R. Cockerill, III. 1997. Multiplex polymerase chain reaction detection of vanA, vanB, vanC-1, and vanC-2/3 genes in enterococci. J. Clin. Microbiol. 35:703-707. 6. Rohlf, F. J. 1994. NTSYS-pc numerical taxonomy and multi-variate analysis system version 1.80. Exeter Publishing, Setauket, NY. 7. Woodburn, M. A., A. A. Yousten, and K. H. Hilu. 1995. Random amplified polymorphic DNA fingerprinting of mosquito pathogenic and nonpathogenic strains of Bacillus sphaericus. Int. J. Syst. Bacteriol. 45(2):212-217. 8. Zhang, J., T. C. Hodgman, L. Krieger, W. Schnetter, and H. U. Schairer. 1997. Cloning and analysis of the first cry gene from Bacillus popilliae. J. Bacteriol. 179(13):4336-4341.

55 CHAPTER THREE

Bacillus popilliae and Bacillus lentimorbus, Bacteria Causing Milky Disease in Japanese Beetles and Related Scarab Larvae

Abstract Bacillus popilliae and B. lentimorbus, causative agents of milky disease in Japanese beetles and related scarab larvae, have been differentiated based upon a small number of phenotypic characteristics, but they have not previously been examined at the molecular level. In this study thirty-four isolates of these bacteria were examined for similarity by DNA reassociation (henceforth referred to as DNA similarity). Three distinct but related similarity groups were identified; the first contained strains of B. popilliae, the second contained strains of B. lentimorbus, and the third contained two strains distinct from but related to B. popilliae. Some strains received as B. popilliae were found to be most closely related to B. lentimorbus and some received as B. lentimorbus were found to be most closely related to B. popilliae. Paraspore formation, believed to be a characteristic unique to B. popilliae, was found to occur among a subgroup of B. lentimorbus strains. Growth in media supplemented with 2% NaCl was found to be less reliable in distinguishing the species than was vancomycin resistance, the latter present only in B. popilliae.

Bacillus popilliae and B. lentimorbus are pathogens of Japanese beetles (Popillia japonica) and related scarab larvae. Larvae feeding in the soil consume spores of these bacteria and following spore germination in the larval gut, vegetative cells penetrate into the hemocoel. A period of vegetative growth is

57 followed by asynchronous sporulation and death of the larvae. At the time of larval death, the hemolymph may contain up to 5 x 1010 spores/ml (1). The milky color of the larval hemolymph at the time of death has given the condition the name “milky disease” (8). Because of its action against economically important insect pests, efforts have been made to develop B. popilliae as a biological control agent. However, the inability to mass produce spores in vitro has prevented large scale manufacture and utilization (5).

Dutky (2) reported a difference in color of the hemolymph in insects infected by either B. popilliae (type A disease) or B. lentimorbus (type B disease). In addition, Dutky (2), Gordon et al. (3) and St. Julian and Bulla (9) suggested that a primary distinguishing characteristic between the two named species is the production of a parasporal body by B. popilliae but the absence of this structure in B. lentimorbus. Serological studies prompted Krywienczyk and Luthy (6) to propose a single species, B. popilliae, with three varieties, B. popilliae var. popilliae, B. popilliae var. lentimorbus and B. popilliae var. melolonthae (the last variety based on a European isolate also known as fribourgensis). This approach was similar to that proposed by Wyss (14) who emphasized physiological and morphological characteristics in his taxonomic arrangement. Milner (7) utilized the position and size of the spore and paraspore in the sporangium to group the bacteria. In this system all milky disease isolates were considered varieties of B. popilliae. The A1 group contained strains with small parasporal bodies overlapping the spore. Group A2 had a large parasporal body separated from the spore. Group B1 had a large central spore and lacked a paraspore and group B2 had a small spore and no paraspore. The utilization of these morphological

58 characteristics in species determination is limited because the paraspore is produced at the time of sporulation which only occurs in living larvae. Therefore, only those laboratories capable of collecting and infecting the larvae are able to identify the species (11). It has been reported that B. popilliae will grow in media containing 2% NaCl whereas B. lentimorbus will not grow under these conditions (11).

The genetic relationship between B. popilliae, B. lentimorbus, and less well-known bacteria producing milky disease is unknown. In this study we utilized DNA reassociation to define relationships between these species. Our results validate the existence of the two species and identified the presence of subgroups within the species. Phenotypic characteristics presented by the species and subgroups were investigated to facilitate identification.

RESULTS DNA similarity. DNA was prepared from 34 strains of bacteria that had been originally isolated from scarab larvae suffering from milky disease. This DNA was compared to labeled reference DNA from the type strains of both B. popilliae and B. lentimorbus as well as to three additional strains, one of which was a European isolate sometimes referred to as B. popilliae var. melolonthae (NRRL B-4081), shown in Table 1. The clusters obtained from the distance and correlation matrices were almost identical in their topology. The cophenetic correlations for both clusters were r=0.98 to 0.99, underscoring the extremely high fit between the original matrices and the phenograms. The distance-based phenogram will be used here because it showed higher resolution within the groups. The similarity study revealed the existence of two groups

59 Table 1. Levels of DNA similarity between Bacillus popilliae and Bacillus lentimorbus as determined by the S1 nuclease method

Refer to Rippere et al. 1998. Molecular systematics of Bacillus popilliae and Bacillus lentimorbus, bacteria causing milky disease in Japanese beetles and related scarab larvae. Int. J. Syst. Bacteriol. 48:395-402. of strains (Fig. 1). The first group showed 84 to 97% similarity to the type strain, B. popilliae ATCC 14706T, and a high similarity to BpPj5, another B. popilliae isolate. These strains were primarily North American in origin and most were isolated from diseased Popillia japonica except for a few from Anomola orientalis (northern masked chafer). Two strains, NRRL B-4081 and Bp3, showing markedly lower similarity (77% and 73% respectively) to the ATCC 14706T reference strain than did the other strains of B. popilliae. Bp3 displayed 82% similarity to NRRL B-4081 whereas the other strains of the B. popilliae group showed only 59% to 67% similarity to that reference strain. Two strains, BlPj1 and NRRL B-2522, were received as B. lentimorbus but showed 95% and 86% similarity to the B. popilliae reference strain and 59% and 62% similarity respectively, to the B. lentimorbus type strain. Following growth and sporulation in Japanese beetle larvae, paraspores were detected in NRRL B-2522 but not in BlPj1.

Eight strains showed higher similarity to the B. lentimorbus reference than to B. popilliae. Only one of these was received as B. lentimorbus, while the other seven were received as B. popilliae.

60 Figure 1. UPGMA dendogram of B. popilliae and B. lentimorbus strains based on a distance matrix generated from DNA similarity analysis.

Refer to Rippere et al. 1998. Molecular systematics of Bacillus popilliae and Bacillus lentimorbus, bacteria causing milky disease in Japanese beetles and related scarab larvae. Int. J. Syst. Bacteriol. 48:395-402.

However, these latter seven strains had lower similarity to B. lentimorbus than one strain, KLN2, received as B. lentimorbus (Table 1). Microscopic examination of hemolymph from Japanese beetles or masked chafers infected with six of these strains, Bp7, DGB1, BpCb1, BpCb2, BpPa1, and BpCp1, revealed the presence of parasporal bodies in the sporangia. Strain Bp1 has not yet been retested.

Growth in 2% NaCl or vancomycin. Growth in media supplemented with 2% NaCl has been used as a characteristic to separate B. popilliae from B. lentimorbus. Although we found this to be an accurate indicator of species for most strains tested (Table 2), there were a few exceptions on both solid and liquid media.

Stahly et al. (12) described a selective medium designed for the recovery of B. popilliae spores from soil. This medium utilized vancomycin to select for B. popilliae while suppressing growth of B. lentimorbus and many other soil microorganisms. Although they reported that B. popilliae was generally resistant to vancomycin, there were several isolates that appeared to be susceptible. When we examined the response to vancomycin of the strains studied by DNA similarity, it was found that strains

61 identified as B. popilliae were resistant to vancomycin and all strains identified as B. lentimorbus were sensitive (Table 4). The strains of B. popilliae that Stahly et al. (12) reported as being sensitive to the antibiotic were found to be B. lentimorbus by DNA similarity and one B. lentimorbus strain, BlPj1, that Stahly et al. reported to be resistant, we have found to be B. popilliae.

Table 2. Phenotypic characteristics of Bacillus popilliae and Bacillus lentimorbus strains used in DNA similarity studies

Refer to Rippere et al. 1998. Molecular systematics of Bacillus popilliae and Bacillus lentimorbus, bacteria causing milky disease in Japanese beetles and related scarab larvae. Int. J. Syst. Bacteriol. 48:395-402.

When the MIC’s were determined for the type strains, B. popilliae was found to be highly resistant, MIC’s ranging from 400 to 800 mg/ml, whereas B. lentimorbus was sensitive to <1 mg/ml. All three strains were sensitive to the related glycopeptide antibiotic, teicoplanin.

Discussion DNA similarity analysis was used to elucidate the genetic relationship between 34 isolates of bacteria causing milky disease in scarab larvae. The strains separated into two species based on greater than 70% similarity to the type strains of the species (4, 10, 13). Twenty four were shown to be B. popilliae by their relatedness to the type strain, ATCC 14706T, and eight were shown to be B. lentimorbus by their relatedness to ATCC 14707T. Strains NRRL B-2522 and BlPj1 were received as B.

62 lentimorbus but were found to be most closely related to B. popilliae. Both strains grew with 2% NaCl in the medium (NRRL B- 2522 only in broth) and both were resistant to vancomycin, traits that are associated with B. popilliae. The European isolate referred to as B. popilliae var. melolonthae (NRRL B- 4081) and the North American isolate Bp3 had lower DNA similarity to the B. popilliae type strain than the remaining isolates of this species. The main body of B. popilliae isolates showed less than 70% similarity to NRRL B-4081, suggesting that these strains may constitute a subspecies of B. popilliae. The vancomycin resistance of these two strains points to their relationship to B. popilliae, however, DNA similarity clearly indicates their uniqueness.

Of the eight strains showing greater than 70% simlarity to the B. lentimorbus type strain, seven had been received as B. popilliae. Only one of these, BpPa1, grew with 2% NaCl in the medium (and then only on plates), and all of them were sensitive to 150 mg/ml vancomycin. Six of these strains displayed parasporal bodies when retested by infecting Japanese beetle or masked chafer larvae. Although the presence of a parasporal body has been used as a distinguishing characteristic of B. popilliae, I have shown that paraspores may also be formed by B. lentimorbus. It appears that the paraspore-forming isolates may constitute a distinct subgroup of this species. The strains that were received as B. popilliae but that are now known to be B. lentimorbus had lower similarity to the type strain (73 to 78 %) than KLN2 (90%) received as B. lentimorbus. It is noteworthy that all of the isolates of the second subgroup were isolated from insects other than Popillia japonica.

63 The observation that vancomycin resistance is a uniform trait among strains of B. popilliae, as that species is defined by the DNA similarity, offers a simple phenotypic test for identifying the species. This test appears to be more reliable than growth in the presence of 2% NaCl.

This study focused mainly on North American isolates that were available in pure culture or that I was able to recover from larval material supplied to me. I have not examined A2 or

B2 isolates, and these may reveal further diversity among the milky disease bacteria. There are also some strains that have been described in the literature solely on their appearance in infected larval hemolymph but which have not been grown in vitro. It would be of value to be able to examine their relationships to the better known strains. An understanding of the genetic relationships among these bacteria and the discovery of subgroups within the species may provide insight into the specificity which these bacteria exhibit in their infection of various species of scarab larvae.

References 1. Bulla, L. A., Jr., R. N. Costilow, and E. S. Sharpe. 1978. Biology of Bacillus popilliae. Adv. Appl. Microbiol. 2:1- 18. 2. Dutky, S. R. 1940. Two new spore-forming bacteria causing milky diseases of Japanese beetle larvae. J. Agri. Res. 61(1):57-68. 3. Gordon, R. E., W. C. Haynes, and C. H.-P. Pang. 1973. The Genus Bacillus, vol. 427. U. S. Department of Agriculture, Washington, D. C.

64 4. Johnson, J. L. 1973. Use of nucleic-acid homologies in the taxonomy of anaerobic bacteria. Int. J. Syst. Bacteriol. 23:308-315. 5. Klein, M. G. 1981. Advances in the Use of Bacillus popilliae for Pest Control, p. 183-192. In H. D. Burges (ed.), Microbial Control of Pests and Plant Diseases 1970- 1980, 1 ed. Academic Press, London. 6. Krywienczyk, J., and P. Luthy. 1974. Serological relationship between three varieties of Bacillus popilliae. J. Invertebr. Pathol. 23:275-279. 7. Milner, R. J. 1981. Identification of the Bacillus popilliae group of Insect Pathogens, p. 45-59. In H. D. Burges (ed.), Microbial Control of Pests and Plant Diseases 1970-1980, 1 ed. Academic Press, London. 8. Splittstoesser, C. M., and C. Y. Kawanishi. 1981. Insect diseases caused by bacilli without toxin mediated pathologies, p. 190-199. In E. W. Davidson (ed.), Pathogenesis of Invertebrate Microbial Diseases. Allanheld, Osmun and Company. 9. St. Julian, G., and L. A. Bulla, Jr. 1973. Milky disease, p. 57-87. In T. C. Cheng (ed.), Current Topics in Comparative Pathobiology. Academic Press, New York. 10. Stackebrandt, E., and B. M. Goebel. 1994. Taxonomic Note: A Place for DNA-DNA Reassociation and 16S rRNA Sequence Analysis in the Present Species Definition in Bacteriology. Int. J. Syst. Bacteriol. 44(4):846-849. 11. Stahly, D. P., R. E. Andrews, and A. A. Yousten. 1992. The genus Bacillus-Insect pathogens, p. 1029-2140. In A. Balows, H. G. Truper, M. Dworkin, W. Harder, and K.-H. Schleifer (ed.), The Prokaryotes, 2 ed, vol. 2. Springer- Verlag, New York.

65 12. Stahly, D. P., D. M. Takefman, C. A. Livasy, and D. W. Dingman. 1992. Selective medium for quantitation of Bacillus popilliae in soil and in commercial spore powders. Appl. Environ. Microbiol. 58(2):740-743. 13. Ursing, J. B., R. A. Rossello-Mora, E. Garcia-Valdes, and J. Lalucat. 1995. Taxonomic note: A pragmatic approach to the nomenclature of phenotypically similar genomic groups. Int. J. Syst. Bacteriol. 45(3):604. 14. Wyss, C. 1971. Sporulationsversuche mit drei varietaten von Bacillus popilliae Dutky. Zentralbl. Bakteriol. Parasitenk. Infektionskr. Hyg. II. 126:461-492.

66 CHAPTER FOUR

Randomly Amplified Polymorphic DNA Analysis of Geographically Distinct Isolates of Bacillus popilliae and Bacillus lentimorbus

Abstract Geographically distinct strains of Bacillus popilliae and Bacillus lentimorbus were analyzed using randomly amplified polymorphic DNA (RAPD). Eight decamer primers were tested against nineteen new and seventeen previously described isolates. Of the new isolates, eight were found to belong to the B. popilliae group containing the type strain ATCC 14706T while two Australian strains grouped with the subgroup of B. popilliae containing isolate NRRL B-4081. Nine isolates belong to the B. lentimorbus species with two isolates in the non-crystal forming subgroup and seven contained within the subgroup of crystal forming B. lentimorbus. Vancomycin resistance and 2% NaCl tolerance were tested for all new isolates.

Thirty-four isolates of B. popilliae and B. lentimorbus were previously studied by both DNA-DNA similarity and RAPD analysis, however all but two of those strains were obtained from the northeastern United States (4). I was interested in determining whether isolates from a wider range of geographical regions would reveal the existence of strains of milky disease bacteria with greater diversity. I have investigated nineteen geographically diverse isolates using RAPD analysis.

Results RAPD Analysis. Nineteen geographically diverse milky disease isolates were tested using eight decamer RAPD primers. Also

67 included in the analysis were seventeen strains that had been included in a RAPD study performed by M. Tran (4). These 17 isolates were chosen to include representatives from each possible subgroup suggested by Tran's RAPD analysis. All of the new isolates fell within the previously described species B. popilliae and B. lentimorbus. Examples of RAPD banding patterns are shown for primers OPA-03 and OPA-15 in Figures 1-4.

Primer OPA-03. Negative control reactions containing no template DNA were run for each primer. Bands appearing in the negative control reactions were compared to the test reactions and any band equal in size to a negative control band was not included in the analysis. Primer OPA-03 generated 15 bands of different size. All B. popilliae strains except NRRL B-4081, RM23 and RM29 (Fig. 2, Lanes 12, 14, and 15) had an intense band of approximately 750 bp that was absent in the B. lentimorbus strains. RM23 and RM29 had identical banding patterns (OPA-03) with the exception of one small band found in RM29. However, the banding patterns obtained from these strains using the other primers indicated that they are not identical. Primer OPA-03 generated two major bands with strain NRRL B-4081 (Figure 2, lane 12) which were shared with some B. popilliae (Fig. 1) and some B. lentimorbus (Fig. 2) isolates. Bacillus lentimorbus strains ATCC 14707, Bp11 and Bp21 all non-crystal forming isolates shared a band of approximately 1.2 kb (Figure 2, lanes 3,7, and 10) which was not found in any other B. lentimorbus isolate.

Primer OPA-15. Primer OPA-15 generated a total of 20 bands of different sizes. All of the B. popilliae (Fig. 3) isolates with the exception of Bp22, NRRL B-4081 and RM29 (Fig. 4, Lanes 12, 14, and 15) shared a distinct band of approximately 1.4 kb

68 which was not found in the B. lentimorbus (Fig. 2) strains. Isolate Bp22 (Fig. 3, Lane 19)

Figure 1. RAPD banding patterns of B. popilliae isolates using primer OPA-03. Lane 1, 1 kb ladder; 2, ATCC 14706; 3, DNG2; 4, DNG11; 5, DNG4; 6, NRRL B-2524; 7, KLN3; 8, BlPj1; 9, KLN1; 10, NRRL B-4145; 11, BpPj1; 12, NRRL B-4154; 13, Bp6; 14, Bp9; 15, Bp10; 16, Bp12; 17, Bp13; 18, Bp14; 19, Bp22; 20, Bp23; 21, 1 kb ladder.

69 Figure 2. RAPD banding patterns of B. popilliae (lanes 2, 12, 14 and 15)and B. lentimorbus isolates (lanes 3-11, 13, and 16- 18) using primer OPA-03. Lane 1, 1 kb ladder; 2, BpF; 3, ATCC 14707; 4, Bp1; 5, BpCb1; 6, BpPa1; 7, Bp11; 8, Bp15; 9, Bp19; 10, Bp21; 11, Bp26; 12, NRRL B-4081; 13, Bp25; 14, RM23; 15, RM29; 16, Bp16; 17, Bp17; 18, Bp18; 19, negative control; 20, 1 kb ladder.

70 Figure 3. RAPD banding patterns of B. popilliae isolates using primer OPA-15. Lane 1, 1 kb ladder; 2, ATCC 14706; 3, DNG2; 4, DNG11; 5, DNG4; 6, NRRL B-2524; 7, KLN3; 8, BlPj1; 9, KLN1; 10, NRRL B-4145; 11, BpPj1; 12, NRRL B-4154; 13, Bp6; 14, Bp9; 15, Bp10; 16, Bp12; 17, Bp13; 18, Bp14; 19, Bp22; 20, Bp23; 21, 1 kb ladder.

71 Figure 4. RAPD banding patterns of B. popilliae (lanes 2, 12, 14 and 15) and B. lentimorbus isolates (lanes 3-11, 13, and 16- 18) using primer OPA-15. Lane 1, 1 kb ladder; 2, BpF; 3, ATCC 14707; 4, Bp1; 5, BpCb1; 6, BpPa1; 7, Bp11; 8, Bp15; 9, Bp19; 10, Bp21; 11, Bp26; 12, NRRL B-4081; 13, Bp25; 14, RM23; 15, RM29; 16, Bp16; 17, Bp17; 18, Bp18; 19, 1 kb ladder; 20, negative control.

72 gave a unique banding pattern not shared by any other strain tested with this primer. This isolate did share a band of 450 bp with all of the B. lentimorbus (Fig. 4) strains. Primer OPA- 15 was not able to distinguish among the crystal forming (Fig. 4, Lanes 4-6, 8, 9, 11, 13, 16-18) and the non-crystal forming (Fig. 4, Lanes 3, 7, 10) B. lentimorbus isolates.

Figure 5. Dendogram illustrating the relationships between strains of B. popilliae and B. lentimorbus generated from RAPD analysis.

73 Analysis. The bands derived from each primer were scored as present or absent for each isolate and combined into a matrix. The matrix was analyzed using the NTSYS-pc computer program (5). A dendogram was generated using the Jaccard coefficient (Figure 5). The analysis of the RAPD bands for each strain identified nine B. lentimorbus isolates and ten B. popilliae strains among the 19 new isolates. The two Australian isolates, RM23 and RM29 were most closely related to the European isolate, NRRL B-4081, described in the previous chapter. There are no apparent groupings according to host insect or the geographic origin of the isolates.

Growth in 2% NaCl or vancomycin. Each isolate was tested for growth on MYPGP plates containing 2% NaCl and MYPGP plates containing 150 mg/ml vancomycin. In addition, each isolate was screened in a PCR based assay for the presence of the vanE ligase gene (refer to Chapter 6). The vanE gene is related to the vanA and vanB ligases found in enterococci. The van ligases, along with several other van gene products confer resistance to vancomycin. These results are shown in Table 1.

Table 1. Characteristics of B. popilliae and B. lentimorbus isolates from diverse host insects and geographical regions. Strain 2% NaCl3 VmR4 vanE5 ATCC 147061 + + + Bp231 + + + BlPj11 + + + NRRL B-41451 +/- + + KLN11 + + + BpPj11 + + + NRRL B-41541 +/- + +

74 Strain 2% NaCl3 VmR vanE5 DNG21 + + + Bp121 - + + Bp131 + + + Bp141 + + + DNG111 + + + DNG41 + + + NRRL B-25241 + + + Bp91 + + + Bp101 + + + BpF1 - + + KLN31 + + + Bp61 + + + Bp221 + - ND6 NRRL B-40811 - + + RM231 - - - RM291 - - - ATCC 147072 - - - Bp112 - - - Bp212 - - - Bp12 - - - BpCb12 - - - BpPa12 V7 - - Bp152 - - - Bp262 - - - Bp192 - - - Bp252 - + + Bp162 - + + Bp172 - + + Bp182 - - - 1 B. popilliae

75 2 B. lentimorbus 3 Growth on MYPGP plates supplemented with 2% NaCl 4 Growth on MYPGP plates supplemented with 150 mg/ml vancomycin 5 vanE gene detected using PCR assay 6 ND = not determined 7 V = variable reactions

When tested for growth on 2% NaCl eighteen out of twenty- three B. popilliae strains were capable of growing on MYPGP plates supplemented with 2% NaCl. Strains Bp12, BpF, RM23, RM29 and NRRL B-4081 were unable to tolerate the NaCl. Three of the five strains (RM23, RM29, NRRL B-4081) are found on a branch of the dendogram that is distinct from the majority of the B. popilliae species. These three isolates form a subgroup of B. popilliae that is almost as different from B. popilliae as B. lentimorbus is different from B. popilliae. Twelve out of thirteen B. lentimorbus strains were negative for growth in 2 % NaCl. Strain BpPa1 had variable reactions to the NaCl concentration.

Twenty of twenty-three B. popilliae isolates (Bp22, RM23 and RM29 were negative) tested positive for growth on MYPGP supplemented with 150 mg/ml vancomycin. The vanE ligase gene was undetectable in these three strains using the PCR assay developed for that purpose. RM23, RM29 and Bp22 are distant from the majority of the B. popilliae strains in the dendogram generated from the RAPD data. Ten out of thirteen B. lentimorbus strains were vancomycin sensitive; Bp16, Bp17 and Bp25 were able to grow on MYPGP plates containing 150 mg/ml

76 vancomycin and the vanE ligase gene was detected in these strains using the PCR.

Discussion

The inclusion of nineteen new B. popilliae and B. lentimorbus strains in a RAPD analysis with seventeen strains that had been previously analyzed by DNA simlarity and RAPD analysis resulted in the identification of ten B. popilliae and nine B. lentimorbus isolates. Two isolates formed a cluster on the dendogram with the B. popilliae NRRL B-4081 subgroup while eight isolates were found in a cluster with the major B. popilliae group. Seven strains (Bp15 through Bp18 in Figure 5) analyzed for the first time were most closely related to the crystal-forming strains of B. lentimorbus (Bp1, BpCb1, and BpPa1) that had previously been studied by DNA similarity and by RAPD. Two (Bp11 and Bp21) were most closely related to the B. lentimorbus type strain. The strains analyzed in this study are diverse both with respect to geographic origin and host insect. There were no apparent grouping patterns relating to either geographic origin or host insect in either species. At present, the data supports the classification of milky disease organisms into two species differentiated only by their reactions to 2 % NaCl and vancomycin (4). Combining strains that had been previously analyzed by the RAPD technique with new strains also being tested by RAPD analysis appears to be of use in integrating two separate studies. One strain representing each terminal group of the previous RAPD study was retested in this study. These seventeen strains retained their original placement in the dendogram generated by this study, indicating that the placement of the nineteen new strains is most likely accurate within the context of the original study (4).

77 Results for two of the eight primers used in the RAPD analysis were shown in Figures 1-4. These primers were chosen as representative of the type of banding patterns obtained from these bacteria. It may be possible to use a RAPD band generated by one of these primers to distinguish groups within these species. For example, the non crystal-forming isolates of B. lentimorbus share a band of 1.2 kb (OPA-03) which could be used to identify them. It is possible that a RAPD band common to all the milky disease bacteria could be used to produce a probe for their rapid identification or that a probe could be produced to distinguish between species.

Bacillus popilliae has been distinguished from B. lentimorbus on the basis of the ability to grow in the presence of 2% NaCl and formation of a parasporal crystal during sporulation (1). Resistance to vancomycin also appears to be useful in distinguishing between the two species (4). The nineteen geographically distinct strains were tested for these characteristics. All of the newly identified B. lentimorbus strains were negative for growth in the presence of 2 % NaCl. However, B. popilliae strains Bp12, BpF, NRRL B-4081, RM23 and RM29 were also negative. According to the RAPD results, NRRL B- 4081, RM23 and RM29 are more closely related to each other than to the other B. popilliae strains but Bp12 and BpF are closely related to other B. popilliae strains that test positive for growth in 2% NaCl. Rm23 and RM29 were isolated from Anoplognathus porosus and Lepidiota picticollis, respectively, in Australia, Bp12 was isolated from Holotrichia oblita in China, NRRL B-4081 was isolated from Melolontha melolontha in Europe and BpF was isolated in Europe (insect unknown). The ability to grow in 2% NaCl appears to be useful in

78 differentiating the species since a high percentage of strains possess this characteristic while it is lacking in the majority of B. lentimorbus strains.

When tested for vancomycin resistance, all of the B. popilliae strains were positive for growth on plates supplemented with the antibiotic and positive for the presence of the vanE gene except isolates Bp22, RM23 and RM29. Bp22 was isolated from a Phyllophaga sp. in Panama, and appears on the RAPD dendogram as a separate cluster. It is almost as dissimilar to the majority of the B. popilliae isolates as strains NRRL B-4081, RM23 and RM29. It is possible that these four represent distinct varieties of the species although additional isolates would be required to verify this suggestion. Six of the nine B. lentimorbus isolates studied by RAPD tested negative for resistance to vancomycin. Bp16, Bp17 and Bp25 were able to grow on plates supplemented with vancomycin and all contain the vanE gene within their genome. These three strains form a single cluster of the dendogram within the B. lentimorbus species and appear to be fairly similar to one another. Bp16 was isolated from Polyphyla comes in North Carolina, Bp17 was isolated from Phyllophaga crinita in Texas and Bp25 was isolated from Cyclocephala parallela in Florida, showing no pattern with regard to either geography or host insect. When both growth in 2% NaCl and vancomycin resistance are considered, differentiation of all of the strains except Rm23 and RM29 is possible. These two characteristics in combination appear to be sufficient for identification of milky disease bacteria as either B. popilliae or B. lentimorbus.

REFERENCES

79 1. Dutky, S. R. 1940. Two new spore-forming bacteria causing milky diseases of Japanese beetle larvae. J. Agri. Res. 61(1):57-68. 2. Krywienczyk, J., and P. Luthy. 1974. Serological relationship between three varieties of Bacillus popilliae. J. Invertebr. Pathol. 23:275-279. 3. Milner, R. J. 1981. Identification of the Bacillus popilliae group of Insect Pathogens, p. 45-59. In H. D. Burges (ed.), Microbial Control of Pests and Plant Diseases 1970-1980, 1st ed. Academic Press, London. 4. Rippere, K. E., M. T. Tran, A. A. Yousten, K. H. Hilu, and M. G. Klein. 1998. Bacillus popilliae and Bacillus lentimorbus, bacteria causing milky disease in Japanese beetles and related scarab larvae. Int. J. Syst. Bacteriol. 48:395-402. 5. Rohlf, F. J. 1994. NTSYS-pc numerical taxonomy and multi- variate analysis system version 1.80. Exeter Publishing, Setauket, NY. 6. White, R. T., and S. R. Dutky. 1940. Effect of the introduction of milky diseases on populations of Japanese beetle larvae. J. Econ. Entomol. 33(2):306-309. 7. Wyss, C. 1971. Sporulationsversuche mit drei varietaten von Bacillus popilliae Dutky. Zentralbl. Bakteriol. Parasitenk. Infektionskr. Hyg. II. 126:461-492. 8. Zhang, J., T. C. Hodgman, L. Krieger, W. Schnetter, and H. U. Schairer. 1997. Cloning and analysis of the first cry gene from Bacillus popilliae. J. Bacteriol. 179(13):4336- 4341.

80 CHAPTER FIVE Identification and detection of the cry gene in strains of Bacillus popilliae and Bacillus lentimorbus

Abstract An assay for detection of the cry18Aa1 gene was developed using a combination of two PCR primer pairs. The cry18Aa1 gene was detected in 31 of 35 B. popilliae isolates and in 1 of 18 B. lentimorbus isolates. When hemolymph smears were examined microscopically, a parasporal crystal was seen in three of the four B. popilliae strains where the PCR primers could not amplify the paraspore gene. The fourth strain was not tested due to the unavailability of infected hemolymph. A paraspore was also detected by microscopic examination in a subgroup of 14 B. lentimorbus strains. Primer CryBp1 products were of the expected size, however they were not identified by sequencing. The ATCC 14706T CryBp2 PCR product was sequenced, compared to the published cry gene sequence and found to vary from the published sequence. In combination, the primer pairs CryBp1 and CryBp2 are effective at detecting the paraspore gene in most B. popilliae isolates, but are unable to detect the B. lentimorbus paraspore gene.

Results Detection of the cry operon. Two sets of PCR primer pairs were designed for the cry18Aa1 gene in B. popilliae. The first pair (CryBp1) amplifies a DNA fragment from position 1058 to position 2020 of the published cry18Aa1 sequence. The second pair (CryBp2) amplifies from bp 441 to bp 1247 of the published sequence(1). This fragment begins in the orf1 that precedes the

81 cry18Aa1 gene and ends in the 5’ end of the cry18Aa1 gene (Figure 1). The PCR products produced by the CryBp2

CryBp1

CryBp2

Figure 1. Structure of the Bacillus popilliae cry18Aa1 operon.

primer pair had fewer non-specific amplification products than the products gained from the CryBp1 pair. As a result, the ATCC 14706 and NRRL B-4081 PCR products obtained from primer pair CryBp2 were sequenced and compared with the published sequence (Figure 3). Primer pair CryBp1 was homologous enough to detect the cry gene in 31 of 35 B. popilliae isolates (the gene was undetected in isolates NRRL B-4154, NRRL B-2522, RM23 and RM29), but was able to identify the gene in only one B. lentimorbus isolate, DGB1. Primer pair CryBp2 was designed as a result of the nonspecific amplification resulting in multiple bands that occurred with the CryBp1 primers. CryBp2 primers were able to detect the cry operon in 28 of 35 B. popilliae strains. Isolates BpPj5, Bp3, Bp22, RM23, RM29, NRRL B-2522, and NRRL B-4145 showed no amplification of the operon under the conditions tested. CryBp2 primers failed to amplify the cry gene of all B. lentimorbus isolates examined in this study (Table 1).

82 Table 1. Detection of the paraspore crystal in strains of B. popilliae and B. lentimorbus by visualization and PCR. Strain Paraspore1 CryBp22 B. popilliae ATCC 14706T ND3 + A8 ND + BlPj1 - + Bp3 ND - Bp6 ND + Bp9 - + Bp10 + + Bp12 - + Bp13 - + Bp14 + + Bp22 + - Bp23 + + BpCh1 ND + BpF ND + BpPj1 ND + BpPj2 ND + BpPj3 ND + BpPj4 ND + BpPj5 ND - DNG1 ND + DNG2 ND + DNG4 ND + DNG10 ND + DNG11 ND + DNG12 ND + KLN1 ND +

83 Strain Paraspore1 CryBp2 KLN3 ND + NRRL B-2309 ND + NRRL B-2522 + - NRRL B-2524 ND + NRRL B-4081 ND + NRRL B-4145 ND + NRRL B-4154 ND - RM23 + - RM29 + - B. lentimorbus ATCC 14707T ND - Bp1 + - Bp7 + - Bp11 - - Bp15 + - Bp16 + - Bp17 + - Bp18 + - Bp19 + - Bp21 - - Bp25 + - Bp26 + - BpCb1 + - BpCb2 + - BpCp1 + - BpPa1 + - DGB1 + - KLN2 ND - 1Paraspore seen in phase contrast microscopic examination of hemolymph smear

84 2Gene encoding paraspore detected using PCR primer pair CryBp2 3ND = not determined

The PCR products for strains ATCC 14706 and NRRL B-4081 were chosen for sequencing. The NRRL B-4081 product obtained from CryBp2 primers was of the expected size (806 bp), however the product from ATCC 14706 was approximately 140 bp shorter than expected (Figure 2).

Figure 2. ATCC 14706 and NRRL B-4081 PCR products using primer pair CryBp2. Lane 1, 1 Kb DNA ladder; Lane 2, ATCC 14706; Lane 3, NRRL B-4081

85 In addition, PCR products obtained from all other isolates using primer pair CryBp2 were approximately 140 bp smaller than expected (Figure 2). The sequences obtained from strains ATCC 14706 and NRRL B-4081 using primer pair CryBp2 were compared to the published sequence(1). The sequence obtained from NRRL B- 4081 was identical to the published sequence, while the sequence found in ATCC 14706 was somewhat different. The nucleic acid sequence comparison is shown in Figure 3 and the amino acid sequence comparison is shown in Figure 4.

453 cry18Aa1 CGATCCCCAT CACAAAGAAA TTTCTATTTG CTGCACAGAA AGTATCTGTA ATCC 14706

503 cry18Aa1 TAGATCATGT ACTGAAATGC AGTGTGGAAA CCAGCCCCCA TCATCATGTG ATCC 14706 ------C------T--

553 cry18Aa1 GACTGCCATC ATGTGGTAGT TCATGATTTG AAAGCAATCC CAATCCGTGA ATCC 14706 ---G----C- ----A------TG------

603 cry18Aa1 AGATCATTGC CGGTTCGTTA AAGTTACGGG GAACTTTCAA TTTCATTATG ATCC 14706 ---G-----T -A---T------T------G--

653 cry18Aa1 TAAAGGATTT G1TAAACCGAA ACACAGGCTG GAATGACCCG AGGCGATTTG ATCC 14706 ------C- -1A-T------A---T------CT-T------GG-

703 RBS cry18Aa1 GATAGATTTG AATGCTCATC ATATGAAGGA GGCTATTGGT ATG2AACAATA ATCC 14706 ---C------C-A ----T-G------C--A-- ---2------T

86 753 cry18Aa1 ATTTTAATGG TGGAAATAAT ACAGGAAATA ACTTTACTGG AAATACTCTA ATCC 14706 -C----T--- :AA--G-::: :::::::::: :TC-A-GC-- -C--CA-A--

803 cry18Aa1 AGCAACGGAA TTTGTACGAA AAAAAATATG AAAGGAACCC TAAGCAGAAC ATCC 14706 -A---TAACG :::---:::- TGG-----C- :::::::::------

853 cry18Aa1 TGCTATATTT TCAGATGGGA TTAGTGATGA TTTAATTTGT TGTCTAGATC ATCC 14706 G---:::::: :::::::::: :::::::::: :----C-::: :::::::::-

903 cry18Aa1 CTATATATAA CAATAACGAT AACAATAACG ATGCTATTTG TGATGAGTTA ATCC 14706 ---C-A------:::-T- --TCG-GGT- --::::--A- -T-C--A-::

953 cry18Aa1 GGTTTAACTC CAATAGATAA CAATACGATA TGCAGTACTG ATTTTACTCC ATCC 14706 ------TTT--- G-T----A-- G-----T---

1003 cry18Aa1 CATAAATGTA ATGAGAACAG ATCCTTTTCG CAAGAAATCA ACACAAGAAC ATCC 14706 --G------C-----A------G-A------T

1053 cry18Aa1 TCACAAGGGA ATGGACAGAA TGGAAAGAAA ATAGTCCTTC TTTGTTTACA ATCC 14706 ---T------A---G------

1103 cry18Aa1 CCGGCAATTG TAGGTGTCGT TACCAGTTTT CTTCTTCAAT CATTAAAAAA ATCC 14706 G-AC------TA------AC------G--G ------

1153 cry18Aa1 ACAAGCAACT AGCTTTCTTT TAAAAACTTT GACAGACCTA TTATTTCCTA ATCC 14706 --T--T-G-G G--AGAG------TGT-A------A----T ------

87 1203 cry18Aa1 ATAACAGT ATCC 14706 --C-----

Figure 3. B. popilliae cry18Aa1 gene sequence. - = identical base : = missing base 1End of orf1 (1) 2Start codon of cry gene (1)

1 Cry18Aa1 MNNNFNGGNN TGNNFTGNTL SNGICTKKNM KGTLSRTAIF SDGISDDLIC ATCC 14706 ---Y-I-KVL S-HHINN-GN GN:::::::: ::------:: ::::::::::

51 Cry18Aa1 CLDPIYNNND NNNDAICDEL GLTPIDNNTI CSTDFTPINV MRTDPFRKKS ATCC 14706 :-T-T:---V -RG-LVTN:: ------F- G-NG-I-R-- T-K-----RT

101 Cry18Aa1 TQELTREWTE WKENSPSLFT PAIVGVVTSF LLQSLKKQAT SFLLKTLTDL ATCC 14706 ---FI------K-A---- AP----I--T --EA---LVA GRV-MS--N-

151 Cry18Aa1 LFPNNS ATCC 14706 ------

Figure 4. Deduced amino acid sequence comparison of cry genes. - = identical amino acid : = missing amino acid

88 Discussion Each strain in this study was examined for the presence of a gene encoding a paraspore protein using two PCR primer pairs. A hemolymph sample was available for 11 B. popilliae isolates and these were examined microscopically for the presence of a parasporal crystal. Four strains, Bp9, Bp12, Bp13, and BlPj1, lacked a parasporal crystal when examined by phase contrast microscopy. These four strains were isolated from different host insects in different geographical regions. The B. lentimorbus strains tested microscopically for the presence of a paraspore revealed a distinctive pattern. A subgroup of B. lentimorbus strains with the ability to produce a paraspore was identified in the original DNA similarity study (Chapter 3). This paraspore-forming subgroup of strains did not group with the type strain of the species, ATCC 14707. The B. lentimorbus strains able to make a parasporal body were also identified through use of RAPD (Chapter 4). They formed a subgroup of the species along with the representative strains from that group chosen from the RAPD study by M. Tran. These strains included Bp15 (Cyclocephala lurida, Texas), Bp26 (Cyclocephala parallela, Florida), Bp19 (Rhopaea morbillosa, Australia), Bp25 (Cyclocephala hirta, New York), Bp16 (Polyphyla comes, North Carolina), Bp17 (Phyllophaga crinita, Texas) and Bp18 (Anomala diversa, Japan) representing a wide range of host insects and geographical regions.

The cry operon was detected in the isolates by use of two PCR primer pairs, CryBp1 and CryBp2. CryBp1 primers amplify a fragment that is internal to the cry18Aa1 gene, beginning near the N-terminal end of the protein. This primer pair gave several bands in addition to the expected product, and as a result, this PCR product was not sequenced. This primer pair

89 could still be used to detect the presence of the parasporal gene in the majority of the B. popilliae strains. These reactions contained fragments of the expected size, indicating that they are most likely the correct amplification product, although this is not certain because the product was not sequenced. Four strains, NRRL B-2522, NRRL B-4154, RM23 and RM29 did not produce a PCR product when tested with this primer pair. This primer pair was also unable to detect the presence of the parasporal gene in any of the B. lentimorbus strains identified in this study. This could be due to a change in one or both of the primer regions causing the primers to be unable to anneal to the template under the conditions tested. This could also occur if these strains produce a paraspore protein with a different amino acid sequence than that found in the B. popilliae strains in which the paraspore gene was detected. Primer pair CryBp2 was designed and used to amplify a portion of orf1, the spacer region and the 5' region of the cry18Aa1 gene.

Primer pair CryBp2 identified the cry18Aa1 gene in 28 of 35 B. popilliae strains and was unable to detect the gene in any of the B. lentimorbus strains tested. This primer pair failed to detect the paraspore gene in B. popilliae strains NRRL B-4154, NRRL B-2522, Bp3, BpPj5, Bp22, RM23 and RM29. Of these strains, Bp22, BpPj5 and Bp3 tested positive with the other primer pair, CryBp1, indicating that the paraspore gene may be similar to the genes in the other B. popilliae strains. The open reading frame that precedes the cry18Aa1 gene in these strains may vary enough from the forward primer sequence that the primer was unable to anneal to the template and produce a reaction product. In addition, since the gene sequence of only one strain has been deposited in the databanks, the reverse primer may be in a region of the paraspore gene that is somewhat more variable than

90 other regions. The four B. popilliae strains RM23, RM29, NRRL B-2522 and NRRL B-4154, which tested negative with both primer pairs could contain paraspore genes that vary widely from the published sequence, preventing the primer from annealing and detecting the gene. Strains RM23 and RM29 do contain a paraspore, as determined microscopically, but this was not determined for NRRL B-4154. It is possible that this strain does not produce a paraspore. Primer pair CryBp2, did not produce a product with any of the paraspore-forming B. lentimorbus strains tested in this study, supporting the idea that these isolates may have a protein that varies widely or is entirely different from the protein that Zhang et al. studied (1).

References 1. Zhang, J., T. C. Hodgman, L. Krieger, W. Schnetter, and H. U. Schairer. 1997. Cloning and analysis of the first cry gene from Bacillus popilliae. J. Bacteriol. 179(13):4336- 4341.

91 CHAPTER SIX

DNA Sequence Resembles vanA and vanB in the Vancomycin-Resistant Biopesticide Bacillus popilliae.

Abstract

Biopesticidal powders containing spores of vancomycin resistant Bacillus popilliae have been used for more than 50 years in the United States for suppression of Japanese beetle populations. The basis for vancomycin resistance in these bacteria was investigated using a polymerase chain reaction assay designed to amplify the vanB ligase genes in enterococci. An amplicon was identified and sequenced. The amplified portion of the putative ligase gene in B. popilliae had 77% and 68-69% nucleotide identity to the sequences of the vanA gene and the vanB genes, respectively. There was 75% and 69-70% identity between the translation of the putative ligase gene in B. popilliae and the deduced amino acid sequence of the vanA gene and the vanB genes, respectively. We have identified a gene resembling vanA and vanB in B. popilliae and determined that it is located either on a plasmid greater than 16 kb in size or on the chromosome. Based on sequence similarity, the gene in B. popilliae may have had an ancestral gene in common with vancomycin resistance genes in enterococci.

Biopesticidal powders containing spores of Bacillus popilliae have been used for more than 50 years in the United States for the suppression of Japanese beetle (Popillia japonica Newman) populations (2, 5). The Japanese beetle feeds on more

82 than 257 different plants and annually destroys turf, field crops, fruits and ornamentals worth millions of US dollars. Bacillus popilliae was the first microorganism registered in the US as a pesticide (15).

Pridham et al. (11) first observed that B. popilliae NRRL B-2309 was vancomycin resistant in the 1960’s. Subsequently, Stahly et al. (14) described several vancomycin-resistant isolates of B. popilliae and described a selective medium containing vancomycin, for quantitation of B. popilliae in soil and in commercial spore powders.

High level vancomycin resistance was first described in enterococci in isolates from 1986 (6). The genes associated with high level vancomycin resistance, vanA and vanB, encode a ligase responsible for the synthesis of the depsipeptide D- alanyl-D-lactate which is incorporated into a pentapeptide peptidoglycan cell wall precursor (which terminates in D-alanyl- D-lactate) to which vancomycin binds poorly. In contrast, in vancomycin-susceptible cells, vancomycin complexes with the D- alanyl-D-alanine termini of normal pentapeptide peptidoglycan cell wall precursor thereby inhibiting cell wall synthesis.

Enterococci cause about five percent of cases of infective endocarditis in humans (a uniformly fatal illness if untreated) and are now the second most common pathogens isolated from hospital acquired infections. Vancomycin-resistant enterococci (VRE) are increasingly isolated from clinical specimens, and infections caused by VRE can be untreatable by any currently available antimicrobial or antimicrobial combination. With the increasing presence of VRE in clinical specimens, there is concern regarding the possibility that vancomycin-resistance

93 genes present in VRE will be transferred to other more virulent gram-positive bacteria. It has been demonstrated, for example, that in the laboratory, vancomycin resistance is readily transferred from enterococci to other gram-positive organisms, including Staphylococcus aureus (7).

The origin of vancomycin resistance genes in enterococci is unknown. One hypothesis as to their origin is that vancomycin resistance present in environmental organisms has been transferred to enterococci, and these transcipients have been selected under the pressure of increased oral and parenteral vancomycin usage in clinical practice. Environmental organisms carrying genes resembling vanA and vanB have not, however, been identified to date.

I hypothesized that vancomycin resistance in B. popilliae might be conferred by a gene resembling the vancomycin- resistance genes in enterococci. Herein we describe the use of a PCR assay originally designed for use in enterococci to detect a gene resembling vanA and vanB, by nucleic acid and amino acid homology studies, in B. popilliae (9).

Results

Previously described multiplex vanA and vanB ligase gene primers were used to amplify the putative ligase gene of B. popilliae ATCC 14706 using PCR (9). The PCR product showed a restriction pattern that was different from those obtained from enterococcal isolates carrying the vanA and vanB genes (Figure 1). The sequence of the amplicon obtained was compared to that of four previously characterized enterococcal isolates carrying the vanB genes (isolates 55, 94, 45, and 91), one previously

94 Figure 1. Multiplex PCR-RFLP of enterococcal isolates carrying the vanA and vanB ligase genes and B. popilliae ATCC 14706. Lane 1, 50 bp ladder; lane 2, empty; lane 3, vanA isolate; lane 4, vanB isolate; lane 5, vanC-1 isolate (negative control); lane 6, vanC-2/3 (negative control); lane 7, ATCC 14706; lane 8, B. lentimorbus ATCC 14707.

95 characterized enterococcal isolate carrying the vanA gene (isolate 1), five previously characterized enterococcal isolates carrying the vanC-1 genes, and six previously characterized isolates carrying the genes vanC-2/3 (10). The 708 bp fragment amplified from B. popilliae ATCC 14706T (figure 1) had 77% nucleotide identity to the sequence of the vanA gene (9), and 68-69% nucleotide identity to the sequences of the vanB genes of isolates 45 and 91 (9). Comparisons of the putative amino acid sequence of the B. popilliae ATCC 14706 ligase gene to that of four previously characterized vanB genes and one previously characterized vanA gene (10) are shown in figure 2. There was 75% identity between the deduced amino acid sequence of the putative ligase gene in B. popilliae ATCC 14706 and that of the deduced amino acid sequence of the previously described vanA gene, and 69-70% identity between the deduced amino acid sequence of the putative ligase gene in B. popilliae and that of the translation of the vanB genes in isolates 45 and 91 (9). When conservatively substituted, non-identical amino acids were considered, the homology increased to 82% (vanA gene) and 78% (vanB gene isolate 45). For comparison, there is 73% nucleic acid and 75% amino acid identity between the vanA gene and the vanB gene of isolate 91 (10). Notably, there was 44-50% nucleotide identity between base pairs 135-490 (Figure 2) of the B. popilliae putative ligase gene and the vanC-1 or vanC-2/3 genes in enterococci or the D-alanine:D-alanine ligase (ddl) genes in Lactobacillus spp. and Leuconostoc mesenteroides (data not shown) (9, 1). I have designated the putative ligase gene in B. popilliae ATCC 14706 “vanE”.

Figure 2. Sequence of the putative ligase gene in B. popilliae ATCC 14706. Eleven other isolates of B. popilliae had identical sequences from position 34 through 572 including Bp6, NRRL B-

96 2309, NRRL B-4145, Bp23, BpPj2, BlPj1, BpPj1, BpPj3, BpPj4, KLN1 and BpCh1. The sequences in Bp3 (12 bp different from ATCC 14706) and NRRL B-4081 (11 bp different from ATCC 14706 are shown). The sequence in BpF (shown) was identical to that in KLN3, DNG1, DNG2, DNG4, DNG10, DNG11, DNG12, Bp9, Bp10, Bp12, Bp14, NRRL B-2524, and Bp13 and differed from that in ATCC 14706 by 7 bp. The sequence in NRRL B-2522 was identical to that in NRRL B-4154 and differed from that in ATCC 14706 by 2 bp. The sequence in Bp17 (shown) was identical to that in Bp25 and Bp16 and differed from that in ATCC 14706 by 10 bp.

Refer to Rippere et al. 1998. A gene resembling vanA and vanB in the vancomycin-resistant biopesticide Bacillus popilliae. J. Infect. Dis. 178:584-588.

Figure 3. Comparison of the deduced amino acid sequence of the putative ligase gene in B. popilliae ATCC 14706 to the deduced amino acid sequences of four previously characterized vanB genes (isolates 55, 94, 45, and 91) and one previously characterized vanA gene (isolate 1) (10).

Refer to Rippere et al. 1998. A gene resembling vanA and vanB in the vancomycin-resistant biopesticide Bacillus popilliae. J. Infect. Dis. 178:584-588.

Patel et al. have previously identified sequence variability in the van genes of clinical isolates of enterococci (10). Given this, I hypothesized that there might be sequence variability in the putative ligase genes of B. popilliae, and that possibly, in some isolates of B. popilliae, I might find a ligase gene with a sequence even more similar to the vanA and

97 vanB genes than was found in B. popilliae ATCC 14706. In order to test this hypothesis, a segment of the putative ligase genes of a collection of B. popilliae isolates were sequenced as follows. Based on the sequence of the ligase gene in B. popilliae ATCC 14706, a set of PCR primers was designed and used to amplify the putative ligase gene in a collection of 33 B. popilliae isolates. Six distinct classes of sequences were found amongst these isolates and are shown in Figure 2. Members of each class are described in the legend of Figure 2. All had homology to the vanA and vanB sequences in enterococci, although none was markedly more similar to the enterococcal van genes than that found in B. popilliae ATCC 14706.

I attempted to determine the location of the vanE ligase gene within the B. popilliae genome by probing both isolated plasmids and isolated large chromosomal fragments. The vanE PCR product was labeled with digoxygenin and then used to probe Southern blots of isolated plasmids. There was no apparent hybridization of the probe with any of the plasmids, but isolated chromosomal DNA showed a strong positive hybridization to the probe. The chromosomal DNA was digested with MboII, cutting at either end of the sequence obtained from the vanE ligase gene. When this digest was run on a gel and probed, the positive signal moved from the top of the gel (undigested DNA) to two bands of approximately 1 kb and 500 bp in size (Figure 4). The 500 bp band is about the expected size of the digested piece determined from the known sequence. The 1 kb band most likely contains the part of the gene adjoining the sequenced portion.

98 Figure 4. Southern blot of digested and undigested B. popilliae chromosomal DNA probed with the vanE PCR product. Lane 1, 1 kb DNA ladder; Lane 2, undigested ATCC 14706 chromosomal DNA; Lane 3, ATCC 14706 chromosomal DNA digested with MboII; Lane 4, vanE PCR product, Lane 5, undigested ATCC 14707 chromosomal DNA; Lane 6, ATCC 14707 chromosomal DNA digested with MboII.

99 Discussion

I have identified a gene resembling vanA and vanB in B. popilliae. This represents the first detection of vanA- and vanB-like genes in an organism other than an enterococcus, where transmission of the gene from an enterococcus was not suspected. Bacillus popilliae ATCC 14706 is an ATCC type strain which was isolated from commercial insecticidal spore dust, and first described in the medical literature in 1961 (4). Furthermore, I was able to amplify the putative ligase gene from an isolate (Bp23) held in dried hemolymph since 1945. There is therefore compelling evidence that the ligase gene present in B. popilliae was not transferred to this organism from an enterococcus (high level vancomycin resistance in enterococci was only described in the late 1980’s). The putative ligase gene present in B. popilliae has homology to both the vanA and vanB genes raising the possibility that it may have been an ancestor to the vanA and vanB genes found in modern clinical isolates of enterococci. Alternatively, the van genes in enterococci and the putative ligase gene in B. popilliae may have had a common ancestor or ancestors. The ligase gene is most likely located on the chromosome of B. popilliae, possibly on a conjugative chromosomal element like that found in enterococci with the VanB phenotype. In addition, the mechanism of resistance in B. popilliae may be similar to that found in the enterococci, involving a change from D-alanyl-D-alanine to D-alanyl-D-lactate in the peptidoglycan. This can be predicted from the similarity found between the vanE ligase and the vanA and vanB ligases.

B. popilliae spores have been introduced into turf in the Eastern United States as a biopesticidal powder since the late

100 1930’s. As an example, in a 14-year period, between 1939 and 1952, approximately 83,600 kg of B. popilliae spore powder, containing a concentration of 1 ´ 108 spores per g, was applied to 194,000 different sites in 14 eastern states (US) and the District of Columbia and to a total of more than 42,000 ha (8). Commercial production of spore powder began in the mid-1940’s, and continues today (8). It has been suggested that spread of B. popilliae spores may have been increased by birds, insects, skunks, moles and mice (13). Such widespread distribution of this organism may have provided the opportunity for its contact with enterococci. In the presence of the increasing use of oral and parenteral vancomycin in humans since the late 1970’s for the treatment of Clostridium difficile and methicillin-resistant staphylococcal infections, respectively, this transfer would potentially have been facilitated. The use of B. popilliae biopesticidal preparations in agricultural practice may have had an impact on bacterial resistance in human pathogens.

References

1. Elisha, B. G. and P. Courvalin. 1995. Analysis of genes encoding D-alanine:D-alanine ligase-related enzymes in Leuconostoc mesenteroides and Lactobacillus spp. Gene 152:79- 83. 2. Fleming, W. E. 1968. Biological control of the Japanese beetle. U. S. Department of Agriculture Technical Bulletin, Vol. 1383. U. S. Department of Agriculture, Washington, D. C. 3. Gerhardt, P., R. G. E. Murray, W. A. Wood and N. R. Krieg. 1994. Methods for general and molecular microbiology. American Society for Microbiology, Washington, D. C.

101 4. Haynes, W., G. St. Julian, M. Shekelton, H. Hall and H. Tashiro. 1961. Preservation of infectious milky disease bacteria by lyophilization. J. Insect Pathol. 3:55-61. 5. Klein, M. G. 1988. Pest management of soil-inhabiting insects with microorganisms. Agric. Ecosyst. Environ. 24:337-349. 6. Leclercq, R., E. Derlot, J. Duval and P. Courvalin. 1988. Plasmid-mediated resistance to vancomycin and teicoplanin in Enterococcus faecium. New Engl. J. Med. 319:157-161. 7. Noble, W. C., Z. Virani and R. G. Cree. 1992. Co-transfer of vancomycin and other resistance genes from Enterococcus faecalis NCTC 12201 to Staphylococcus aureus. FEMS Microbial. Lett. 72:195-198. 8. Obenchain, F. D. and B. J. Ellis. 1990. Safety considerations in the use of Bacillus popilliae, the milky disease pathogen of Scarabaeidae, pp. 189-201. In M. Laird, E. Lacey and E. Davidson (ed), Safety of microbial insecticides. CRC Press, Boca Raton, FL. 9. Patel, R., J. R. Uhl, P. Kohner, M. K. Hopkins and F. R. Cockerill. 1997. Multiplex polymerase chain reaction detection of vanA, vanB, vanC-1 and vanC-2/3 genes in enterococci. J. Clin. Microbiol. 35:703-707. 10. Patel, R., J. R. Uhl, P. Kohner, M. K. Hopkins, J. M. Steckelberg, B. Kline and F. R. Cockerill. 1998. DNA sequence variation within vanA, vanB, vanC-1 and vanC-2/3 genes of clinical Enterococcus spp. isolates. Antimicrob. Agent. Chemother. 42:202-205. 11. Pridham, T. G., H. H. Hall and R. W. Jackson. 1965. Effects of antimicrobial agents on the milky disease bacteria Bacillus popilliae and Bacillus lentimorbus. Appl. Microbiol. 13:1000-1004. 12. Rippere, K. E., M. T. Tran, A. A. Yousten, K. Hilu and M. Klein. Bacillus popilliae and Bacillus lentimorbus,

102 bacteria causing milky disease in Japanese beetles and related scarab larvae. Int. J. Syst. Bacteriol. In press. 13. St. Julian, G. and L. A. Bulla. 1973. Milky Disease, pp. 57-87. In T. C. Cheng (ed) Current topics in comparative pathobiology. Academic Press Inc. 14. Stahly, D. P., D. M. Takeman, C. A. Livasy and D. W. Dingman. Selective medium for quantitation of Bacillus popilliae in soil and in commercial powders. Appl. Environ. Microbiol. 58:740-743. 15. Zhang, J., T. C. Hodgman, L. Krieger, W. Schnetter and H. U. Schairer. 1997. Cloning and analysis of the first cry gene from Bacillus popilliae. J. Bacteriol. 179:4336-4341.

103 CHAPTER SEVEN

DNA Similarities among Mosquito-Pathogenic and Nonpathogenic Strains of Bacillus sphaericus.

Abstract

Bacillus sphaericus strains isolated on the basis of pathogenicity for mosquito larvae and strains isolated on the basis of a reaction with a B. sphaericus DNA homology group IIA 16S rRNA probe were analyzed for DNA similarity. All of the pathogens belonged to homology group IIA, but this group also contained nonpathogens. It appears inappropriate to designate this homology group a species based solely upon pathogenicity.

Aerobic bacilli that form shperical endospores are common in soil and water and are usually classified as Bacillus sphaericus. There are few useful phenotypic tests for identification of these bacteria. Spore morphology combined with negative reactions in tests for fermentation products and extracellular enzymes have been the basis for taxonomic placement. The species was found to be comprised of at least five distinct homology groups, each sufficiently separated from the others to merit species status (5). Representative strains of the homology groups have also been examined by rRNA gene restriction fragment length polymorphisms analyses (ribotyping), and these analyses confirmed that there are distinct groups within the B. sphaericus complex (2). Recently, randomly amplified polymorphic DNA analysis has also clearly distinguished the groups originally identified by DNA similarity analysis (9). These five groups have not been designated

104 separate species because of the lack of readily utilizable phenotypic tests to distinguish them.

In the original study of Krych et al. (5), group II was divided into two subgroups based on levels of DNA similarity and DNA heteroduplex stability. It was of considerable interest that all of the isolates in group IIA were pathogenic for mosquito larvae. No mosquito pathogens were found in any other group. These bacteria are pathogenic because they produce one or more of four toxins, a binary toxin composed of two distinct proteins and three additional toxins designated Mtx, Mtx2, and Mtx3 (6-8), Strains that produce the binary toxin are highly toxic (50% lethal concentrations, around 102 to 103 cells ml-1), and strains that produce only toxins Mtx, Mtx2, and Mtx3 have low toxicity (50% lethal concentrations, about 105 to 107 cells ml-1). It appeared that the group IIA mosquito pathogens might be designated a separate species. However, only seven pathogenic isolates were available at the time of the original DNA similarity study. Now, many more pathogenic isolates from many geographic locations are available, and although they have been referred to as group IIA strains on the basis of ribotyping data, DNA similarity studies have never actually been performed with them. In this paper we report DNA similarity results for a large number of strains from diverse geographic locations.

Bacteria and DNA isolation.

The strains of B. sphaericus used in this study are listed in Table 1 . The bacteria were grown in NY broth (Difco nutrient broth supplemented with 0.05% yeast extract) at 30oC with shaking at 150 rpm. Cells were recovered by centrifugation, suspended in 20 ml of pH 8.0 buffer (10 mM Tris,

105 1.0 mM EDTA, 0.35 M sucrose, 0.1 mg of lysozyme per ml), and incubated at 37oC for 30 min. A 20-ml portion of lysing solution (100 mM Tris, 20 mM EDTA, 0.3 M NaCl, 2% [wt/vol] sodium dodecyl sulfate, 2% [vol/vol] b-mercaptoethanol, 100 mg of proteinase K Table 1 . Bacillus sphaericus isolates examined by DNA reassociation.

Refer to Rippere et al. 1997. DNA similarities among mosquito- pathogenic and nonpathogenic strains of Bacillus sphaericus. Int. J. Syst. Bacteriol. 47:214-216.

per ml) was added to each preparation, and the mixture was incubated at 55oC for 1 h. Protein was removed by multiple phenol-chloroform extractions, and DNA was precipitated with 0.6 volume of isopropanol. The DNA was dried and suspended in 20 ml of TE, 250 ml of an RNase solution (1 mg of RNase A per ml, 4,000

U of RNase T1 per ml) was added, and the preparation was incubated 1 h at 37oC. The DNA was chloroform extracted and precipitated with ethanol. The precipitated DNA was dissolved in 3 ml of TE and frozen.

DNA similarities

DNA was sheared in a French pressure cell and labeled with 125I, and a hybridization analysis was performed by using the S1 nuclease method (4). DNA samples were heated for 5 min at 60oC before they were used. Reaction tubes containing 10 ml of labeled DNA (0.4 mg/ml), 50 ml of unlabeled DNA (0.4 mg/ml), and 50 ml of buffer (13.2´ SSC, 5 mM HEPES; pH 7.0 [1´ SSC is 0.15 M NaCl plus 0.015 M sodium citrate]) were incubated at 60oC for 24 h to allow reassociation. Following this incubation, 1 ml of

106 buffer (0.3 M NaCl, 0.05 M acetic acid, 0.5 mM ZnCl2), 100 U S1 nuclease, and 50 ml of denatured salmon sperm DNA (0.4 mg/ml) were added to each reaction mixture, and the mixture was incubated for 1 h at 50oC. Then 0.5 ml of HCl buffer (1 M HCl,

1% Na4P2O7, 1% NaH2PO4) and 50 ml of native salmon sperm DNA (1.2 mg/ml) were added to the reaction mixture, and the preparation was incubated for 1 h at 4oC to precipitate the DNA. The precipitated DNA was collected on Whatman glass fiber filters and counted with a gamma counter.

Results and Discussion

The strains used as reference strains for the homology groups were the same as those used in the study of Krych et al. (5). An additional strain, strain Gt1-a, was labeled and also used as a reference. Also, six of the seven pathogenic strains included in group IIA in the original study were analyzed again. A total of 27 additional mosquito pathogen from a variety of geographic locations were included in the study. These pathogens had been isolated on the basis of their ability to kill mosquito larvae. Each of these isolates, regardless of its level of toxicity, was found to be a member of homology group IIA (Table 2). This suggests that the four genes that have been identified as being responsible for toxicity in these bacteria have not been transferred beyond this genetically defined group. As long as isolations were made on the basis of mosquito pathogenicity, it appeared that homology group IIA might contain only these distinctive pathogenic bacteria.

Jahnz et al. (3) utilized an oligonucleotide probe based on a specific region of 16S rRNA from group IIA strains (1) to

107 isolate group IIA strains not on the basis of pathogenicity but on the basis of membership in homology group IIA. These authors recovered 20 strains from Brazilian soil that produced ribotype and isozyme patterns typical of group IIA. However, these strains lacked mosquito pathogenicity, and probes for the binary toxin and Mtx toxin revealed that the genes for these toxins were absent. We included five of these strains in this study (strains G4a, Gt1-a, Gt1-d, R1e, and R4a) and utilized Gt1-a as a labeled reference strain. The high levels of homology of these strains to 1593, the group IIA reference strain, leaves no doubt that they are in fact members of homology group IIA. In addition, the group IIA pathogens exhibited high levels of homology to Gt1-a. Therefore, it appears that although all of the pathogens belong to homology group IIA, this homology group also contains nonpathogens. It is interesting that Jahnz et al. (3) recovered only nonpathogens when they used their probe. These authors suggested that the nonpathogens may, in fact be more common in soil than the homology group IIA pathogens. Whether a pathogen or a nonpathogen is isolated may simply depend on the method used for selection (i.e., pathogenicity or response to the group IIA probe).

Table 2. Levels of DNA similarity among strains of B. sphaericus.

Refer to Rippere et al. 1997. DNA similarities among mosquito- pathogenic and nonpathogenic strains of Bacillus sphaericus. Int. J. Syst. Bacteriol. 47:214-216.

In view of this, it does not seem appropriate to utilize mosquito pathogenicity as the sole characteristic for defining a new species based on homology group IIA.

108 References

1. Aquino de Muro, M., and F. Priest. 1994. A colony hybridization procedure for the identification of mosquitocidal strains of Bacillus sphaericus on isolation plates. J. Invertebr. Pathol. 63:310-313. 2. Aquino de Muro, M., W. Mitchell, and F. Priest. 1992. Differentiation of mosquito-pathogenic strains of Bacillus sphaericus from non-toxic varieties by ribosomal rRNA gene restriction patterns. J. Gen. Microbiol. 138:1159-1166. 3. Jahnz, U., A. Fitch, and F. Priest. 1996. Evaluation of an rRNA-targeted oligonucleotide probe for the detection of mosquitocidal strains of Bacillus sphaericus in soils; characterization of novel strains lacking toxin genes. FEMS Microbiol. Ecol. 20:91-99. 4. Johnson, J. L. 1994. Similarity analysis of DNAs, p. 655- 682. In P. Gerhardt, R. G. E. Murray, W. A. Wood, and N. R. Krieg (ed.), Methods for general and molecular bacteriology. ASM Press, Washington, D. C. 5. Krych, V., J. Johnson, and A. Yousten. 1980. Deoxyribonucleotide acid homologies among strains of Bacillus sphaericus. Int. J. Syst. Bacteriol. 30:476-484. 6. Liu, J.-W., A. Porter, B. Y. Wee, and T. Thanabalu. 1996. New gene from nine Bacillus sphaericus strains encoding highly conserved 35.8 kilodalton mosquitocidal toxins. Appl. Environ. Microbiol. 62:2174-2176. 7. Porter, A., E. Davidson, and J.-W. Liu. 1993. Mosquito toxins of bacilli and their genetic manipulation for effective biological control of mosquitoes. Microbiol. Rev. 57:838-861. 8. Thanabalu, T., and A. Porter. 1996. A Bacillus sphaericus gene encoding a novel type of mosquitocidal toxin of 31.8 kDa. Gene 170:85-89.

109 9. Woodburn, M. A., A. Yousten, and K. Hilu. 1995. Random amplified polymorphic DNA fingerprinting of mosquito-pathogenic and nonpathogenic strains of Bacillus sphaericus. Int. J. Syst. Bacteriol. 45:212-217.

110 SUMMARY

Bacillus popilliae and B. lentimorbus, pathogens of the Japanese beetle, have been differentiated by the production of a parasporal crystal at the time of sporulation and the ability to grow in the presence of 2% NaCl in B. popilliae and the lack of these characteristics in B. lentimorbus. Many different classification systems have been proposed for these bacteria, but the classification has not been studied using molecular techniques. Bacillus popilliae and B. lentimorbus were examined using both DNA similarity studies and randomly amplified polymorphic DNA (RAPD) analysis. Thirty-four isolates of B. popilliae and B. lentimorbus were examined for DNA similarity using the S1 nuclease method. Three distinct but related similarity groups were identified; the first contained strains of B. popilliae, the second contained strains of B. lentimorbus, and the third contained two strains (NRRL B-4081 and Bp3) distinct from but related to B. popilliae. Twenty-five isolates were identified as B. popilliae, while 19 isolates were identified as B. lentimorbus. Two B. popilliae isolates (NRRL B-2522 and BlPj1) were originally received as B. lentimorbus while seven B. lentimorbus isolates (DGB1, Bp1, Bp7, BpCb1, BpCb2, BpPa1 and BpCp1) were originally received as B. popilliae. These seven strains of B. lentimorbus produce a paraspore during sporulation, possibly leading to their original misidentification. All B. popilliae isolates with the exception of NRRL B-4081 were positive for growth when tested using a combination of broth and plates supplemented with 2 % NaCl. The B. lentimorbus isolates (except KLN2, which was positive) were all negative for growth in 2 % NaCl. When tested for vancomycin

110 resistance, all of the B. popilliae isolates were positive and all of the B. lentimorbus isolates were negative.

Nineteen milky disease isolates from various geographic regions were subjected to RAPD analysis. It was hypothesized that due to their diversity, these strains might reveal new subgroups of B. popilliae and B. lentimorbus. Included in this analysis were seventeen strains that had been previously analyzed by M. Tran. Ten new B. popilliae and nine new B. lentimorbus isolates were identified, but there were no new subgroups identified for either species. Patterns relating groups to either the geographic region or host insect were not identified. All of the B. lentimorbus strains were negative for growth in 2 % NaCl, however four B. popilliae strains were also negative (Bp12, BpF, RM23 and RM 29). When tested for vancomycin resistance, 16 (Bp22, RM23 and RM29) B. popilliae strains were positive and 3 (Bp16, Bp17 and Bp25) B. lentimorbus strains were also positive. Seven B. lentimorbus isolates (Bp15, Bp16, Bp17, Bp18, Bp19, Bp25 and Bp26) were most closely related to the crystal-forming subgroup identified in the DNA similarity study. These isolates did form a crystal during sporulation as detected through microscopic examination of a hemolymph smear. DNA similarity and RAPD analysis of B. popilliae and B. lentimorbus has validated the existence of the two species originally identified by Dutky (1940).

Bacillus popilliae and Bacillus lentimorbus isolates were examined for the presence of the cry and van genes. The paraspore is only detectable in sporulated cells of B. popilliae and B. lentimorbus. This limits the use of this characteristic in identification to those laboratories capable of infecting insect larvae with the bacteria. The design of a rapid assay to

111 detect the cry gene enables all laboratories with access to a thermalcycler to use paraspore production in the identification of these species. A PCR assay designed to amplify the cry gene detected the gene in 31 of 35 B. popilliae isolates and in only 1 of 18 B. lentimorbus isolates. This assay is effective at detecting the cry gene in B. popilliae but not in B. lentimorbus. Transferable vancomycin resistance is an emerging problem in clinical strains of enterococci. B. popilliae was shown to be vancomycin resistant by Stahly et. al. (1992), but the mechanism of resistance was unknown. A PCR-RFLP assay designed to detect the van ligase genes in enterococci was used to detect a gene in B. popilliae that is related to the enterococcal van ligase genes. The sequence of the "vanE" gene in B. popilliae had 76.8 % and 68.4 % nucleotide identity to the vanA and vanB genes. The vanE gene is located either on a large plasmid or on the chromosome of B. popilliae. This gene is predicted to be part of an operon responsible for vancomycin resistance in B. popilliae. It is yet to be determined if vancomycin resistance in B. popilliae is transferable, but this is likely due to the similarity of the vanE gene to both the vanA and vanB genes.

DNA similarity analysis was used to examine the classification of 34 B. sphaericus isolates pathogenic for mosquitoes and 5 non-pathogenic B. sphaericus isolates identified by a 16S rRNA probe. All of the isolates were members of the B. sphaericus homology group IIA. As the non- pathogens were also included in group IIA, it appears to be inappropriate to designate group IIA a species based only upon pathogenicity.

112 CONCLUSIONS

1. Most strains of B. popilliae will grow in the presence of 2

% NaCl. In contrast, most strains of B. lentimorbus will not

grow in the presence of 2 % NaCl.

2. Parasporal bodies are present in both B. popilliae and in a

subgroup of B. lentimorbus. Therefore, paraspore formation can

no longer be used as a reliable means to distinguish between the

species.

3. Most strains of B. popilliae are resistant to the antibiotic

vancomycin while most strains of B. lentimorbus are sensitive to

vancomycin.

4. Subgroups of strains were identified among the B. popilliae

isolates studied. There were no apparent relationships between

these strains and the insect from which they were isolated or

between the strains and their geographic origin.

5. PCR primers based upon the published cry18Aa1 nucleotide

sequence (Zhang et al. 1997) detected the presence of the gene

in most strains of B. popilliae known to produce parasporal

inclusions. These primers did not detect the gene in B.

lentimorbus isolates known to produce paraspores. The amount by

113 which the B. lentimorbus parasporal gene differs from the B.

popilliae paraspore gene is unknown.

6. A gene was identified and sequenced in B. popilliae that is

related to the vanA and vanB ligase genes in enterococci. This

gene has been designated vanE and encodes a ligase putatively

involved in vancomycin resistance in B. popilliae.

7. The vanE gene in B. popilliae has been localized to either the chromosome or a large plasmid.

8. All B. sphaericus mosquito pathogens examined to date belong to DNA similarity group IIA. A few non-pathogens isolated on the basis of 16S rRNA similarity also belong to this group. It may be premature to give species status to this DNA similarity group based solely upon mosquito pathogenicity.

114 Karen Elaine Rippere Lampe

20712 Crystal Hill Circle #G Germantown, MD 20874 301-972-2708

EDUCATION

Doctor of Philosophy, Microbiology, September, 1998 Virginia Polytechnic Institute and State University, Blacksburg, VA Dissertation: Systematics of the entomopathogenic bacteria Bacillus popilliae, Bacillus lentimorbus and Bacillus sphaericus. Project includes: DNA:DNA similarity analysis, phenotypic analysis, plasmid curing, Southern blots, hybridizations, nonradioactive detection of probes, RAPD analysis, DNA sequencing, high performance liquid chromatography, polymerase chain reaction, multiplex PCR-RFLP, SDS PAGE, Sephadex gel filtration column chromatography Major Advisor: Dr. Allan Yousten, Professor of Microbiology

Bachelor of Science,Biology, December 1993 Virginia Polytechnic Institute and State University, Blacksburg, VA Basic biological principles and microbiological techniques

Secondary Education South River High School, June 1990

PROFESSIONAL EXPERIENCE

Teaching Graduate Teaching Assistant: Laboratory Instructor, Department of Biology Virginia Polytechnic Institute and State University, Blacksburg, VA January 1995-present Taught laboratory sections in General Biology, General Microbiology, Aquatic Microbiology, Microbial Physiology, Pathogenic Bacteriology and Molecular Plant Systematics

Teaching Experience: Spring 1998 Molecular Plant Systematics laboratory Fall 1996 & 1997 Pathogenic Bacteriology laboratory Spring 1997 General Microbiology laboratory Spring 1996 Aquatic Microbiology laboratory Fall 1995 Microbial Physiology laboratory Spring 1995 General Biology laboratory

115 Related Experience Graduate Research Assistant, Biology Department Virginia Polytechnic Institute and State University, Blacksburg, VA Spring 1995, Summer 1995, 1996 & 1997 Designed and performed experiments related to dissertation

Laboratory Technician Reichardt Animal Hospital, Edgewater, MD January 1994-August 1994 Performed clinical diagnostic tests

Data Entry Clerk, Drug Listing Branch Food and Drug Administration, Rockville, MD Data entry into computers, streamlined drug listing process

PUBLICATIONS

Rippere, K. E., R. Patel, J. R. Uhl, K. Piper, J. M. Steckelberg, B. Kline, F. R. Cockerill and A. A. Yousten. A Gene Resembling vanA and vanB in the Vancomycin-Resistant Biopesticide Bacillus popilliae. J. Infect. Dis. 178:584- 588.

Alban, P. S., D. L. Popham, K. E. Rippere and N. R. Krieg. Identification of a Gene for a Rubrerythrin/Nigerythrin-Like Protein in Spirillum volutans by Using Amino Acid Sequence Data from Mass Spectrometry and NH2-terminal Sequencing. J. Appl. Microbiol. In press.

Rippere, K. E., M. T. Tran, A. A. Yousten, K. H. Hilu and M. G. Klein. Molecular Systematics of Bacillus popilliae and Bacillus lentimorbus, Bacteria causing Milky Disease in Japanese Beetles and Related Scarab Larvae. Int. J. Syst. Bacteriol. 48:395-402.

Yousten, A. A. and K. E. Rippere. 1997. DNA Similarity Analysis of a Putative Ancient Bacterial Isolate Obtained From Amber. FEMS Lett. 152:345-347.

Rippere, K. E., J. L. Johnson and A. A. Yousten. 1997. DNA Similarities among Mosquito-pathogenic and Nonpathogenic Strains of Bacillus sphaericus. Int. J. Syst. Bacteriol. 47:214-216.

Pettersson, B., K. E. Rippere, A. A. Yousten and F. G. Priest. Transfer of Bacillus lentimorbus and Bacillus popilliae to the Genus Paenibacillus with Descriptions of Paenibacillus lentimorbus comb. nov. and Paenibacillus popilliae comb. nov. In review.

116 ABSTRACTS & PRESENTATIONS

Rippere, K. E. and A. A. Yousten. Studies of the Beetle Pathogens, Bacillus popilliae and Bacillus lentimorbus. Presented at 6o Simposio de Controle Biologico, May 1998.

Rippere, K. E, R. Patel and A. A. Yousten. A Gene Resembling vanA and vanB in the Biopesticide Bacillus popilliae. Presented at the national American Society for Microbiology meeting, May 1998.

Rippere, K. E., M. T. Tran, K. Hilu, M. Klein and A. A. Yousten. Molecular Systematics of Milky Disease Bacteria. Presented at the Society for Invertebrate Pathology meeting, Aug. 1997.

Rippere, K. E., J. L. Johnson and A. A. Yousten. DNA Homologies Among Strains of Milky Disease Bacteria. Presented at the national American Society for Microbiology meeting, May 1996

Lampe, R. C., K. E. Rippere, J. L. Johnson, T. Phelps and R. E. Benoit. Characterization of a Deep Subsurface Microaerophile Using 16S rRNA Sequencing and DNA DNA Reassociation. Presented at the national American Society for Microbiology meeting, May 1996

Rippere, K. E. DNA Homologies Among Strains of Milky Disease Bacteria. Presented at theVirginia Branch ASM meeting Dec., 1995

HONORS & AWARDS

1998 Sigma Xi Research Grant awarded 1997 Sigma Xi Research Grant awarded 1996 Phi Kappa Phi Honor Society

PROFESSIONAL MEMBERSHIPS

American Society for Microbiology American Society for Microbiology (Virginia Chapter) Society for Invertebrate Pathology Sigma Xi

REFERENCES

Available upon request

117