iii A MICROBIOLOGICAL STUDY OF ERWINIA AMYLOVORA EXOPOLYSACCHARIDE OOZE By Suzanne Marie Slack A THESIS Submitted to Michigan State University in partial fulfillment of the requirements for the degree of Plant Pathology Master of Science 2016 ABSTRACT A MICROBIOLOGICAL STUDY OF ERWINIA AMYLOVORA EXOPOLYSACCHARIDE OOZE By Suzanne Marie Slack Fire blight, caused by the pathogen Erwinia amylovora (Burrill) Winslow et al, is the most devastating b acterial disease of pome fruits around the world . The primary dispersal of E. amylovora is through ooze, a mass of exopolysaccharid es and bacterial cells that is exuded from infected host tissue. Over the 2013 and 2014 field season s , 631 ooze droplets (201 in 2013 and 435 in 2014) were collected from field inoculated trees. Populations of E . amylovora in ooze drop let s range from 10 8 to 10 11 colony forming units per micro liter (cfu/µl) . I n the host tissue surrounding the droplets e ven larger populations of E. amylovora reside in the surrounding 1 cm of tissue . Three apple cultivars with varying levels of resistance were also infected with four Michigan E. amylovora strains. U sing scanning electron microscopy , host tissue was examined for the origin of the ooze droplets and erumpent mounds and small (10 µm) tears were the o nly bacterial sources observed. Genetic expression analysis indicated that E. amylovora cells in stem sections located above ooze drops and in ooze drops were actively expressing virulence genes . I f disseminated to susceptible host tissue , these cells would be primed for infection. The c urrent study suggests the following: high populations of E. amylovora are present in ooze droplets which larger populations found in darker pigmented, smaller volume droplets. These droplets are rupturing out from the parenchyma and epidermis of the host, with evidence of immense pressure being involved from SEM observations. Ooze droplet volume and population can vary between host cultivar and the virulence of a specific E. amylovora strain. Genetic expression analysis of virulent factors in E. amylovora i ndicated that the bacteria in ooze were primed and ready to infect a new susceptible host. iii ACKNOWLEDGMENTS I would like to thank my advisor Dr. George Su ndin, who put up with me for two years. I would like to thank Dr. Tyre Proffer and Dr. Ray Hammerschmidt for being on my committee. I would like to thank Dr. Quan Zeng, who without him the genetic aspects of this project would not have been accomplished. Finally, I would like to thank my family for being supportive of me and my passions. iv TABLE OF CONTENTS LIST OF TABLES vi i LIST OF FIGURES x i LITERATURE REVIEW OF ERWINIA AMYLOVORA 1 1. Abstract 2 2. Introduction and Historical Importance of Erwinia amylovora 2 2.1 Introduction to ooze - The dispersion method 3 2.2 E. amylovora creates External Ooze Drops 4 2.3 Importance of Ooze in the Disease Cycle 4 3 . E. amylovora Cell Movement in Internal Host Tissue 6 3.1 E. amylovora Internal Movement Hypotheses 7 4 . E. amylovora Escapes from the Host Tissue to Form Ooze Droplets 8 4.1 The Multiplication Hypothesis 8 8 4.3 External forces acting on E. amylovora in the host to produce ooze drops 9 5. E. amylov ora Exits Host Tissue as Aerial Strands besides Ooze Droplets 9 6. E. amylovora Ooze Drops Range in Color 10 7. Weather Factors that may have an Effect on E. amylovora Ooze Drops 10 7.1 Fire Blight Forecasting Systems 11 8. Exopolysacchar ides produced by E. amylovora 11 8.1 Amylovoran 12 8.2 Levansucrase 12 8.3 The role of exopolysaccharides in ooze 12 9. Questions that Merit Further Investigation 13 CHAPTER 1: A MICROBIOLOGICAL EXAMINATION OF EXOPOLYS ACCHARIDE OOZE 15 1. Abstract 16 2. Introduction 17 3. Materials and Methods 2 1 3.1 Shoot blight ooze collection methods 2 1 3.1.1 Ea110 ooze study during the 2013 and 2014 field seasons 21 3.1.2 Stem i noculation of three distinct apple cultivars with four E. amylovora strains during the 2014 field season 22 3.1.3 Internal population samples from apple stems 23 3.2 Blossom Blight 24 3.3. Scanning Electron Microscopy 25 3.4 Analysis of virulence gene expression of E. amylovora in apple stem tissue and ooze 25 3.5 Spectra analysis of ooze and apple tissue pigment colors 26 v 4. Results 26 4.1 Ea110 ooze study during the 2013 and 2014 field se asons 26 4.1.1 Ea110 Population size in ooze droplets 27 4.1.2 Ea110 Ooze Droplet Volume 33 4.2 Internal population samples from apple stems 37 4.3 Stem inoculation of three distinct apple cultivars with four E. amylovora strains during the 2014 field season 40 4.3.1 Differences in population between apple cultivars and E. amylovora strains 40 4.3.2 Differences in ooze droplet volume between apple cultivars and E. amylovora strains 41 4.4 Scanning Electron Microscopy 44 4.5 Analysis of virulence gene expression of E. amylovora in apple stem tissue and ooze 48 4.6 Spectra analysis of ooze and apple tissue pigment colors 51 5. Discussion 52 C HAPTER 2: A FIRST - YEAR REPORT ON THE BIOLOGICIAL CONTROL COMPOUND BLOSSOM - PROTECT IN MICHIGAN ORCHARDS 55 1. Abstract 56 2. Introduction 56 3. Materials and Methods 59 3.1 Laboratory Studies of Blossom - Protect Methods 60 3.2 2014 Field Aspect of Blossom - Protect Methods 61 3.3 Statistical analysis 62 4. Results 62 4.1 Laboratory Studies of Blossom - Protect Results 62 4.2 2014 Field Results 66 4. Discussion 69 CHAPTER 3: AN ATTEMPT AT UV - C MUTAGENESIS OF FUNGAL PATHOGENS OF TREE FRUIT TO CONFER RESISTANCE TO SUCCINATE DEHYDROGENASE INHIBITORS FUNGICIDES 71 1. Abstract 72 2. Introduction 72 2.1 Background on tree fruit fungicides 72 2.1.1 Cherry and Other Stone Fruit 72 2.1.2 Apple 73 2.2 Background of SDHIs 74 3. Materials and Methods 76 3.1 Monilinia fructicola 76 3.1.1 UV - C dose determination for M. fructicola for a 80 - 90% conidial kill curve 76 3.1.2 Creating UV - C Mutant Isolates 77 3.1.3 Screening Mutant Isolates for Fungicide Resistance 78 3.2 V. inaequalis and B. jaapii Mutant Isolation and Screening 78 vi 4. Featured M. fructicola Results 78 5. Discussion 84 WORKS CITED 85 vii LIST OF TABLES Table 1 - 1: Inoculation dates for 2013 and 2014 field seasons along with the E. amylovora strain Ea110 21 Table 1 - 2: Ooze droplets totals collected both in 2013 and 2014 broken down by color 27 Table 1 - 3: E. amylovora Population averages and ranges by color for each the 2013 and 2014 field season 29 Table 1 - 4: 2013 Weather data comparing different weather vari ables (maximum and minimum relative humidity, maximum and minimum air temperature, maximum wind speed, and precipitation ) with ooze droplet population 30 Table 1 - 5: 2014 Weather data comparing different weather variables (maximum and minimum relativ e humidity, maximum and minimum air temperature, maximum wind speed, and prec ipitation) with ooze population 31 Table 1 - 6: Ooze droplet size averages and ranges by color for each the 2013 and 2014 E. amylovora strain Ea110 34 Table 1 - 7: 2013 Weather data comparing different weather variables (maximum and minimum relative humidity, maximum and minimum air temperature, maximum wind speed, and precipitation) with ooze droplet volum e 36 Table 1 - 8: 2014 Weather data comparing different weather variables (maximum and minimum relative humidity, maximum and minimum air temperature, maximum wind speed, and precipi tation) to ooze droplet volume 37 Table 1 - 9: Ooze droplet totals collected from 3 cultivars and 4 strains in 2014 40 Table 1 - 10: 2014 Chart of all insignificant interactions between variables for the 2014 inoculation of four strains onto three cultivars 43 Table 1 - 11: Chart of the four strains from the 2014 field season showing that there was significant difference between strains in regards t o droplet volume and population 44 Table 2 - 1: 2014 field season Blossom - Protect treatment list along with product rates used and the percen t bloom timing of each treatment 60 Table 2 - 2: 2014 field study dates indicating the percent bloom sampled for Blossom - Protect 61 Table 3 - 1: SDHI chemistries available and sorted into chemical group and common name. 75 viii Table 3 - 2: The nu mber of mutants isolated from three M. fructicola isolates (RBOM72, SCHM13, and BR36) from various rounds of UV - C treatments 81 Table 3 - 3: Preliminary screening doses of boscalid and fluxapyroxad resistance in M. fructicola isolate SCHM1 3 and generate d SCHM13 mutants 82 Table 3 - 4: Higher screening doses of boscalid for resistance in M. fructicola isolate SCHM13 and generated SCHM1 3 mutants 83 Table 3 - 5: Higher screening doses of fluxapyroxad for resistance in M. fructicola isolate SCHM 13 and generated SCHM13 mutants 83 ix LIST OF FIGURES Figure 1 - 1: A color scale of E. amylovora ooze droplets seen in the field 22 Figure 1 - 2: Diagram of sections A, B. and C and the 1 cm cuts used in sampling for internal stem populations surrounding the ooze droplets 24 Figure 1 - 3: E. amylovora population in ooze droplets sorted by color for the 2013 field season 29 F igure 1 - 4: E. amylovora population in ooze droplets sorted by color for the 2014 field season 30 Figure 1 - 5: E. amylovora total ooze droplet population per date sampled in 2013 d ivided by color of the droplet 31 Figure 1 - 6: E. amylovora total ooze droplet population by date sampled in 2014 divided by color of droplet 32 Figure 1 - 7: Average Log10 population of E. amylovora by droplet volume for t he 2013 and 2014 field seasons 32 Figure 1 - 8: E. amylovora ooze droplet volume sorted by color for the 2013 field season. 34 Figure 1 - 9: E. amylovora ooze droplet volume sorted by color for the 2014 field season 35 Figure 1 - 10: E. amylovora average ooze droplet volume per date by color in 2013 35 Figure 1 - 11: E. amylovora average ooze droplet volume per sample date divided by color in 2014 36 Figu re 1 - 12: E. amylovora internal popu lations sampled from the field 38 Figure 1 - 13: E. amylovora internal populati ons sampled from growth chamber 38 Figure 1 - 14: E. amylovora average internal population of section B compared to droplets f rom field 39 Fig ure 1 - 15: E. amylovora average internal population of section B compared to droplets from growth chamber 39 Figure 1 - 16: E. amylovora population in ooze droplets compared between strains and cultivars 41 Figure 1 - 17: E. amylovora droplet volume compared between strains and cultivars 42 x Figure 1 - 18: Average E. amylovora ooze drop volume over the twenty - eight day sampling period in 2014 combine d for all cultivars and strains 43 Figure 1 - 19: SEM micrographs of a c ollapsed ooze droplet that formed on a naturally occurring E. amylovora infected immature apple fruit 45 Figure 1 - 20: SEM micrograph zoomed to 450x of a collapsed ooze droplet that formed on a naturally occurring E. amylovora infected immature apple fr uit 46 Figure 1 - 21: SEM micrographs displaying a gradual close up of an exit point by an E. amylovora ooze drop on an inoculated stem 46 Figure 1 - 22: An erumpent mound found underneath a large ooze droplet on a leaf p etiole from an inoculated stem 47 Figure 1 - 23: Various images of dried ooze formed around erumpent mounds 47 Figure 1 - 24: Relative expression of hrpL in different infected plant tissues, ooze drops, and Hrp inducing minimal medium in comparison to expression levels in LB medium 49 Figure 1 - 25: Relative expression of dspE in different infected plant tissues, ooze drops, and Hrp inducing minimal medium in comparison to expression levels in LB medium 49 Figure 1 - 26: Relative expression of lsc in different infected plant tissues, ooze drops, and Hrp inducing minimal medium in comparison to expression levels in LB medium 50 Figure 1 - 27: Relative expression of amsK in different infected plant tissues, ooze drops, and Hrp inducing minimal medium in comparis on to expression levels in LB medium 50 Figure 1 - 28: Spectra of various tissues containing pigment (ooze droplets and orange - tipped fire blight shoots) against controls (uninfected apple stem tissue) 51 Figure 2 - 1: Total bacterial populations fr om flowers sampled in spring 2014 which were collected at the percent bloom in dicated in the horizontal axis 64 Figure 2 - 2: Fungal populations from flowers sampled in spring 2014 which were collected at the percent bloom indicated in the horizontal axis 64 Figure 2 - 3: A. pullulans populations from flowers sampled in spring 2014 which were collected at the percent bloom in dicated in the horizontal axis 65 Figure 2 - 4: E. amylovora populations a t 100% bloom for each treatment 65 Figure 2 - 5: Average pe rcent of blossom prints taken in the field that tested positive for A. pullulans at each sampling percentage 66 xi Figure 2 - 6: 2014 field data for Blossom - Protect treatment trials indicating the percent of infected apple l eaves and fruit with apple scab 67 Figure 2 - 7: 2014 field data for Blossom - Protect treatment trials indicating the percent of infected apple shoot s and blossoms with fire blight 68 Figure 2 - 8: 2014 field data for Blossom - Protect treatment trials indicating the percent of fruit having more than 10% russeting 68 Figure 3 - 1: Survival rates of M. fructicola isolate SCHM13 conidia exposed to various doses of UV - C to determine a kill curve 80 Figure 3 - 2: Survival rates of M. fructicola isolate BR36 conidia exposed to vario us doses of UV - C to determine a kill curve 80 Figure 3 - 3: Preliminary screening of boscalid and fluxapyroxad resistance in M. fructicola isolate SCHM 13 and generated SCHM13 mutants 81 Figure 3 - 4: Higher doses of boscalid and fluxapyroxad resistance in M. fructicola isolate SCHM 13 and generated SCHM13 mutants 82 1 LITE RATURE REVIEW OF ERWINIA AMYLOVORA OOZE 2 1. Abstract Thought to be the main dispersal method of Erwinia amylovora , the causal agent of fire blight, ooze has been neglected in the fire blight literature for many years, as the last fully dedicated paper to ooze was published in 1939 by E.M. Hildebrand (Hildebr and, 1939). Since then, most work directly involving ooze has been based more on the exopolysaccharide characterizati on (Bennett and Billings, 1979) and E. amylovora aerial st r ands (Eden - Green and Billing, 1972). Ooze droplets have been mentioned as a side note in disease dispersal (Eden - Green and Knee, 1974; Bennett and Billing, 1978; Vanneste, 2000) and in internal movement (Schouten, 1989; 1990; 1991). This literature revie w seeks to address the available information on ooze droplets as well as other pe rtinent areas of E. amylovora biology that is related to exopolysaccharide ooze. 2. Introduction and Historical Importance of Erwinia amylovora Erwinia amylovora (Burrill) Winslow et a l., a rod - shaped, gram negative bacterium that is a member of the Enterbacteriaceae family, is the causal agent of fire blight, the most serious bacterial disease present in apples and pears. E. amylovora is thought to be native to North America, infecting hawthorns and other native Rosceae species. C ommercial apples and pears are not native to North America and are typically susceptible to the disease, which can ultimately result in the death of the host (van der Zwet and Keil, 1979; Vanneste, 2000). Since the 1780s, fire blight has been reported along the east coast of the United States and the disease eventuall y spread West with the pioneers through the rest of the continent (van der Zwet and Keil, 1979). On the east coast of the United States fire blight still has a major presence; the disease is so severe in pear tree s that the industry is non - existent (Vanneste, 2000). T he apple industry in the East h owever , is successful, with Michigan, New York, Pennsylvania, and Virginia producing 3 large yields yearly. The main reason apple production is successful is that apples have higher rates of overall resistance to the disease than pears (Vanneste, 2000; Zamski et al., 2006). Even still, apple orchards can be hit with de vastating fire blight epidemics as consumers favor varieties that tend to be mo re susceptible (van der Zewt and Beer, 1995; Vanneste, 2000). In 2000, Southwest Michigan experienced one of these epidemics causing around $42 Million in losses over five years (Longstorth, 2001). Unfortunately, fire blight epidemics are sporadic and most ly unpredictable (Schroth et al, 1974; Vanneste, 2000). 2.1 Introduction to ooze - The dispersion method The main dispersal method of E. amylovora is via exopolysaccharide ooze. Ooze contains E. amylovora that forms external droplets on This slimy secretion consists of bacterial cells and exopolysaccharides (Eden - Green and Knee, 1974; Bennett and Billing, 1979; Vanneste, 2000). However , others describe ooze as containing plant materials as well, thought the paper never define d what plant materials are present (Zamski et al, 2006). In addition to sticky droplets, t he escaped cells may also form thin strings of bacteria called aerial strands (Keil and van der Zwet, 1972; Schouten, 1990). These external masses of tissue can be win d or insect dispersed throughout an orchard, possibly leading to epidemics. Aerial strands and ooze drops once they dry and the E. amylovora cells inside ooze drops can stay viable for at least 25 months (Hildebrand, 1939; Bauske, 1968; Keil and van der Zw et, 1972; Southey and Harper, 1972). In most weather conditions, the ooze dries and hardens in under 24 hours (Slack, observation ). In the event of rain or hail storms, the ooze can be dissolved in the water and spread quickly. It is particularly dangerous to have ooze, fresh or dried, present during hail or wind storms as the damage to the tree allows for quick infection by the bacteria (Keil and van der Zwet, 1972). 4 2.2 E. amylovora creates External Ooze Drops Ooze typically emerges from infected tissue showing symptoms of disease, but also can be observed occurring further from the infection area (Hildebrand, 1939). Aerial strands have a lso been observed originating f r o m ooze droplets (Slack, personal observati on). Though ooze drops offer protection from desiccation (Koczan et al., 2011) and provide bactericide resistance (Hildebrand, 1939; Sutherland, 1988), they do not protect from extreme heat (50 o C) (Hildebrand, 1939). 2.3 Importance of Ooze in the Disease C ycle The primary inoculum for fire blight is thought to come from the overwintering cankers formed from infection during the previous season (Schroth et al., 1974; Vanneste, 2000). In the spring, these cankers begin the disease process with the release of ooze drops (Thomas and Ark, 1934). However , some cankers may not ooze until later in the season and some do not have any viable bacteria present (Schroth et al., 1974; Beer and Norelli, 1977; van der Zwet and Beer, 1995). Because cankers do no t always pro duce ooze at bloom, it is thought that epiphytic E. amylovora could also act as primary inoculum (van der Zwet and Be er, 1995). The bacteria that do form ooze from viable cankers are thought to be dispersed via rain, wind, and insects until enough cells arrive at flowers to infect (Schroth et al., 1974). From the initially infected flower, the bacteria can be spread by pollinating insects to other flowers (Thomas and Ark, 1934; Schroth et al, 1974). Ooze droplets can have populations higher than 1x10 10 c fu/ µ l, which is similar to other pathogenic bacteria (Vieira Lelis et al., 2013). Pollinating insects, in particular bees, are these secondary carriers of cells from infected flowers throughout an orchard (Schroth et al., 1974; Vanneste, 2000). The initial number of infected blossoms needed to create an 5 epidemic can be quite small (Billing, 2011). This is important because there only needs to be a minimum population of 1.04x10 2 CFU/ml on the blossoms for disease to take hold (Schroth et al., 1974). However, once on the stigma of a flower, populations of E. amylovora can reach up to 1x10 6 cfu/ µ l ( Kozan et al., 2009). From those numbers, if there is ample ooze available, the occurrence of an epidemic seems more likely. This stage is known as blossom blight which can seriously affect fruit yield and can establish E. amylovora in an orchard. Secondary infections occur from the flow ers or cankers and spreads to the shoots, causing shoot blight (Vanneste, 2000). During and following bloom, the ooze can appear all over infected blossom clusters and shoots and continue to spread blight through an orchard. Shoot Blight is known best by i , which is caused by the wiling of a shoot (van der Zwet and Beer, 1995; Vanneste, 2000). Bacteria can easily spread from shoot to shoot and quickly over take an orchard with blight, as only 38 cells are needed for injured s hoots to become infected (Crosse et al., 1972; Schroth et al., 1974; Vanneste, 2000). This is partially due to fast growing, succulent shoots, large quantities of ooze, and early summer rains. Mass outbreaks of shoot blight are thought to occur from insects or weather events similar to blossom blight. When E. amylovora becomes internal, entering either the infected flowers or shoots, the bacteria become systemic and eventually make their way down the scion and into the rootstock (Schroth et al., 1974 ; Norelli et al., 2000; Vanneste, 2000). This stage, called rootstock or collar blight, is the most devastating to an orchard , if the disease progresses to the collar blight stage it almost always causes tree death, especially in susceptible cultivars (va n der Zwet and Beer, 1995). Rootstock blight is so devastating that an outbreak can cost up to $2,500 an acre with only a 10% infection rate (Norelli et al., 2000). 6 The last form of fire blight is called trauma blight (van der Zwet and B eer, 1995; Vannest e, 2000). Trama blight is associated with tree injury from external forces, such as storm s or being near a dirt road, that cause damage to the tree (Schroth et al., 1974; Vanneste, 2000). Trauma blight occurring from a storm can be considered the most dama ging, since the w indblown r ain disperses bacteria large distances and the trees can be heavily damaged, giving cells easy access to cause infection. All four types of blight can lead to cankers (Vanneste, 2000). The tree can try to block off these cankers using natural defenses, however the cankers can stay active with viable bacteria for long periods of time (Schroth et al., 1974; Vanneste, 2000). When the cankers start to ooze in the spring, the cycle starts all over again. 3. E. amylovora Cell Movement in Internal Host Tissue Understanding how E. amylovora cells moves through a shoot is vital for explaining how ooze is able to escape from inside the host. However, there is still a debate as to the spreading mechanism of cells (Zamski et al., 2006). In t he symptomatic shoot tissue with necrosis present, bacteria can spread as fast as 2.5 cm a day (Momol et al., 1999; Blachinsky, 2003). It is also known that E. amylovora can be isolated internally some distance from symptomatic tissue (van der Zwet, 1969; Vanneste, 2000), which has led to the recommendation of removing infected tissue at least 30 cm beneath the symptoms of fire blight (Vanneste, 2000). This seems like a good recommendation, as ooze can be found as far as 11 cm beneath shoot symptoms (Vannes te, 2000). 7 3.1 E. amylovora Internal Movement Hypotheses There are many hypotheses of how E. amylovora cells move through a host tree; over the last century many theories have been cast, and many have been discarded. However papers still debate on exact ly how E. amylovora moves systemically though the plant. Literature suggests bacteria could move through either the vascular system via the xylem and/or phloem and/or the intracellular spaces (Crosse et al. 1972; Seemuller and Beer, 1977; Huang and Goodman , 1977; Ayers et al, 1979; Schouten, 1990; Boges et al., 1998; Koczan, 2011). The xylem hypothesis is further supported by a scanning electron microscopy study showing E amylovora blocking the xylem (Kozan et al., 2009). This seems to be most accepted theory on how the bacteria cells move though the host (Vanneste, 2000). There is another theory , however , that cells only move through the intracellular space of the cortical parenchyma cells and do not trave l through the vascular system at all (Bachmann, 1913; Tullis, 1929; Rosenburg, 1936; Huang and Goodman, 1976; Zamski et al., 2006). Zamiski et al. suggests that the reason why cells can be found in the xylem are because of embolisms allowing c ells to ente r the spaces. However , other sources suggest that the embolisms are a result of the E. amylovora population or exopolysaccharide build up and eventual cell wall destru ction (Schouten, 1989). However, when these embolisms occur, they render the xylem useles s as they no longer have turgor pressure and could not move bacteria (Zamski et al., 2006). 8 4. E. amylovora Escapes from the Host Tissue to Form Ooze Droplets 4.1 The Multiplication Hypothesis There are also conflicting theories on how the ooze escapes from the host tissue. Some of through natural openings such as lenticels or stomata (Zamski et al., 2006). However, not every lenticel or stomata on infected apple fruit or shoots ooze (Slack et al., unpublished). Soft tissue is more easily torn allowing build up and movement of cells (Fisher, 1959; Schouten 1991). The force behind the buildup of exopolysaccharides and bacterial cells could potentially rupture the epidermis (Schouten, 1989). T he multiplication hypothesis, in which the bacteria seep or burst out of the plant since internally (Schouten, 1990). This theory w as historically popular, but due to the large amount s of exopolysaccharides produced by the bacterial cells this may only happen in some instances (Gooden et al., 1974; Bennett and Billings, 1980; Schouten, 1989). bacteria moves through the host tissue without necessarily relying on the xylem or phloem bacterial multiplication or the absorption of water (by exopolysaccharides, possibly amylovoran but not named in all citations) increases the physical pressure in the intracellular spaces (Schouten, 1989; Vanneste, 2000; Zamski et al., 20 06). Since no cell wall degrading enzymes are produced by E. amylovora , the breaking through cell walls is probably caused by the pressure of 9 the expanding exopolysaccharides (Seemuller and Beer, 1976; Schouten, 1989). There is direct and model evidence th at exopolysaccharides absorb water and cause swelling (Schouten, 1989; Schouten, 1990; Schouten, 1991). This swelling would lead to the bacteria moving through spaces without the host pushing or pulling the bacteria along the vascular system. It is thought that this pressure could be responsible for the exudation of ooze droplets in asymptomatic tissue (Schouten, 1989; Zamski et al., 2006). The pressure hypothesis is dependent on c ertain weather conditions, and is discussed more in the weather section of th is review. This hypothesis might only be true for apples, since instead of oozing, Hawthorne trees form blisters and only when physically ruptured the ooze emerges (Schouten, 1989). 4.3 External forces acting on E. amylovora in the host to produce ooze dro ps Besides internal forces, the application of external substances, such as pesticide oils can also cause bacterial cells to rupture from the epidermis. H owever they are usually seen in the form of aerial stands and not ooze droplets (Bauske, 1968; Eden - Green and Billings, 1972; Keil and van der Zwet, 1972). The coating of the oils on the epidermis may block air flow, causing pressure to build faster, causing the strains to spew out in the thin rup tures that do occur. 5. E. amylovora Exits Host Tissue as Aerial Strands besides Ooze Droplet s Aerial stands are thought to indicate internal pressure and usually appear during initial stages of infection. E. amylovora strand formation may correspond to long internal, longitudinally oriented strands in healthy t issue (Hockenhull, 1974; Wilson et al, 1987). Aerial strands were 80% matrix and 20% cells (Keil and van der Zwet, 1972). Keil and van der Zwet (1972) f ound that E. amylovora isolated from aerial strands were more virulent on wounded trees than unwounded t rees, however it is known that wou nded trees are more susceptible (Crosse et 10 al, 1972; Keil and van der Zwet, 1972) . The aerial stands can be colorless, amber, or brick red from apple (Eden - Green and Billing, 1972). Eden - Green and Billing (1972) Counted E. amylovora populations in strands that were between 1.6x10 7 - 3.2x10 8 cells/mm 3 in dry strands, and fresh aerial strand populations were not counted. 6 . E. amylovora Ooze Drops Range in Color Hildebrand (1939) states colors of ooze that are not white or dark brown are thought to be contaminants of other organisms on pears (Hildebrand, 1939). The colors could possibly be due to contamination of the drops by another bacteria or yeast. Ooze has be en observed in shades of yellow, brown, and red as well as black, green, and other shades of various colors (van der Zwet, 2012). Aerial stands are thought to be amber to dark brick red (Keil and van der Zwet, 1972). 7 . Weather Fact ors that may have an Ef fect on E. amylovora Ooze Drops night, or early morning before sunrise when water potential is increasing. If the day time temperature is high (20 - 30 o C) oozing could increase according to this model. However oozing is rarely observed in the day time or after the plants harden off (Slack, personal observations). At night time there is very high water potential and the pressure and potential for ooze is the highest. (Sch outen, 1991.) Other known weather conditions important for disease incidence are in temperatures under 19 o C (66 o F ) for shoot blight and even 13 o C (55 o F) during the critical blossom period. It can also become widespread in conditions with 70% relative humidity, heavy fog, heavy dew, and high wind speeds (McManus and Jones, 1994; van der Zwet and Beer, 1995). 11 7 .1 Fire Blight Forecasting Systems There is no model that directly uses the presence of ooze or a quantitative analysis to determine epidemic severity. However, there are many programs that monitor certain weather conditions and predict when an outbreak of fire blight may occur. The oldest system is the Mills system, circa 1955. This forecasting system is based on temperature and moisture during bloom (Mills, 1955; Parker et al., 1956; van der Zwet and Beer, 1995). MARYBLYT, developed in Maryland, uses some forecasting events other than weather to determine whether or not there could potentially be a fire blight outbreak. These factors are: blossom or canker symptoms develop, if insect vectors are available, and the daily temperatures (15.6 o C or above) (Steiner and Lightner, 1992 ; van der Zwet and B eer, 1995.) 8. Exopolysaccharides produced by E. amylovora E. amylovora produces an exopolysaccharide (EPS) capsule, formed from amylovoran and levan, around the bacteria cell (Koczan et al., 2009). These two components are found in ooze and are essential to the pathogenicity or virulence of E. amylovora (Oh and Beer, 2005; Koczan et al., 2009) . Mutants without the capability to make any EPS are non - virulent (Koczan et al., 2009). Besides being important in virulence or pathogenicity, the EPS can shield th e bacterium from host defenses and possibly even an tibiotics (Geider et al, 1993). These EPS compounds have been known, albeit not named, for many years as the source of . A compound isolated from external ooze was noted as the fire blight toxin (Hildebrand, 1939). Many studies and papers have noted the toxin, (van der Zwet, 1966; Keil and van der Zwet, 1972; Beer and Norelli, 1974), however the mystery compound 12 search in Science that had identified the changed to amylovoran. 8.1 Amylovoran Amylovoran is a heterogeneous, acidic polysaccharide made up of repeating subunits of galactose molecules all linked to a glucuronic acid residue (Goodman et al, 1974; Bennett and Billing, 1978). This EPS is essential for pathogenicity and the formation of biofilm (Koczan et al, 2009; 2011). Without amylovoran, there is no biofilm produced (Koczan et al., 2011). 8.2 Levansucrase E. amylovora also produces a fructose homopolymer called levan (Gross e al., 1992). Sucrose is needed for the levansucrase to make levan (Gross et al., 1992). This levan EPS is an important virule nce factor as kno ckout mutants have dramatically decrease virulence (Koczan et al, 2011). Of the main components of apple and pear blossom nectar is sucrose, suggesting that the production of levansucrase by the bacteria would intensely aide in procuring enough sugar assi milation (Gross et al., 1992). The levan also aides in expansion as part of the biofilm (Schouten, 1989). 8.3 The role of exopolysaccharides in ooze Besides bacteria, the rest of the ooze droplet is formed from EPS (Bennett and Billing, 1978). Thus, the e xopolysaccharide (EPS) capsule plays a huge role in ooze production. Without EPS, no bacteria would be able to bind together, as EPS is thought to be the basis for biofilms 13 (Koczan et al., 2011). The other major functions of EPS, such as water and nutrient retention, would be vital for ooze production and bacteria viability inside the droplet (Goodman et al., 1974; Sutherland, 1988; Koczan et al., 2011). The binding of water could be important for the internal swelling of the ooze, allowing for the dramatic burst exiting. The nutrient retention could explain the phenomenon of bacteria being viable in ooze droplets for at least a year as well (Hildebrand, 1939). 9. Questions that Merit Further Investigation Research focusing on ooze is minimal , and many questions have not been answered. Though Eden - Green and Billing (1972) counted E. amylovora populations in dry strands, they did not look at ooze drops or fresh aerial strands (Eden - Green and Billing, 1972). From the literature, no papers have p ublished the quantities of E. amylovora present in ooze. There have only been notes of the color differences, no experiments on quantification of the ooze colors or if the colors have anything to do with disease progression have been completed. Do the ooze droplets vary in volume, population, and color? Are there any significant interactions between these three variables, and could they tell us more about E. amylovora ? There is evidence that E. amylovora can build enough pressure to rupture xylem vessels ( Schouten, 1989; Geider et al, 1993), so why can ooze not rupture out of the epidermis? More evidence is needed to determine how the ooze is escaping from the host. There has been work done on the internal population of E. amylovora in the host, but the pop ulations were not quantified, just tracked or confirmed. What are the internal populations in regards to the ooze drops? How does the disease allocate population for dispersal 14 and subsequent new infection of hosts? With new genetic capabilities, advances i n microscopy techniques, bacterial tracking , and biochemical analysis further study on ooze should be conducted to answer the questions the literature review raised as well as look into genetic expression of pathogenicity and virulence factors to see if th e ooze - dwelling E. amylovora are primed for invasion of new host tissues. 15 CHAPTER 1 A MICROBIOLOGICAL EXAMINATION OF EXOPOLYSACCHARIDE OOZE 16 1. Abstract Fire blight, caused by the pathogen Erwinia amylovora (Burrill) Winslow et al, is the most devastating bacterial disease of pome fruits in North America and around the world. The primary dispersal method of E. amylovora is through ooze, a mass of exopolysaccharides and bacterial cells that is exuded from infected host tissue. Over the 2013 and 2014 field season, 631 ooze droplets (201 in 2013 and 435 in 2014) were collected from field inoculated trees. Populations of E. amylovora in ooze drops range from 10 7 to 10 11 colony forming units per micro liter (cfu/µl) . The droplets that had higher populations were typically smaller in total volume and had darker coloring, such as orange, red, or dark red hues. These darker colors may be more attractive to insects that disper se ooze. When the examining host tissue for the origin of the ooze droplets using scanning electron microscopy, no natural openings were discovered in the vicinity of the droplets; erumpent mounds and small (10 µm) tears were the only bacterial sources obs erved. Even though large amounts of bacteria and exopolysaccharide ooze are forcing out of the host parenchyma and epidermis, even larger populations reside in the host. The ooze droplet is at most one third of the population in the surrounding 1 cm of ti ssue, meaning that the pathogen is allotting the most population resources to further infection of the current host instead of spread or survival. Three distinct cultivars with different levels of fire blight resistance were infected with four native Michi gan E. amylovora strains, which indicated that ooze production can vary between host and the virulence of the strain. Genetic expression analysis indicated that E. amylovora cells in stem sections located above ooze drops and in ooze drops were actively expressing virulence genes suggesting that these cells would be primed for infection if disseminated to susceptible host tissue. The current study suggests the following: high populations of E. amylovora are present in ooze droplets which larger population s found in darker pigmented, smaller volume droplets. These droplets are rupturing out from the 17 parenchyma and epidermis of the host, with evidence of immense pressure being involved from SEM observations . Ooze droplet volume and population can vary betwee n host cultivar and the virulence of a specific E. amylovora strain. Genetic expression analysis of virulent factors in E. amylovora indicated that the bacteria in ooze were primed and ready to infect a new susceptible host. 2. Introduction Erwinia amylovora (Burrill) Winslow et al., a rod - shaped, gram negative bacterium that is a member of the Enterbacteriaceae family, is the causal agent of fire blight, the most serious bacterial disease present in apples and pears. E. amylovora is thought to be native to North America, infecting hawthorns and other native Rosaceae species. Since commercial apples and pears are usually susceptible to fire blight, which can ultimately result in the death of the host (van der Zwet and Keil, 1979; Vanneste, 200 0). The disease can spread quite rapidly and cause large amounts of destruction, hence the name fire blight. The main reason E. amylovora can be dispersed so quickly between trees and within or between orchards is because of the ability of cells to extrude to the surface of infected tissue in droplets. Bacterial ooze droplets form outside of the host and can serve as a source for a large amount of bacteria which can be spread through an orchard. Even though ooze droplets are an important epidemiological tra it, no major studies have been conducted on ooze droplets since the 1930s. The last fully dedicated paper to ooze was published in 1939 by E.M. Hildebrand (Hildebrand, 1939). Since then, most work involving ooze has been based more on the exopolysaccharide identification (Bennett and Billings, 1979) and aerial stands (Eden - Green and Billing, 1972). Ooze has been mentioned as a side note in E. amylovora dispersal (Eden - Green and Knee, 1974; Bennett and Billing, 1978; Vanneste, 2000) 18 and in movement (Schouten , 1989; 1990; 1991). Ooze was also mentioned in Blachinsky et al, 2006, however they defined ooze being inside the plant, as opposed to ooze droplets. Ooze is both the primary and a secondary form of inoculum for the transmission of fire blight. The fire blight primary inoculum stage consists of ooze that forms on the surface of over wintering cankers (Schroth et al., 1974; Vanneste, 2000). The ooze from viable cankers is thought to be dispersed via rain, wind, and insects until enough cells build on flowe rs to infect, causing blossom blight (Schroth et al., 1974). This does not bode well as a minimum population of 1.04x10 2 CFU/ml of E. amylovora present on the flowers is needed for blight to occur (Schroth et al., 1974). For an epidemic to occur there onl y needs to be a few flowers infected initially (Billing, 2011). After the initial infection, ooze is readily produced from flower pedicels or shoots. Bacterial cells in the ooze then function as secondary inoculum that can easily spread from shoot to shoot and quickly over take an orchard with blight (Schroth et al., 1974; Vanneste, 2000). This is due to the abundance of fast growing, succulent shoots, large quantities of ooze, and early summe r rains. Shoot blight can be more devastating to the orchard than blossom blight since it can spread faster through a planting (van der Zwet and Beer, 1995; Vanneste, 2000). Mass outbreaks of shoot blight are thought to occur from insect or weather - mediate d events. Ooze consists of bacterial cells, EPS, and possibly plant materials as well, which is discussed later in this chapter. Thus, the EPS capsule plays a main role in ooze production. The main exopolysaccharides present in ooze are amylovoran and levan (Oh and Beer, 2005; Koczan et al., 2009). thought to be the basis for biofilm (Koczan et al., 2011). The other major functions of EPS, such as water and nutrient retention, would be vital for ooze production and bacterial viability inside the droplet (Goodman et al., 1974; Sutherland, 1988; Koczan et al., 2011). The binding of water 19 could be important for the internal swelling of the ooze, allowing for the dramatic burst exi ting. The nutrient retention could also help to explain the phenomenon of bacteria being viable in ooze droplets for at least one year (Hildebrand, 1939). Not much is known about the color of ooze from the literature, other than it can be colorful (Hildebr and, 1939; Keil and van der Zwet, 1972; van der Zwet et al., 2012). The method of escape of ooze from the host tissue has been widely debated in the phytopathology literature for decades (Hildebrand, 1939; Fisher, 1959; Seemuller and Beer, 1976; Billings, 1981; Schouten, 1991; Zamski et al., 2006). There does not seem to be a consensus on whether ooze passively seeps from the plant or bursts out from the parenchyma cell layer. Since there are no cell wall degrading enzymes produced by E. amylovora , the brea king through compartmental walls is probably caused by the pressure of the expanding exopolysaccharides (EPS) (Seemuller and Beer, 1976; Schouten, 1989). There is direct and model evidence that EPS absorb water and cause swelling (Schouten, 1989; Schouten, 1990; Schouten, 1991). This swelling would lead to the bacteria moving through spaces without the host pushing or pulling the bacteria along the vascular system (Schouten, 1988). Weather is always cited as being a major contributing factor to the spread of fire blight, thus it would seem that weather and ooze production must somehow be tied together. According to Schouten (1991), when water potential is high, typically around 20 - 30 o C, oozing could increase. There is also evidence that epidemics occur in conditions with 70% relative humidity, heavy fog, heavy dew, and high wind speeds (McManus and Jones, 1994; van der Zwet and Beer, 1995). Based from the literature, we proposed a microbiological study on E. amylovora ooze to fill some of the gaps abou t the dissemination mechanism. We created a list of seven objectives 20 and with them seven hypotheses based on previous literature. This chapter seeks to answer these questions: Objectives: 1. Determine the size and population of E. amylovora in ooze droplets. 2. Determine if there is a correlation between droplet color, size, and bacterial population. 3. Determine the cause of the differences in ooze droplet color. 4. Determine if there is a correlation between ooze droplets and internal shoot E. amylovora populations. 5. Determine the mechanism of ooze escape from the host tissue. 6. Determine if host cultivar or E. amylovora strain is a factor in ooze population or size. 7. Determine what, if any, weather factors influence the development of ooze. Hypotheses: 1. There is a relatio nship between size and population of ooze droplets. 2. There is a correlation between droplet color, size, and population. 3. The color differences could be an indicator of internal damage. 4. There is a correlation between ooze droplets and internal E. amylovora p opulations. 5. The mechanism of ooze escape from the host tissue is not mainly through natural openings. 6. Host cultivar and E. amylovora strain are factors in ooze population or size. 7. Weather factors influence the development of ooze, especially the temperatur e and relative humidity. 21 3. Materials and Methods 3.1 Shoot blight ooze collection methods 3.1.1 Ea110 ooze study during the 2013 and 2014 field seasons Erwinia amylovora Ea110, a spontaneous rifampicin resistant mutant isolated in Michigan (Zhao et al., 2005), was manually inoculated in an orchard located in East Lansing, Malus x domestica ) grafted onto M9 rootstock. The inocu lum was prepared using cell suspensions grown in Luria - Bertani (LB) broth at 28 o C for 12 hrs. Prior to inoculation populations were adjusted using a Tecan Safire spectrometer to 1x10 6 CFU/ml in 0.5x phosphate buffered saline (PBS). These adjusted cell susp ensions were kept on ice until inoculation, occurring less than an hour after the suspension was prepared. Using sterile scissors dipped into the cell suspension, the growing tips of leaves on healthy shoots were cut horizontally across the midvein, removi ng a quarter to a third of the tip of the leaf (McGhee et al., 2011) . Inoculation occurred weekly throughout the growing season, starting in late May and ending in early July, resulting in seven inoculations in both 2013 and 2014 (Table 1 - 1) . Table 1 - 1: number of shoots inoculated with E. amylovora strain Ea110 . In 2014, three additional E. amylovora strains , EL01, GH9, and K2, were inoculated on the same day as Ea110 on culti vars this cultivar. Inoculation 1 2 3 4 5 6 7 2013 Date 24 May 28 Ma y 4 June 7 June 15 June 20 June 25 June 2013 Shoot No. 1 - 20 21 - 40 41 - 45 46 - 55 46 - 65 66 - 75 76 - 80 2014 Date 22 May 29 May 5 June 10 June 19 June 26 June 30 June 2014 Shoot No. 1 - 20 21 - 40 41 - 60 61 - 70 71 - 80 81 - 90 91 - 95* When present in the early morning, bacterial ooze was collected using sterile .6 ml tubes by scraping the ooze off the branch using the inner lip of the lid cylinder. This scraping insured 22 that the entire ooze drop was in the microcentrifuge tube. Only ooze deemed fres h, not hardened and less than 24 hrs old, w as collected. 100 µ l of 0.5x PBS was added to the 0 .6 ml tube in a sterile lab hood. The ooze drop was vortexted until completel y dissolved. Next, 100 µ l of the solution was removed from the 0 .6 ml tube and added to a 1.5 ml micr ocentrifuge tube containing 900 µ l of 0.5x PBS to be further diluted. The remaining solution in the original .6 ml tube was r emoved using a 10 µ l pipette 1ul at a time until all liquid was removed. This amount was determined to be the original volume of the ooze droplet. The 100 µ l of the original solution was then diluted and the appropriate dilutions were drop plated onto LB medium am ended with cycloheximide at 50 µ g ml - 1 and rifampicin at 100 µ g ml - 1 to inhibit fungal and unwanted bacterial growth respectively. The colony counts were used to determine the original population of the ooze drop. With each ooze drop, the color (Figure 1 - 1) , physical location, date, current weather conditions, and shoot number were collected. In 2014 , the distance from the ooze drop to the bud scar was also collected. Figure 1 - 1: A color scale of E. amylovora ooze droplets seen in the field . From left to right the colors are white, yellow, orange, red, and dark red. This figure represents the sc ale used to define what color category each ooze drop was assigned. 3.1.2 Stem i noculation of three distinct apple cultivars with four E. amylovora strains during the 2014 field season In 2014, three additional E. amylovora strains (K2, GH9, and EL01) isolated from Michigan orchards (McGhee and Sundin, 2012) were tested on three cultivars of apple, M. x 23 domestica, ( Linda Mac , Kit Jon , and September Wonder Fuji ) to determine if there were differences in ooze production compared to the Ea110 strain used previously in 2013. Each strain, including Ea110 , was inoculated, collected, and processed using the methods listed above with the exception that the LB media pla tes were not amended with rifampicin for the new strains. Also, only orange droplets were collected for this study inste ad of every ooze drop produced to allow for a higher population sample with one less variable. 3.1.3 Internal population samples from a pple stems In addition to collecting the ooze drop, 1 cm sections of the tissue surrounding the exuded ooze (labeled A, B, and C) were collected and evaluated for internal populations. The inoculation was the same as for the Ea110 ooze drop collection met hods described previously. When the ooze was collected using the same method listed as above, the surrounding tissue was also collected and brought back to the laboratory immediately for making the sections. The middle section (B) consisted of 0.5 cm on ei ther side of the ooze drop, resulting in the 1 cm section. The top section (A), was the 1 cm piece above section B and the bottom (C) was directly under section B (Figure 1 - 2) . Each piece was weighed and then chopped up using a sterile razor blade. The cho pped stem pieces were placed into glass sonication blanks containing 9 ml of 0.5x PBS. The solution was then homogenized using a PT 10 - 35 polytron (Brinkmann, Westbury, NY) for 3 - 5 seconds (McGhee et al, 2011) . The solution was then plated using methods pr eviously listed for ooze drops to determine population levels . Final population counts were standardized for weight based on the original stem piece. This was repeated twice in a growth chamber, once in 2013 then again in 2014 usin ye G ala In spring/summer 2014 this was again repeated in the field using Ea110 on Kit Jonathans. 24 Figure 1 - 2: Diagram of sections A, B. and C and the 1 cm cuts used in sampling for internal stem populations surrounding the ooze droplets . This method was used in both the growth chamber and in the 2014 field season. 3.2 Blossom B light In Spring 2014, E. amylovora strains Ea110 , K2, GH9, and EL01 (isolated from various Michigan orchards) were tested on different cultivars of apple, M. x domestica, ( Linda Mac , Kit Jon , and September Wonder Fuji ) to determine if there were any differences in ooze production from blossom infection. Populations were adjusted to 1x10 6 cell density in 0.5x PBS as described for shoot inoculum. The suspensions were kept chilled en route to the o rchard and until inoculation. Fifty blossoms per cultivar and strain were inoculated by hand pipetting 5 µ l of inoculum directly onto the stigma of king bloom flowers . A total of 600 king bloom flower s were inoculated. After inoculation, ooze was harvested and processed as previously described in the shoot blight ooze section. Besides ooze; percen t infection, severity of infection, and weather conditions were also recorded. 25 3.3. S canning Electron Microscopy All scanning electron m icroscopy (SEM) samples were processed in the Center for Advance M icroscopy at MSU. Samples were fixed in paraformaldehyde/glutaraldehyde (2.5% of each compound in 0.1 M sodium cacodylate buffer) (Electron Microscopy Sciences, 151 Hatfield, PA). The tissue was ethanol dehydrated, then critical point dried (Balzers CPD, L ichtenstein). Tissue was sliced after critical point drying to reduce potential artifacts from the fixation process and , then mounted on aluminum mounting stubs using carbon tabs (Electron Microscopy Sciences). 3.4 Analysis of virulence gene expression of E. amylovora in apple stem tissue and ooze E. amylovora Ea110 was cultured in LB and inoculated on apple shoots using a previously describ ed dipping scissors method. Seventeen ooze drops were collected from 17 individual apple shoots from four different c ultivars ( Linda M ac , 5; Kit J on , 5; Gala , 3; Fuji , 4), Ooze drops were collected 4 to 6 days post inoculation, on five sampling dates ( 3 June , 10 June , 17 June , 23 June , and 30 June 2014). Early stage infected shoots were collected at 48 hrs post inoculation on 14 June. Section A samples were collected from shoots 1 cm above the ooze drops on the same sampling date of the ooze drops ( 23 June and 30 June). Ea110 was cultured in LB an d induced in Hrp - inducing minimal medium (Hrp MM) (Guttman et al., 2002) for 12 hrs . Total RNA was isolated from the LB culture and the Hrp MM culture as negative and positive con trols for T3SS gene expression. Total RNA was isolated on the same sampling d ates using an E.Z.N.A.Plant RNA Kit (Omega Biotek) and a miRNeasy kit (Qiagen). The quality and quantity of RNA isolated was tested using a Nanodrop1000 (Thermo Scientific) and a bioan alyzer (Agilent Technologies). Reverse transcription was performed usin g TaqMan reverse transcription reagents (Applied biosystems). qRTPCR was 26 performed by a StepOnePlus Real - Time PCR System (Life technologies), using a SYBR Green PCR ma ster mix (Applied biosystems). The PCR amplification specificity was confirmed using a me lting curve method. Housekeeping gene recA was used as an endogenous control. The ized by the endogenous control. Gene expression levels in ooze drops and plant samples were presented as the expression fold in comparison to the expression in Ea110 LB culture. Pri mers used in real time PCR are recA (F: - ATCATTGTTGACTCCGTTGC, R: - CATTGCCTGGCTCATCATAC), hrpL (F: - GATCTGGAGCAAATGACCTG, R: - TTTAAGGCAATGCCAAACAC), dspE (F: - CGCAACATCGGAACCATTAA, R: - TGCGACCTGCGGATTAGC), lsc (F: - ACCAGACGGAAGAGCAGAAC, R: - CACGTTTCCTTCAAACAGCA), and amsK (F: - CGGCACGCTGAAATCATTC, R: - TGCCGCAAAGGGCTTTT). 3.5 Spectra analysis of ooze and apple tissue pigment c olor s Samples of red, orange, and yellow ooze collected from the field in spring 2014, orange - tipped fire blight strikes, healthy apple tissue, and cultured E. amylovora were prepared in a 10% diH 2 O 90% methanol solution and scanned with a Cary 50 UV - Visible spectrometer (Varian, Walnut creek, CA) to gauge the chemical composition of pigments. 4. Results 4.1 Ea110 ooze study during the 2013 and 2014 field seasons In the fire blight field seasons of 2013 and 2014, which lasted from late M ay to early July, a total of 201 and 116 ooze droplets respectively were collected from E. amylovora strain Ea110 - ooze droplets: 15 dark red, 39 red, 91 orange, 29 yellow, and 28 white (Table 1 - 2). The 116 27 collected droplets in 2014 droplet colors were: 9 dark red, 13 red, 44 orange, 22 yellow, and 22 white (Table 1 - 2). These ooze droplets were examined for E. amylovora population, droplet volume, weather event si gnificance, and the internal population remaining in host tissue related to ooze droplet location. Ooze droplets were also observed over time in field to determine if they changed color when drying. There were no observances of the droplets changing colo r. Table 1 - 2: Ooze droplets totals collected both in 2013 and 2014 broken down by color. Year/Color Dark Red Red Orange Yellow White Total for year 2013 15 39 91 44 22 201 2014 9 13 44 22 22 116 Total 24 52 150 66 44 317 4.1.1 Ea110 Population size in ooze droplets E. amylovora populations recovered red ooze droplets were the highest in the 2013 collection year with an average population 6.17x10 10 CFU/µl with a range 2.51x10 9 - 3.72x10 11 CFU/µl. However the dark red ooze droplets were not signific antly different than the dark red (average population was 5.75x10 10 CFU/ µl; r ange was 2.88x10 9 - 1.95x10 11 CFU / µl ) and orange colored droplets (average population of 5.50x10 10 CFU / µl and range 2.24x10 8 - 3.31x10 11 CFU/ µl ) (Table 1 - 3; Figure 1 - 1). In 2014 the red droplets with an average population of 1.86x10 10 CFU / µl (range of 9.55x10 8 - 7.08x10 10 CFU / µl ) were the largest in population size (Table 1 - 2; Figure 1 - 2). However, the red ooze droplets were not significantly dif ferent than the orange (average population was 1.66x10 10 CFU / µl , range was 3.55x10 7 - 1.95x10 11 CFU / µl ) and dark red ( 9.77x10 9 CFU / µl range of 1.48x10 8 - 3.80x10 10 CFU / µl ) (Table 1 - 2; Figure 1 - 2). Even 28 though in both sample years had red, orange and dark red colored droplets had the highest populations, there was a significant difference between the two field seasons (Figures 4.1.1). White and yellow pigmented ooze droplets harbored the lowest populati ons both years (Figures 1 - 1 and 1 - 2). In 2013 white dro plets had an average cell population of 3.24x10 10 CFU /µl and a range from 4.27x10 8 - 1.05x10 11 CFU µl and in 2014 had a population average of 4.37x10 9 CFU /µl and ranged 8.51x10 6 - 5.37x10 10 CFU /µl (Table 1 - 2). Yellow ooze droplets were also on average lower i n population than the other dark shades; in 2013 the population averaged 2.88x10 10 CFU /µl (range 5.37x10 8 - 3.31x10 11 CFU / µl ) and in 2014 averaged 1.92x10 10 CFU / µl (ranged 1.17x10 7 - 7.08x10 10 CFU /µl ) (Table 1 - 2; Figures 1 - 3 and 1 - 4). In 2013 the lighter colors of yellow and white were statistically different than the darker colors. However in 2014 only white was significantly less than red and orange droplets. (Table 1 - 3; Figures 1 - 3 and 1 - 4). When comparing populations of E. amylov ora on a day to day basis during the 2013 and 2014 field seasons there were two statistically occurring groups in 2013 and three groups in 2014 (Figures 1 - 5 and 1 - 6). Not every color of ooze was observed on each sampling date. The populations were differen t between colors on the same sampling date as well (Figures 1 - 5 and 1 - 6). There was not an observed trend in E. amylovora population size in either 2013 or 2014 as the sea son progressed (Figure 1 - 5 and 1 - 6). Weather data was also collected from the Envir o - weather station in East Lansing, MI (MSUHORT). In 2013 the only corresponding significance between any specific date and weather event was minimum air temperature (Table 1 - 4). There was no significant interaction of any weather factor and population size from 2014 (Table 1 - 5). When comparing the average population of the same size ooze droplets between 2013 and 2014, droplet sizes of 1 and 2 microliters had statistically different population sizes (Figure 1 - 7). 29 Table 1 - 3: E. amylovora Population averages and ranges by color for each the 2013 and 2014 field season E. amylovora strain Ea110. Field Season CFU Color of Ooze Droplets Dark Red Red Orange Yellow White 2013 Average 5.75x10 10 6.17x10 10 5.50x10 10 2.88x10 10 3.24x10 10 Range 2.88x10 9 - 1.95x10 11 2.51x10 9 - 3.72x10 11 2.24x10 8 - 3.31x10 11 5.37x10 8 - 3.31x10 11 4.27x10 8 - 1.05x10 11 2014 Average 9.77x10 9 1.86x10 10 1.66x10 10 1.92x10 10 4.37x10 9 Range 1.48x10 8 - 3.80x10 10 9.55x10 8 - 7.08x10 10 3.55x10 7 - 1.95x10 11 1.17x10 7 - 7.08x10 10 8.51x10 6 - 5.37x10 10 Figure 1 - 3: E. amylovora log 10 population in ooze droplets sorted by color for the 2013 field season E. amylovora strain Ea110. Error bars represent standard error, P <.0001 A A A B B 0 1E+10 2E+10 3E+10 4E+10 5E+10 6E+10 7E+10 8E+10 Red Dark red Orange White Yellow Colony Forming units per microliter (CFU/µl) 2013 30 Figure 1 - 4: E. amylovora log 10 population in ooze droplets sorted by color for the 2014 field season E. amylovora strain Ea110. Error bars represent standard error, P <.0001 Table 1 - 4: 2013 Weather data comparing different weather variables (maximum and minimum relative humidity, maximum and minimum air temperature, maximum wind speed, and precipitation) with ooze droplet population. Variables P - value Bacteria population i n ooze and date sampled in relation to max. relative humidity .4051 Bacteria population in ooze and date sampled in relation to min. relative humidity .8326 Bacteria population in ooze and date sampled in relation to maximum air temp. .1423 Bacteria pop ulation in ooze and date sampled in relation to minimum air temp. <.0001 Bacteria population in ooze and date sampled in relation to maximum wind speed .8448 Bacteria population in ooze and date sampled in relation to precipitation .5250 A A AB AB B 0.00E+00 5.00E+09 1.00E+10 1.50E+10 2.00E+10 2.50E+10 3.00E+10 Red Orange Dark Red Yellow White Colony Forming units per microliter (CFU/µl) 2014 31 Table 1 - 5: 2014 Weather data comparing different weather variables (maximum and minimum relative humidity, maximum and minimum air temperature, maximum wind speed, and precipitation) with ooze population. Variables P - value Bacteria population in ooze an d date sampled in relation to max. relative humidity .4493 Bacteria population in ooze and date sampled in relation to maximum air temperature .5730 Bacteria population in ooze and date sampled in relation to minimum air temperature .7134 Bacteria population in ooze and date sampled in relation to maximum wind speed .1700 Bacteria population in ooze and date sampled in relation to precipitation .2913 Figure 1 - 5: E. amylovora total ooze droplet log 10 population per date sampled in 20 13 divided by color of the droplet . There was no observable trend in E. amylovora population averages throughout the apple shoot growing period. All samples were collected on sixteen dates from E. amylovora strain Ea110. 10.5 10.21 10.43 10.56 9.28 8.69 10.6 10.35 10.8 9.98 10.67 9.9 9.51 10.6 9.85 10.39 10.27 11.14 10.78 10.02 10.5 10.48 10.12 10.85 10.57 10.83 9.4 10.37 10.7 10.46 10.48 10.38 9.75 10.35 11.24 10.33 10.67 10.2 10.92 10.74 10.69 10.53 10.66 0 10 20 30 40 50 60 Log 10 Total Colony forming Units per microliter (CFU/µl) Date Sampled Dark Red Orange Red Yellow White 32 Figure 1 - 6: E. amylovora total ooze droplet log 10 population by date sampled in 2014 divided by color of droplet . There was no observable trend in E. amylovora population averages throughout the apple shoot growing period. All samples were collected on twenty - E. amylovora strain Ea110. Figure 1 - 7: Average Log 10 population of E. amylovora by droplet volu me for the 2013 and 2014 field seasons . Over the 2013 and 2014 seasons 317 ooze droplets were collected from field - standard error, P <.0001. Asterisks (*) repr esent significant difference in populations between field seasons. 9.26 10.02 10.13 10.45 9.91 10.25 9.00 8.95 8.49 10.02 9.77 9.09 9.42 10.73 8.83 9.57 8.50 9.49 9.14 9.17 9.13 10.03 8.48 9.37 8.93 8.24 9.01 9.36 9.18 9.87 9.60 10.28 9.41 9.57 9.89 10.03 8.99 10.62 8.31 9.61 9.39 8.95 10.58 9.25 9.64 8.17 10.45 8.91 10.15 9.91 9.18 0 5 10 15 20 25 30 35 40 45 50 5/28/14 5/29/14 5/30/14 5/31/14 6/1/14 6/2/14 6/4/14 6/5/14 6/6/14 6/7/14 6/8/14 6/9/14 6/14/14 6/15/14 6/16/14 6/17/14 6/24/14 6/26/14 6/27/14 6/29/14 6/30/14 7/2/14 7/8/14 7/9/14 Log 10 Total Colony forming Units per microliter (CFU/µl) Date Sampled Dark Red Orange Yellow White Red * * * * 0 2 4 6 8 10 12 14 0.5 1 2 3 4 5 6 7 8 9 10 11 13 14 15 20 Log 10 Colony forming Units per microliter (CFU/ µ l) Droplet Volume ( µ l) 2013 2014 33 4.1.2 Ea110 Ooze Droplet Volume Droplet volumes collected from 2013 and 2014 revealed that yellow ooze droplets were the highest in volume both years with an average volume of 4.9 µl in 2 013 and 6.11 in 2014. In 2014 the ooze droplets were larger overall than in 2013, however statistically only white and orange droplets were different from each other in each year. In 2013, yellow droplets (averaging 4.9 µl, range 1 - 20 µl) were significantl y larger than white (average 2.25 µl, range 1 - 10ul) and orange (average 1.9 µl, range 0.5 - 10 µl) and equivalent to red (3.46 µl average, 1 - 10 µl range) and dark red (averaging 2.5 µl, range of 1 - 5 µl) (Table 1 - 7; Figure 1 - 8). For 2014 the volume of the dr oplets was different than in 2013; The yellow droplets ooze droplets (6.11 µl, range 1 - 20 µl) was only significantly equivalent to dark red droplets (average 5.22 µl, range 1 - 9 µl). the yellow droplets were significantly larger than dark red (average 5.22 µl, range 1 - 9 µl) orange (average 5.5 µl, range 1 - 11 µl), white (average 4.27 µl, range 1 - 15 µl), and red(average 3.92 µl, range 1 - 9 µl) (Table 1 - 6; Figure 1 - 9). When comparing volume of ooze droplets on a day to day basis during the 2013 and 2014 field seasons there were three statistically occurring groups in 2013 and 2014 (Figures 1 - 10 and 1 - 11). The ooze droplet sizes were different between colors on the same sampling date as well (Figures 1 - 10 and 1 - 11). There was not an observable trend in E. amylov ora population size in either 2013 or 2014 as the season progressed (Figure 1 - 4 and 1 - 5). Weather data was also collected from the Enviro - weather station in East Lansing, MI (MSUHORT). In 2013 the only corresponding significance between droplet size and weather events was minimum air temperature, which also corresponded to E. amylovora population size 34 (Table 1 - 4 and 1 - 7). For 2014 there was a significant interaction between droplet size and maximum wind speed; however this was not the case population size (Tables 1 - 5 and 1 - 8). Table 1 - 6: Ooze droplet size averages and ranges by color for each the 2013 and 2014 field season. E. amylovora strain Ea110. Field Season Microliter (µl) Color of Ooze Droplet Dark Red Red Orange Yellow White 2013 Average 2.5 3.46 1.9 4.9 2.25 Range 1 - 5 1 - 10 0.5 - 10 µl 1 - 20 1 - 10 2014 Average 5.22 3.92 5.5 6.11 4.27 Range 1 - 9 1 - 9 1 - 11 1 - 20 1 - 15 Figure 1 - 8: E. amylovora ooze droplet volume sorted by color for the 2013 field season . All E. amylovora strain Ea110. Error bars represent standard error, P <.0001. A AB B BC C 0 1 2 3 4 5 6 Yellow Red Dark Red White Orange 13 Droplet Size ( µ l) 35 Figure 1 - 9: E. amylovora ooze droplet volume sorted by color for the 2014 fiel d season . All E. amylovora strain Ea110. Error bars represent standard error P <.0001. Figure 1 - 10: E. amylovora average ooze droplet volume per date by color in 2013. Ooze droplets were co llected from field - to 10 µl with overall daily averages falling in between 1.5 µl and 5.6 µl from sixteen sample dates. A AB B B B 0 1 2 3 4 5 6 7 Yellow Dark Red Orange White Red 14 E. amylovora Average Ooze Droplet Volume ( µ l) Color by Field Season 0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 5/29/13 5/30/13 6/2/13 6/3/13 6/4/13 6/5/13 6/6/13 6/14/13 6/18/13 6/20/13 6/23/13 6/24/13 6/26/13 6/27/13 6/29/13 6/30/13 Average Volume of Ooze Droplets (µl) Date Sampled in 2013 Yellow White Red Orange Dark Red 36 Figure 1 - 11: E. amylovora average ooze droplet volume per sample date divided by color in 2014. Ooze droplets were collected from field - from 1 µl to 20 µl with daily averages between 1.5 µl and 8.8 µl from twenty - eight sample dates. Table 1 - 7: 2013 Weather data compa ring different weather variables (maximum and minimum relative humidity, maximum and minimum air temperature, maximum wind speed, and precipitation) with ooze droplet volume. Variables P - value Ooze droplet volume and date sampled in relation to maximum relative humidity .5878 Ooze droplet volume and date sampled in relation to minimum relative humidity .7020 Ooze droplet volume and date sampled in relation to maximum air temp. .7953 Ooze droplet volume and date sampled in relation to minimum air temp. .0137 Ooze droplet volume and date sampled in relation to maximum wind speed .6128 Ooze droplet volume and date sampled in relation to precipitation .6451 0 5 10 15 20 25 30 35 40 45 5/28/14 5/29/14 5/30/14 5/31/14 6/1/14 6/2/14 6/4/14 6/5/14 6/6/14 6/7/14 6/8/14 6/9/14 6/14/14 6/15/14 6/16/14 6/17/14 6/24/14 6/26/14 6/27/14 6/29/14 6/30/14 7/2/14 7/8/14 7/9/14 Average Volume of Ooze Droplets ( µl) Date Sampled in 2014 Dark Red Orange Yellow White Red 37 Table 1 - 8: 2014 Weather data comparing different weather variables (maximum and minimum relative humidity, maximum and minimum air temperature, maximum wind speed, and precipitation) to ooze droplet volume. Variables P - value Ooze droplet volume and date sampled in relation to maximum relative humidity .9054 Ooze droplet volume and date sampled in relation to maximum air temperature .7432 Ooze droplet volume and date sampled in relation to minimum air temperature .9509 Ooze droplet volume and date sampled in relation to maximum wind speed .0005 Ooze droplet volume and date sampled in relation to precipitation .0743 4.2 Internal population samples from apple stems In 2014, the internal populations from field inocula ted trees were not statistically different between the three sections; however the ooze droplet population was significantly lower than all the sections (Figure 1 - 12). While the growth chamber experiments from 2013 and 2014 also show that there was no sign ificant difference between the internal population sections (Figure 1 - 13), the ooze droplet population was only significantly different than Section B (Figures 1 - 13). The overall percentage of bacteria found in and on section B, including the ooze droplet, showed that the E. amylovora population in ooze comprises a relatively small percentage of the total bacteria present compared to the internal population, ranging between 7% and 36% of the total E. amylovora population (Figures 1 - 15 and 1 - 14). 38 Figure 1 - 12: E. amylovora internal populations sampled from the field bars represent standard error; P <.0001. Figure 1 - 13: E. amylovora internal populations sampled from growth chamber . 1 cm sections of and 2014. Error bars represent standard error; P = <.05. A A A B 1.00E+08 1.00E+09 1.00E+10 1.00E+11 1.00E+12 Section A Section B Section C Ooze Colony forming Units (CFU) AB A AB B 0 1E+10 2E+10 3E+10 4E+10 5E+10 Section A Section B Section C Ooze Colony forming units (CFU) 39 Figure 1 - 14: E. amylovora average internal population of section B compared to droplets from apple shoots and average ooze drop population from the 2014 field season . The percentage indicates ho w much bacteria was found in each area, ooze or internal, of section B. Figure 1 - 15: E. amylovora average internal population of section B compared to droplets from growth chamber . E. amylovora average internal populations present in the 1 cm section B of and 2014. The percentage indicates how much bacteria was found in each area, ooze or internal, of section B. 93% 7% Section B Ooze 63% 37% Section B Ooze 40 4.3 Stem inoculation of three distinct apple cult ivars with four E. amylovora strains during the 2014 field season In 2014, 364 ooze droplets were collected from late May to early July from three inoculated with four strains of E. amylovora (Ea110, EL01, GH9, and K2 ) (Table 1 - 9 ). At least thirt y ooze droplets were collected over the season from each cultivar and strain combination oplets as readily as the other cultivar and strain combinations (Table 1 - 9 ). Table 1 - 9 : Ooze droplet totals collected from 3 cultivars and 4 strains in 2014. Cultivar Strain Number of droplets Ea110 30 EL01 30 GH9 30 K2 30 Ea110 44 EL01 30 GH9 30 K2 30 Ea110 30 EL01 30 GH9 21 K2 30 4.3.1 Differences in population between apple cultivars and E. amylovora strains Populations of strain Ea110 were statistically significant between each cultivar (Figure 1 - 16). The populations of the three other strains (EL01, GH9, and K2) were significantly different between cultivars; all three of the strains exhibited a significant reduction in population in the 41 - 16). There was no significant interaction between strain, cultivar, and date in relation to population (Table 1 - 10 ). Figure 1 - 16: E. amylovora population in ooze droplets compared between strai ns and cultivars. the field - inoculated trees Error bars represent standard er ror; P <.0001. 4.3.2 Differences in ooze droplet volume between apple cultivars and E. amylovora strains two cultivars, as well as in comparison with the other strain BC B A B BC AB A AB BC C BC C 0 2E+09 4E+09 6E+09 8E+09 1E+10 1.2E+10 1.4E+10 Ea110 EL01 GH9 K2 Ea110 EL01 GH9 K2 Ea110 EL01 GH9 K2 September Wonder Fuji Kit Jonathan Linda Mac Colony Forming Units per microliter (CFU/µl) 42 respec tively (Figure 17). Comparisons between the strains Ea110 and GH9 were statistically the same, whereas again EL01 and K2 had larger volumes in different cultivars (Figure 1 - 17). When the strains and cultivar ooze droplet volumes were combined by date samp led, the 17 June 2014 was significantly different than the rest of the sampling dates (Figure 1 - 18). When further explored however, there was no significance between volume, cultivar, and strain ( P =.3314) or for the time of inoculation ( P =.3560). The only significant weather variable was max wind speed, which was 37.8 miles per hour ( P <.0001). When the population, cultivar, and droplet volume were statistically examined as a whole, there was no significant difference between the cultivars (Table 1 - 10 ). How ever the strains were all significantly different when examined against droplet population and volume (Table 1 - 1 1 ). Figure 1 - 17: E. amylovora droplet volume compared between strains and cultivars . Strains listed are Ea110, EL01, GH9, and K2 as well as c the field - inoculated trees. Error bars represent standard error, P <.0001. CD BC CD A CD CD CD CD D AB CD CD 0 2 4 6 8 10 12 14 EA110 EL01 GH9 K2 EA110 EL01 GH9 K2 EA110 EL01 GH9 K2 Fuji Kit Jonathan Linda Mac Droplet Volume (µl) 43 Figure 1 - 18: Average E. amylovora ooze drop volume over the twenty - eight day sampling period in 2014 combined for all cultivars and strains . Error bars represent standard error, P <.0001. Table 1 - 10 : 2014 Chart of all insignificant interactions between variables for the 2014 inoculation of four stra ins onto three cultivars . The variables shown are: Strain, cultivar, date collected, distance from bud scar, and time of inoculation along with the p - value of significance. Insignificant interaction P - value Strain, cultivar, and date in relation to population volume. .2631 Strain, cultivar, and date in relation to droplet volume. .5128 Distance from bud scar, date, and time of inoculation in relation to population volume. .5178 Population and cultivar in relation to ooze volume. p - value for inte raction 0.1278 0.0549 0.4314 0.1058 C BC BC C BC B BC C BC C C C BC C B A C BC C BC C C BC C C C C C 0 2 4 6 8 10 12 14 5/27/2014 5/28/2014 5/30/2014 5/31/2014 6/1/2014 6/2/2014 6/4/2014 6/5/2014 6/6/2014 6/7/2014 6/8/2014 6/9/2014 6/14/2014 6/15/2014 6/16/2014 6/17/2014 6/24/2014 6/25/2014 6/26/2014 6/27/2014 6/28/2014 6/29/2014 6/30/2014 7/1/2014 7/2/2014 7/8/2014 7/9/2014 7/10/2014 Average Droplet Volume (µl) Samping Date 44 Table 1 - 11 : Chart of the four strains from the 2014 field season showing that there was significant difference between strains in regards to droplet volume and population. Strain P - value GH9 .0032 EA110 .0131 K2 .0001 EL01 <.0001 4.4 Scanning Electron Microscopy In SEM micrographs of collapsed ooze droplets, bacterial cells can clearly be observed surrounding what could possibly be the remains of exopolysaccharides (Figure 1 - 19). Another notable observation is the masses of bacteria which appear to emerge from the exit point for the ooze droplet from the fruit. There were no natural openings observed around the collapsed ooze droplets; the tearing in the skin of the apple was an artifact from the SEM fixation, as it was also seen in uninfected fruit from the same v ariety and age (Figure 1 - 19 B; Figure 1 - 20). When the ooze dried before fixation, the ooze would be more rigid than the waspy - looking edges of the collapsed droplets (Figure 1 - 19 A; Figure 1 - 20). Again, no natural opening was observed. This circular shap ed structure from which the ooze emerged does not appear to be a natural opening, as it is only 10 µm in diameter (Figure 1 - 21). When submerged in ethanol, ooze droplets did not dissolve when shaken (personal observations). Only when 10% water was added an d the droplets shaken vigorously did the dried ooze dissolve, meaning that the early stages of critical point drying (50% ethanol 50% diH20) could have possibly affected the shape and texture of the ooze droplets as they were processed. However the SEM cri tical point drying 45 protocol does not result in samples being vigorously shaken, if anything the opposite is desired as not to harm the specimens in this fragile form. Erumpent mounds, large convex shaped mounds with a central tear, were found underneath l arge ooze droplets (Figure 1 - 22). These mounds were also visible with the naked eye after the drop was removed (personal observations). These mounds appear to have been formed from internal pressure based on the shape of the stem or petiole tissue (Figure 1 - 22). Bacteria was also found around the erumpent mounds where the ooze drop had been removed. Dried bacterial ooze is prevalent even when ooze droplets are physically removed before drying (Figure 1 - 23). The dried ooze could also stick to trichomes in th e surrounding area of the erumpent mound. Along the edges of the erumpent mounds bacterial cells and exopolysaccharides could be found attached to the surface of the plant tissue (Figure 1 - 23 C). There were observances of ooze emerging from the same wound underneath an older, dried droplet in the field. Figure 1 - 19: SEM micrographs of a collapsed ooze droplet that formed on a naturally occurring E. amylovora infected immature apple fruit . The ooze droplet was not removed before SEM critical point dry ing of the fruit. A: Collapsed ooze droplet on an infected immature apple fruit. B: Close up of the edge of the collapsed ooze droplet with bacteria still attached to the fruit and exopolysaccharides. 46 Figure 1 - 20: SEM micrograph zoomed to 450x of a coll apsed ooze droplet that formed on a naturally occurring E. amylovora infected immature apple fruit . The ooze droplet was fresh and not removed before SEM critical point drying of the fruit. The mound of bacteria pictured is thought to be the exit point of the E. amylovora from the internal apple fruit tissue. Figure 1 - 21: SEM micrographs displaying a gradual close up of an exit point by an E. amylovora ooze drop on an inoculated stem. The ooze droplet was allowed to dry before SEM critical point fixati on processing. The close up fixates on the dried mass of exopolysaccharides and bacteria present near the circular exit hole. 47 Figure 1 - 22: An erumpent mound found underneath a large ooze droplet on a leaf petiole from an inoculated stem . The ooze drop let was removed prior to critical point drying for SEM fixation. The red arrow points to exit point of ooze droplet. Figure 1 - 23: Various images of dried ooze formed around erumpent mounds. The fresh ooze was removed prior to SEM critical point fixatio n, however dried ooze was still seen around area near exiting wound. A: dried ooze on surface nearby an erumpent mound found on an apple stem. B: Dried ooze nearby the erumpent mound, ooze is also present on nearby trichomes. C: Area of an erumpent mound a t the edge of the exit point. 48 4.5 Analysis of virulence gene expression of E. amylovora in apple stem tissue and ooze Expression of hrpL is an indicator for activation of the type III secretion system which is required for infection of host plant cells. Gene expression of hrpL was different between the hrpL expression in the 48 hour (early stage) infected shoots while bacteria isolated from stem Section A dwelling bacteria had less hrpL - 24). Ea110 E. amylovora in ooze expressed hrpL similar to the Hrp MM control for both cult ivars (Figure 1 - 24). The expression of the dspE gene is also an indication of the type III secretion system and is required for pathogenicity of E. amylovora . dspE gene expression was reduced for all Ea110 from the field compared to the HRP minimal mediu m, however it was still being expressed at a higher level in the infected plant tissue compared to the LB grown E. amylovora (Figure 1 - 25). The gene lsc encodes levansucrase, which is an important virulence factor in E. amylovora that allows for sugar ac quisition and biofilm formation. For both cultivars, bacteria isolated from section A had the largest expression of lsc , around 6 - 7 fold higher expression than the Hrp MM bacteria (Figure 1 - als - 26). Amylovoran is an important pathogenicity factor in E. amylovora and production i s measured by the expression of gene amsK. Without actively producing some level of amylovoran, E. amylovora is not able to infect a host. The E. amylovora in ooze were lower expressing of amsK than the bacteria in the Hrp MM, however they are probably not significantly different. The fold of expression for amsK are low, and there is probably not any significant difference between the cultivars or tissues sampled (Figure 1 - 27). 49 Figure 1 - 24: Relative expression of hrpL in different infected plant tissues, ooze drops, and Hrp inducing minimal medium in comparison to expression levels in LB medium. The expression endogenous control recA. Figure 1 - 25: Relative expression of dspE in different infected plant tissues, ooze drops, and Hrp inducing minimal medium in comparison to expression levels in LB medium. The expression CT) method, normalized by an endogenous control recA. 0 10 20 30 40 50 60 70 80 90 100 McIntosh Jonathan Hrp MM Fold of expression of hrpL compared to LB Early stage infected shoots Section A Ooze drops 0 5 10 15 20 25 30 35 McIntosh Jonathan Hrp MM Fold of expression of dspE compared to LB Early stage infected shoots Section A Ooze drops 50 Figure 1 - 26: Relative expression of lsc in different infected plant tissues, ooze drops, and Hrp inducing minimal medium in comparison to expression levels in LB medium. The expression levels of targe endogenous control recA. Figure 1 - 27: Relative expression of amsK in different infected plant tissues, ooze drops, and Hrp inducing minimal medium in comparison to expression lev els in LB medium . The expression endogenous control recA. 0 1 2 3 4 5 6 7 8 9 McIntosh Jonathan Hrp MM Fold of expression of lsc compared to LB Early stage infected shoots Section A Ooze drops 0 2 4 6 8 10 12 14 16 McIntosh Jonathan Hrp MM Fold of expression of amsK compared to LB Early stage infected shoots Section A Ooze drops 51 4.6 Spectra analysis of ooze and apple tissue pigment colors When ran in the spectrometer, there was a consistent peak and shoulder at different concentrations present for every sample that contained pigment (dark red, red, orange, yellow ooze droplets and orange - tipped fire blight shoots) (Figure 1 - 28). The peak wa s slightly visible for the control of a non - infected apple stem; however the stem tissue had been wounded in preparation for the spectrometer. The E. amylovora grown in culture did not contain the peak or shoulder present in the other samples (Figure 1 - 28) . The peak and shoulder is consistent with flavanones, which are known defense chemicals produced against E. amylovora. Figure 1 - 28: Spectra of various tissues containing pigment (ooze droplets and orange - tipped fire blight shoots) against controls (uni nfected apple stem tissue). 52 5. Discussion Our field and laboratory data indicated that E. amylovora populations within ooze droplets were extremely large, that populations were consistently higher in darker - colored ooze drops, and that the cells present in ooze drops were expressing virulence genes suggesting that these cells would be primed for infecti on if disseminat ed to susceptible host tissue. When considering the average volume of ooze droplets and the population size determined per microliter, overall populations per individual ooze droplet ranged from 10 8 to 10 11 cells. E. amylovora ooze function s in the epidemiology of fire blight through both direct dissemination of virulent bacterial cells via rain - splash, wind, or insect activity ( Schroth et al., 1974 ) or in long - term survival, as E. amylovora cells in dried ooze drops can remain viable and pa thogenic for periods of 1 - 2 years ( Hildebrand, 1939; van der Zwet et al , 2000). Ooze populations are many orders of magnitude larger than those necessary to establish infections in inoculated, wounded apple shoots. For example, Crosse et al (1972) determin ed a median inoculum dose of only 38 bacterial cells following pipetting of a droplet of cells onto the cut end of the main leaf veins of an apple leaf at the shoot tip, and Ruz et al (2008) reported a median effective pathogen dose of 2.01 x 10 5 cells using the most effective inoculatio n method in a comparative study: cutting an apple leaf with scissors dipped in cell inoculum. Why then might E. amylovora ooze populations be so large? In nature, ooze from cankers supplies primary inoculum for fl ower infection, and is primarily disseminated by insects such as flies. When considering dissemination of E. amylovora from cankers among native Rosaceae trees, these trees were likely scattered randomly in the landscape, with potentially large distances b etween them. Flies contacting ooze on a canker would likely only pick up fractions of a microliter of ooze, necessitating dense bacterial populations to ensure successful carriage and 53 cell delivery between canker and flower. There would also be a time comp onent between acquisition and delivery; it is possible that the exopolysaccharide matrix of ooze would enhance survival during this dissemination phase. For shoot infection in orchards, freshly - injured tissue at shoot tips is not always available; thus, la rger cell populations are probably required to initiate shoot infections than those reported previously in wound inoculation studies (Crosse et al. 1972; Ruz et al. 2008). This again would signify a requirement for excessively large E. amylovora population s in ooze. In addition, during the long - term period of cell survival in dried ooze, large amounts of the population may die and a high initial population would be needed to allow for the survival of the pathogen. While the E. amylovora populations in ooze drops are very large, our analyses indicated that these populations represent at most 36% of the total E. amylovora population within the infected shoot. Thus, ooze represents an allocation of infecting cells to dissemination, meaning that a larger propor tion of cells are available for continued systemic infection of the host. Based on our SEM imaging studies, ooze drops appear to have been formed after a rupturing of the parenchyma and epidermis layers of the cell. Since E. amylovora does not secrete cell wall degrading enzymes, the pressure hypothesis formed by Schouten (1989) that enables cel ls in exopolysaccharide to move through the host could also explain ooze drop formation (Seemuller and Beer, 1976; Schouten, 1989) . When ooze drops were examined in the field, they ranged in color from white to dark red; the color of the drops did not change if the ooze drops were allowed to dry. The E. amylovora populations were found to be consistently higher in the darker shades of ooze than white and yellow. When ooze is released from cankers, it is typically orange to dark red. Our field observations confirm those made by others in observing a high frequency of visits by flies to 54 oozing cankers, with the flies walking through and appearing to consume the ooze dro plets (Ark and Thomas, 1936; Hildebrand 1939; Van der Zwet and Keil, 1979; Hildebrand et al., 2001). Hildebrand et al (2001) conducted a large field study of insect dispersal of E. amylovora , however he did not address cell counts (Hildebrand et al., 2001) . These flies may be attracted to the darker red pigment, however, whether the pigment is directly involved or the source of the pigment is the attractant is still unknown. The pigment is likely a flavanone, particularly flavanone - 3 - hydroxlase or 3 - dexpyfl avonoid derived compounds. Besides being a potential defense response to E. amylovora , the role of these compounds in host response is unknown or circumstantial at best (Flachowsky et al., 2012). The E. amylovora present in ooze droplets had expression levels of virulence genes that were reduced compared to actively invading cells, however they were still high enough to be considered primed compared to the HRP MM or LB grown bacteria. E. amylovora could possibly have regulators that when e xposed to oxygen, may contribute to the oppression of gene expression. Since the ooze droplets are external and exposed to more oxygen than in intracellular spaces, these regulators may play a role in the suppression compared to the bacterial cells that re main inside the host. 55 CHAPTER 2 A FIRST - YEAR REPORT ON THE BIOLOGICIAL CONTROL COMPOUND BLOSSOM - PROTECT IN MICHIGAN ORCHARDS 56 1. Abstract As the National Organic Program of antibiotic synthetic compound s for fire blig ht comes to a close, along with streptomycin resistance, there needs to be new ways to protect pome orchards from Erwinia amylovora, the causal agent of fire blight . One way this could be achieved is by finding biological control agents with a lower va riability in effectiveness. In Michigan, trials are being run of a prod uct called Blossom - Protect, which is a yeast - like fungus Aureobasidium pullulans . A. pullulans strains 10 and 40 have had success in suppressing other plant pathogens. Due to biological control agents in Michigan lack of reliability year to year, other organic compounds were added to the trials (Sundin et al., 2009). Blossom blight control was significantly greater for Blossom - Protect + Cueva at 3 sprays (BP+C3), Blossom - Protect + Oxidat e at 2 sprays (BP+O2), Blossom - Protect at 2 late sprays (BP+2L) and Blossom - Protect + Nordox + Umbrella at 3 sprays (BP+N+U3) compared to the unsprayed control (USC); Blossom - Protect at 4 sprays (BP+4) was the same as USC in percent infection. BP+O2 also resulted in no significant russeting and had significantly decreased apple scab presence. In year one of the study, there were no significant differences between pure Blossom - protect sprays in regards to russeting. 2. Introduction Fire Blight, caus ed by the gram negative bacterium Erwinia amylovora , is a devastating disease of pome fruit that occurs in orchards around the world . Traditionally , antibiotics such as copper and streptomycin have been used to combat the primary stage of infection which occurs at flower bloom (Sundin et al, 2009). The need to use a chemical defense against E. amylovora is so crucial even organic apple and pea r growers could use antibiotics; t he National Organic Program allowed the use of antibiotic synthetic compounds from 2002 to 21 Oct 2014 on tree 57 fruit fo r fire blight control (Granatstein, 2014 ). As 2014 was the last year that organic growers had antibiotics available in their disease combat tool box, new methods of chemical control for fire blight are needed. The other conc ern about traditional antibiotics is the level of streptomycin resistance occurring in E. amylovora (Jones and Schnabel, 2000; Sundin et al, 2009; Vanneste, have been gaining momentum and are continuing to promise new methods of disease control (Vanneste, 2011). Biological control is the use of another organism or the products produced by an organism as an antagonizer against an unwanted population of organis ms. Since flowers are the primary infection site, specifically the stigma (Thomson, 1986), many biological controls have tried to target that ecological niche in the flower (Vanneste, 2011). This biological control mode of action is known as competition (P al and McSpadden Gardener, 2006). The other mode of action taken advantage of by fire blight antagonists are antibiotic - mediated suppression (Pal and McSpadden Gardener, 2006). There are currently a few biological controls available for apple and pear gro wers against fire blight (Sundin et al, 2009; Vanneste, 2011). The majority of control agents for fire blight are bacterial antagonists, including Pseudomonas fluorescens strain A506 (BlightBan A506), Pantoea agglomerans strain E235 (Bloomtime FD Biopestic ide), Pantoea vagans strain C9 - 1 (Blightban C9 - 1), and Bacillus subtilis strain 713 (Serenade) (Sundin et al., 2009; Vanneste, 2011). The challenge with biological control is the variability in effectiveness (Sundin et al, 2009; Vanneste, 2011). In a stud y conducted by Sundin et al. (2009) that spanned over three states and six years, there were many inconsistencies with performance of the biologicals they tested: Pseudomonas fluorescens strain A506 (BlightBan A506), Pantoea agglomerans strain 58 E235 ( Bloomtime FD Biopesticide), Pantoea vagans strain C9 - 1 (Blightban C9 - 1), and a mixture of Pseudomonas fluorescens strain A506 (BlightBan A506 ) and Pantoea vagans strain C9 - 1 (Blightban C9 - 1). Ultimately, they found that streptomycin was still the best cont rol for fire blight over the four biological control treatments (Sundin et al., 2009). Besides bacteria, yeasts have been screened for potential biological control agents as well (Pusey et al., 2009). However, like bacteria, yeasts have shown mixed result s as biological control agents. A yeast success as a biological control was for postharvest apple diseases grey mold ( Botris cinerea ) and blue mold ( Penicillum expansum ) by Cryptococcus laurentii LS28 and Aureobasidium pullulans LS30 (Lima et al., 2003). An example of a failed commercial product was Aspire, containing Canidida oleophilas , which could not control postharvest diseases (Droby et al. 1998). Aureobasidium pullulans is a black yeast - like fungus that has tested to have antagonistic properties a gainst many plant pathogens (Lima et al., 2003; Kunz, 2004; Duffy et al., 2006; Pusey et al., 2009). One way that the yeast is currently being marketed is as Blossom - Protect, a mixture of two strains, A. pullulans 10 and 40 ( Ap CF10 and Ap CF40 respectivel y) (Kunz, 2004). Developed in the early 2000s in Germany, Blossom - Protect could be a successful biological control against E. amylovora (Kunz, 2004; Duffy et al., 2006; Pusey et al., 2009). The mode of action of A. pullulans is thought to be antibiosis of nutrient acquisition and n iche competition (Duffy et al., 2006). However in another study involving control of molds, A. pullulans produced antifungal compounds to inhibit mold growth and no antibiosis was observed (Castoria et al., 2001). There have been many studies on the efficacy of Blossom - Protect on controlling fire blight before the biological was released . Duffy et al. (2006) found that Blossom - Protect had 59 significant fire blight control on seven cultivars, with variation, and found that 24 - 48 hour prior to bacterial inoculation reduced the bacterial population by 3 log of colony forming units (log CFU ) (Duffy et al., 2006). Both Kunz (2004) and Pusey e t al. (2009) also report that A. pullulans colonized detached apple blossoms better than other ant agonistic yeasts and reduced E. amylovora populations as well (Kunz, 2004; Pusey et al., 2009). Other field trials in Germany have reported that two to four applications of yeasts as biological controls as can result in fruit russeting (Pusey et al., 2009) . Kunz (2004) however states that no increase in russeting was found in a two year organic study on various cultivars (Kunz, 2004). The label for Blossom - Protect however indicates that russeting may occur in susceptible fruit in late blossom. In year one of this study , Blossom - Protect was evaluated for use in Michigan. Different overall rates of the product were tested in the field as with a laboratory component observing the A. pullulans , E. amylovora, fungal, and total bacterial viability on the blossoms . Incidences of russeting, blossom and shoot blight, as well as apple scab were also reported later in the season. 3. Materials and Methods In 2014 Blossom - Protect (Westbridge Agricultural Products, Vista, CA) was applied to McIntosh trees at different bl oom intervals on the MSU Plant Pathology Research Farm in East Lansing, MI. Along with Blossom - Protect, Cueva (Certis, USA) a copper compound, Nordox 75 WG (Nordox AS, Oslo, Norway) another copper bactericide, and Umbrella (Agrian, Fresno, CA USA) which co ntains terpene resins, tall oil fatty acids and alkyl phenol ethoxylate. Each treatment had four single - tree replicates (Table 2 - 1).These treatments and replicates were arranged in a complete randomized block design. In each treatment, E. amylovora was in oculated at 80 - 100% bloom. The E. amylovora 60 strain Ea110, a spontaneous rifampicin mutant native to Michigan, was used in this trial. Prior to use, Ea110 was stored at - 80 °C in 15% glycerol. Prior to inoculation, cells were grown in Luria - Burtani broth (LB ) and adjusted to 1x10 6 CFU/ml in 0.5x PBS. The suspension was kept on ice to the field then sprayed on the trees with an 11.4 - liter pump mist sprayer (Solo, Newport News, VA ). Table 2 - 1: 2014 field season Blossom - Protect treatment list along with product rates used and the percent bloom timing of each treatment . Treatment abbreviation is used to denote which compounds are used in each treatment and how many times the treatment was applied. Treatment Abbreviation Treatment and product per acre Treatment Timings (Percent bloom) 1 BP - 4 Blossom - Protect 1.34 lb. Buffer A 9.35 lb. 10%, 40%, 70 - 80%, 100% 2 BP+C3 Blossom - Protect 1.34 lb. Buffer A 9.35 lb. Cueva .5gal/100gal 40%, 70 - 80%, 100% 4 BP+O2 Oxidate Blossom - Protect 1.34 lb. Buffer A 9.35 lb. 70 - 80% (before and after), 100% 5 BP+L2 Blossom - Protect 1.34 lb. Buffer A 9.35 lb. 70 - 80%, 100% 8 BP+N+U3 Blossom - Protect 1.34 lb. Buffer A 9.35 lb. Nordox 75WG 1.25lbs/Acre Umbrella 16 FL oz./100gal 40%, 70 - 80%, 100% 10 USC Unsprayed Control 3.1 Laboratory Studies of Blossom - Protect Methods For each time point, eight blossoms were randomly collected from four trees (one each rep). See Table 2 - 1 for percent bloom time point that treatment was applied and sampled. Each rep was separately added to a glass sonication tube with 20 mL .5x PBS and s onicated for seven minutes (McGhee et al., 2011) The resulting solution was then diluted and spread plated onto 61 non - amended unless noted to see the full scope of microorg anisms present. Counts were taken two days after plating for yeast, bacteria (with no differentiating E. amylovora from other colonies unless noted), and non - yeast - like fungal colonies. For BP+O2, an oxidate spray was applied four hours before the other 70 - 80% sprays. Sampling occurred prior to this spray and again four hours later. E. amylovora was inoculated on 17 May 2014 between 80% - 100% bloom as described above. On the 100% bloom date, E . amylovora was also directly counted using KB plates amended with rifampicin 100 µg ml 1 and cycloheximide at 50 µg ml 1 to inhibit fungal growth. When the blossoms were sampled at each time point, 28 other flowers were pressed to NAG media to observe yeast colonies. Table 2 - 2: 2014 field study dates indicating the percent bloom sampled for Blossom - Protect . The X indicates that no spray was applied. Date 2014 13 May 14 May 15 May 16 May 17 May 18 May Percent Bloom Sampled 10% 40% X 80% Inoculation 100% 3.2 2014 Field Aspect of Blossom - Protect Methods On 25 July 2014, ratings were taken for prevalence of blossom and shoot blight. Blossom and shoot rating were percent of infected blossoms/shoots out of a hundred randomly selected shoots. Ratings for apple scab were also taken for this experiment on 8 September 2014 , as crucial apple scab fu ngicide sprays were not applied as to not kill the yeast. Apple scab was rated by counting all the leaves on a randomly selected shoot and giving a percentage of infected leaves. This was repeated twenty times per rep. For fruit, 100 randomly selected fruit per rep were observed for scab. Fruit russeting was also rated, as noted by Pusey et al (2009) that is could be a concern. The rating system for this was 60 random fruit were rated by the USDA 62 grade fancy protocol (USDA, 2002). Any fruit with over 10% russet was determined unmarketable. The calyx and stem flesh of the apples were also rated on russeting according to the USDA standards for fancy fruit (USDA, 2002). 3.3 Statistical analysis All statistical analysis was conducted using SAS 9.4, PROC GLM and PROC Mixed using complete randomized blocking. 4. Results 4.1 Laboratory Studies of Blossom - Protect Results The oxidate spray along with the Nordox and Umbrella treatment were the only one s that significantly lowered bacterial populations ; comparing the 4 0% and 80% populations were significantly less at 1 00% bloom for both treatments (Figure 2 - 1) . In oxidate treatment, there were no bacteria present at the 80% sampling which was a fter the o xidate spray. Populations did recover at the 100% sampling and but was statistically less than before the oxidate spray (Figure 2 - 1). There were significantly less bacteria a t 100% bloom at the recommended sprays (BP+4) compared to 80%, but was statistical ly at the 40% treatment. T he other two treatments , Cueva an d the late Blossom Protect, were statistically similar throughout the time points. (Figure 2 - 1). The 2 late sprays were statistically the same as the 4 recommended sprays. The fungal population was significantly reduced after the oxidate spray (Figure 2 - 2). The reduction in fungal growth however, was similar to populations at 100% blo om of Cueva . Cueva ( BP+C3 ) had a significant raise in fungal populations in 80%, but dropped back to levels at 40% at 100% bloom (Figure 2 - 2). The recommended Blossom Protect rate had higher fungal populations as bloom went on. 63 For A. pullulans populations, 10% bloom was sampled after the first spray of Blossom - Protect was applied. For the reco mmended spray (BP - 4) there was a significant sharp population decline from 10% to 40% bloom (Figure 2 - 3) . However the A. pullulans populations signific antly recovered at 80% but failed to maintain population at 100% (Figure 2 - 3). The Cueva ( BP+C3 ) gained a jump in population from 40% bloom to 80% bloom, but reverted back down to the lowest population level at 100% bloom (Figure 2 - 3). The Oxidate treatment ( BP+O2 ) , where Blossom - Protect was applied four hours after the oxidate burst, saw high levels of A. p ullulans populations for both 80% and 100% bloom (Figure 2 - 3). BP+2L showed a significant increase in A. pullulans populations from 80% to 100% bloom, whereas BP+N+U3 stayed consistent through 40%, 80% and 100% bloom (Figure 2 - 3). The recommended spray (BP - 4) had the highest yeast population levels. For each treatment, E. amylovora populations were sampled at 100% bloom. All of the treatments have significantly less E. amylovora present than USC (Figure 2 - 4). There were some differences in the treatments a s well; the Cueva treatment had no E. amylovora present (Figure 2 - 4). The Oxidate treatment had significantly more A. amylovora present than the other three treatments with populations present (Figure 2 - 4). For the blossom presses indicating where or not yeast is present inside the blossom, BP+C3 shows a loss of A. pullulans statistically significant (Figure 2 - 5). For BP+4, there is also no significan t difference in the screening for blossom protect (Figure 2 - 5). Besides the pre oxidate spray of BP+O2, the only other significant drop off is in BP+N+U3, where 100% bloom is statistically lower than any other of the blossom prints (Figure 2 - 5). 64 Figure 2 - 1: Total bacterial populations from flowers sampled in spring 2014 which were collected at the percent bloom indicated in the horizontal axis . Error bars represent standard error, P = >.05. Axis legend: Blossom - Protect + Cueva at 3 sprays (BP+C3), Blosso m - Protect + Oxidate at 2 sprays (BP+O2), Blossom - Protect at 2 late sprays (BP+2L), Blossom - Protect + Nordox + Umbrella at 3 sprays (BP+N+U3), Blossom - Protect at 4 sprays (BP+4). Figure 2 - 2: Fungal populations from flowers sampled in spring 2014 which were collected at the percent bloom indicated in the horizontal axis . Error bars represent standard error, P = >.05. Axis legend: Blossom - Protect + Cueva at 3 sprays (BP+C3), Blossom - Protect + Oxidate at 2 sprays (BP+O2), Blossom - Protect at 2 late sprays ( BP+2L), Blossom - Protect + Nordox + Umbrella at 3 sprays (BP+N+U3), Blossom - Protect at 4 sprays (BP+4). B A B B B B A B B B AB B C 1.00E+00 1.00E+01 1.00E+02 1.00E+03 1.00E+04 1.00E+05 1.00E+06 1.00E+07 40% 80% 100% 40% 80% 100% Preoxidate 80% 100% 80% 100% 40% 80% 100% BP-4 BP+C3 Bp+O2 BP+L2 BP+N+U3 Average Bacterial Colony forming Units (CFU/g) Percent Bloom of Blossoms by Treatment B A A C AB C A C B B B B B 1.00E+00 1.00E+01 1.00E+02 1.00E+03 1.00E+04 1.00E+05 1.00E+06 40% 80% 100% 40% 80% 100% Preoxidate 80% 100% 80% 100% 40% 80% 100% BP-4 BP+C3 Bp+O2 BP+L2 BP+N+U3 Average Fungal Colonies per Gram Percent Bloom of Blossoms by Treatment 65 Figure 2 - 3: A. pullulans populations from flowers sampled in spring 2014 which were collected at the percent bloom indicated in the horizontal axis . E rror bars represent standard error, P = >.05. Axis legend: Blossom - Protect + Cueva at 3 sprays (BP+C3), Blossom - Protect + Oxidate at 2 sprays (BP+O2), Blossom - Protect at 2 late sprays (BP+2L), Blossom - Protect + Nordox + Umbrella at 3 sprays (BP+N+U3), Blo s som - Protect at 4 sprays (BP+4) Figure 2 - 4: E. amylovora populations at 100% bloom for each treatment . Error bars represent standard error, P = >.05. Axis legend: Blossom - Protect + Cueva at 3 sprays (BP+C3), Blossom - Protect + Oxidate at 2 sprays (BP+O2), Blossom - Protect at 2 late sprays (BP+2L), Blossom - Protect + Nordox + Umbrella at 3 sprays (BP+N+U3), Blossom - Protect at 4 sprays (BP+4). A D AB BC D C F C B C C C CD C 1.00E+00 1.00E+01 1.00E+02 1.00E+03 1.00E+04 1.00E+05 1.00E+06 1.00E+07 10% 40% 80% 100% 40% 80% 100% Preoxidate 80% 100% 80% 100% 40% 80% 100% BP-4 BP+C3 Bp+O2 BP+L2 BP+N+U3 A. pullulans Population per Gram) Percent Bloom of Blossoms by Treatment C B C C A 1.00E+00 1.00E+01 1.00E+02 1.00E+03 1.00E+04 1.00E+05 1.00E+06 1.00E+07 E. amylovora colony forming units (CFU/g) Treatment at 100% bloom 66 Figure 2 - 5: Average percent of blossom prints taken in the field that tested positiv e for A. pullulans at each sampling percentage . Error bars represent standard error, P = >.05. Axis legend: Blossom - Protect + Cueva at 3 sprays (BP+C3), Blossom - Protect + Oxidate at 2 sprays (BP+O2), Blossom - Protect at 2 late sprays (BP+2L), Blossom - Protec t + Nordox + Umbrella at 3 sprays (BP+N+U3), Blossom - Protect at 4 sprays (BP+4). 4.2 2014 Field Results On average, each treatment, though not always significant, was lower than USC for overall scab prevalence (Figure 2 - 6). Surprisingly, a few of the treatments were significantly lower than USC; for leaf and fruit scab BP+O2 and BP+N+U3 were significantly lo wer (Figure 2 - 6). Also lower than the USC was BP+C3 for fruit scab, however BP+4 and BP+2L were significantly the same as the USC (Figure 2 - 6). In regards to shoot blight, all but one treatment had significantly lower incidence of blight; BP - 4 actually h ad a higher prevalence of shoot blight than USC, but statistically lower blossom blight (Figure 2 - 7). The other treatments; BP+C3, BP+O2, BP+2L, and BP+N+U3 all had significantly lower incidence of blossom and shoot blight than USC, but there was no A A AB AB AB AB B D A A A A A A C 0 20 40 60 80 100 120 10 40 80 100 40 80 100 Pre ox 80 100 80 100 40 80 100 BP-4 BP+C3 BP+O2 BP+2L BP+N+U3 Percent of blossoms with yeast present (%) Percent bloom per treatment 67 signi ficant difference between them (Figure 2 - 7). Between shoot and blossom blight however there are some differences; BP - 4, BP+O2 and the USC had significantly lower incidence of shoot blight than blossom blight (Figure 2 - 7). For fruit russeting, there was n o significant difference in russeting between the stem ( P =.1424) or calyx ( P =1540), however there was a difference in over ten percent total fruit russeting (Figure 2 - 8). BP+N+U3 had 11% of the fruit having more than 10% russet, which was significantly d ifferent than the rest of the treatments. BP+4, BP+O2, and BP+2L had statistically the same amount of russet as the USC; while BP+C3 had more than the other treatments, the average was still lower than BP+N+U3 (Figure 2 - 8). Figure 2 - 6: 2014 field data f or Blossom - Protect treatment trials indicating the percent of infected apple leaves and fruit with apple scab . There were no significant differences between the leaf and fruit percent infected per treatment ( P = .5532). Error bars represent standard error, P = >.05. Axis legend: Blossom - Protect + Cueva at 3 sprays (BP+C3), Blossom - Protect + Oxidate at 2 sprays (BP+O2), Blossom - Protect at 2 late sprays (BP+2L), Blossom - Protect + Nordox + Umbrella at 3 sprays (BP+N+U3), Blossom - Protect at 4 sprays (BP+4). AB AB C AB BC A ab b b ab b a 0% 5% 10% 15% 20% 25% 30% 35% 40% 45% BP-4 BP+C3 BP+O2 BP+2L BP+N+U3 USC Apple Scab, Percent Infected (%) Treatment Apple scab Leaf Apple scab Fruit 68 Figure 2 - 7: 2014 field data for Blossom - Protect treatment trials indicating the percent of infected apple shoots and blossoms with fire blight . There were significant differences (not noted in figure) between BP - 4, BP - O2, and the USC shoot and blossom perc ent infected per treatment ( P = >.05). Error bars represent standard error, P = >.05. Axis legend: Blossom - Protect + Cueva at 3 sprays (BP+C3), Blossom - Protect + Oxidate at 2 sprays (BP+O2), Blossom - Protect at 2 late sprays (BP+2L), Blossom - Protect + Nord ox + Umbrella at 3 sprays (BP+N+U3), Blossom - Protect at 4 sprays (BP+4). Figure 2 - 8: 2014 field data for Blossom - Protect treatment trials indicating the percent of fruit having more than 10% russeting . Error bars represent standard error, P = >.05. Axis legend: Blossom - Protect + Cueva at 3 sprays (BP+C3), Blossom - Protect + Oxidate at 2 sprays (BP+O2), Blossom - Protect at 2 late sprays (BP+2L), Blossom - Protect + Nordox + Umbrella at 3 sprays (BP+N+U3), Blossom - Protect at 4 sprays (BP+4). A B B B B A b b b b b a 0 10 20 30 40 50 60 BP-4 BP+C3 BP+O2 BP+2L BP+N+U3 USC Percent Infection (%) Treatment Fire Blight Percent Infected Shoots Fire Blight Percent Infected Blossoms BC B C C A C 0 2 4 6 8 10 12 14 BP-4 BP+C3 BP+O2 BP+2L BP+N+U3 USC Percent russeting (%) Treatment 69 4. Discussion Overall there did seem to be some differences in the treatments for Blossom - Protect. One misleading treatment was BP+O2. BP+O2 may have done worse over all in the field, there seemed to be a decrease in fruit set for this treatment (personal observations) . The flowers had a severe phytotoxic reaction to the oxidate spray, causing lower fruit yield. The oxidate spray may have had an effect on scab presence however, as the treatment had lower fungal populations present (Figure 2 - 2) and the incidence of scab was significantly lower (Figure 2 - 6). It might be advantageous to use the oxidate spray, if in fact that no significant decrease in yield is observed, simply to try to replace the fungicides that cannot be applied. Surprisingly, the treatment with the m ost russeting was not BP +O2 with the oxidate application but BP+N+U3, which had the Nordox and Umbrella components. Nordox was observed to have higher incidences of russeting in other treatments in the orchard (personal observations). The treatment that sh ould have shown the highest amount of russeting according to other trials was BP+4, as it had four applications of Blossom - Protect. However though this treatment was not significantly different the control, the treatment did have higher incidence of russet ing on average (Figure 2 - 8). The Cueva component to BP+C3 could have resulted in higher russeting as well. Over all, the treatments that were purely Blossom - Protect did not show any increase in russet from the untreated control. Even though BP+C3 had no E. amylovora present on the plates from the lab (Figure 2 - 7), the incidence of fire blight was still significantly the same as other treatments that contained higher populations of E. amylovora on plates (Figure 2 - 7). It is possible that uneven blossoming could have allowed for blossoms to not come in contact with Blossom - protect and allow 70 infection to occur. It is also possible that the blight that occurred was n ot from the applied Ea110, but from E. amylovora previously sprayed in that particular orchard. The Blossom - Protect treatment did reduce incidence of blossom blight in spring 2014. Laboratory results showed that the yeast is able to compete and survive epiphytically on the flowers. More field trials in different weather seasons are needed to confir m that this product would be good for Michigan growers. 71 CHAPTER 3: AN ATTEMPT AT UV - C MUTAGENESIS OF FUNGAL PATHOGENS OF TREE FRUIT TO CONFER RESISTANCE TO SUCCINATE DEHYDROGENASE INHIB I TORS FUNGICIDES 72 1. Abstract New chemistries of S D HIs have been introduced f o r Michigan fruit trees to help combat fungal pathogens already resistant to other fungicide chemistries. With the advent of these new chemistries, cross - resistance may become a problem and render the new SDHIs useless. By using UV - C mutagenesis, new mutations can be selected in vitro that could allow for laboratory studies of cross - resistance before isolating resistant fungal pathogens in the field becomes a reality. Unfortunately, the UV - C mutage nesis conducted failed to induce mutations in M onilinia fructicola conferring 100% resistance to boscalid or f luxapyroxad. UV - C mutagenesis also failed to produce any level of SDHI fungicide resistance in Blumeriella jaapii or Venturia inaequalis. As a result of the failed mutagenesis, no cross - resistance was detected as well. 2. Introduction 2.1 Background on tree fruit fungicides 2.1.1 Cherry and Other Stone Fruit In stone fruit, there are three board spectrum chemicals used for cherry leaf spot control: captan, chlorothalonil and c opper. Cholorothanlonil cannot be used after chuck spilt, and copper can cause phytotoxicity to trees if applied under drought or low water conditions ((McManus et al., 2007). There are two major fungal diseases of che rry: American brown rot ( Monilinia fructicola ) and Cherry Leaf Spot ( Blumeriella jaapii ). Powdery mildew on cherries ( Podosphaera clandestina ) is also controlled by the fungicides used for these other two diseases. Besides broad spectrum fungicides, sing le - site fungicides are also used in stone fruit production. For cherry leaf spot, demethylation inhibitors (DMIs) were traditionally used all season for control and were effective against American brown rot and powdery mildew 73 (McManus & Weidman, 2001) . Due to consistent spraying of DMIs Cherry leaf spot resistance, especially in Michigan, has been observed and there have been notable failures in control (Sundin et al., 2005; Proffer et al., 2006). There has also been documented cross resistance to multiple fungicides that use DMIs as the mode of action (Proffer et al., 2006). A possibly reason for the resistance is the overexpression of CYP51 , which is known to cause DMI resistance, has been found in DMI resistant B. jaapii isolates (Ma et al., 2006). The o ther major used class of fungicides are strobilurins (QoIs), which are used on American brown rot, powdery mildew and cherry leaf spot, and this class of fungicide also has resistance concerns (McManus et al., 2007). The newest mode of action released fo r fungal disease of stone fruit are succinate d e hydrogenase i nhibitors (SDHIs) and were released in 2004. There is a chance that this mode of action that tree frui t may see in a while, given current fungicide development trends. 2.1.2 Apple In apple , two major diseases are controlled with fungicides: apple scab ( Venturia inaequalis ) and powdery mildew ( Podosphaera leucotricha ). There are other diseases, such as Black rot ( Botryosphaeria obtusa ) or Flyspeck ( Schizothyrium pomi ) that fungicides can also control in the field. Typic ally in a commercial orchard, broa d spectrum fungicides are applied to trees first to protect against initial infection . For apples, thes e broad spectrums include c aptan, which has a seasonal use limit, and e thylenebisdithiocarbamates (EBDCs) which also have a seasonal use limit and a 77 day pre - harvest interval (PHI), meaning that the fungicide cannot be used however many days indicated be fore fruit harvest. 74 After the br oad spectrum fungicides, single - site fungicides are also used in commercial orchards for disease control. There is already evidence for resistance for QoIs in V. inaequalis (Steinfeld et al., 2002), as well as DMIs (Gao et all, 2009). The SDHIs, as like stone fruit, are the newest and possibly the only new chemistry that growers can utilize in their programs for years to come. 2.2 Background of SDHIs The target enzyme for SDHI chemistries is succinate dehydrogenase whic h is part of the tricarboxylic cycle in the mitochondrial transport chain. This enzyme has four subunits which form a binding site by B, C, and D which usually bind ubiquinone; this binding site is the target for SDHI fungicides ( Kuhn, 1984; Sierotzki and Scalliet, 2009; Avenot and Michaildies, 2010; Proffer et al., 2013 ; Sierotzski and Scalliet, 2013 ). There are four SDHI compo unds available for tree fruit: penthiopyrad, b oscalid, f luo pyram, and f luxapyroxad. Boscalid was introduced in Michigan for the 20 04 growing season and is mostly used on cherry for cherry leaf spot and powdery mildew (P roffer et al., 2013). However, b oscalid resistance has already been reported in other fungal diseases (Prof fer et al., 2013; Sierotzski and Scalliet, 2013). Penthiopy rad was intr oduced in Michigan in 2012 and fluxapyroxad and f luopyram were both introduced in Michigan in 2013. Some of these compounds share chemical groups according to the FRAC (Table 3 - 1). As of 2010, there was evidence of no cross - resistance reported between SDHIs and other classes of fungicides, like QoIs (Avenot and Michaildies, 2010). However, cross - resistance between SDHI compounds have bee n reported, especially between b oscalid and penthiopyrad ( Fraaije et al., 2012; Proffer, 2013). However, fluop yram has a different binding chemistry and cross - resistance may take longer to develop (Fraaije et al., 2012). 75 Single - site fungicides typically have high risk of resistance development and one way FRAC and chemical companies are trying to preserve chemist ries from resistance is by selling compounds as premixes. Most new fungicides, including the SDHIs, are sold as these premixes. Unfortunately for tree fruit, there is already resistance present for the second chemistry added to the SDHIs as mentioned above ; premixes are meant for broad use of fungal diseases and crops and thus added selection pressure is placed on SDHI site. In general, the two fungicides in premixes are typically unequal in efficacy, length of use, and overall risk of resistance developmen t. Table 3 - 1: SDHI chemistries available and sorted into chemical group and common name. Common name in bold are for use on tree fruit. Taken from the FRAC website CODE TARGET SITE OF ACTION CHEMICAL GROUP COMMON NAME 7 Complex II; succinate - dehydrogenase Phenyl - benzamides Benodanil Flutolanil Mepronil Pyridinyl - ethyl - benzamide Fluopyram Furan - carboxamides Fenfuram Oxathiin - carboxamides Carboxin Oxycarboxin Thiazole - carboxamides Thifluzamide Pyrazole - carboxamides Bixafen Fluxapyroxad Furametpyr Isopyrazam Penflufen Penthiopyrad Sedaxane Pyridine - carboxamides Boscalid With the emergence of more SDHI chemistries in Michigan, more studies involved in detecting cross - resistance between the compounds is needed to ensure proper rotations and to 76 avoid resistance in general. By using UV - C generated mutants, we can observe if a ny cross - resistance is even possible before finding isolates from the field. 3. Materials and Methods 3.1 Monilinia fructicola 3.1.1 UV - C dose determination for M. fructicola fo r a 80 - 90% conidial kill curve M. fructicola conidia were collected from 3 - 4 day old cultures grown on V8 agar and transferred to 1.7 ml microcentrifuge tubes containing 500 µl of sterile diH 2 0. Three to four glass beads were also placed in the tubes. The tubes were then vortexed for 30 to 60 seconds to break up the conidial chains. The solution was poured into a 14 ml tube that had three layers of sterile cheesecloth plugged into the opening of the tube; this was completed using sterilized forceps. The filtered solution in the bottom of the 14 ml tube after being strained by the cheesecloth was then free of mycelium and large chains of conidia that were not broken by the vortex and beads. This new suspension was quantified using a hemocytometer with a light microscope then adjusted to 1x10 5 conidia /ml by diluting the suspensi on with additional sterile diH 2 0 if needed. The suspension was then placed into a glass petri dish (Pyrex). The bottom of the dish was wrapped in a layer of parafilm to prevent slippage in the subsequent steps of the experiment before the conidia suspensio n was added. To determine the correct dose for a conidia survival rate of 10 - 20%, these cultures were exposed to either ~100, 250, 500, 750, or 1000 J m 2 of UVC (254 nm) radiation from an XX - 15 UV lamp (UVP Products, San Gabriel, CA) placed horizontally a t a fixed height above the conidia suspension. The lamp was turned on 15 minutes prior to use to allow for stabilization of the UV output. The XX - 15 UV lamp was monitored with a UV - X radiometer fitted with a UV - 25 sensor (UVP Products) and determined to be 1.3 J m 2 s 1 (Weigand and Sundin, 2009). The spore solution s were mixed during the UVC dosage by protected hand. After the UV - C dose was given to the conidia, the dishes were sealed and 77 wrapped in heavy grade aluminum foil and placed on a shaker for 24 h rs to prevent photoreactivation. A control glass plate was also used in each round of UV - C exposure and treated exactly the same just not subjected to the UV - C. After the dark incubation period was over, the conidia were recounted using the same method li sted above. The spores were then adjusted if needed to 1x10 3 /µl using sterile diH 2 O and 100 µl of the diluted suspension was transferred to water agar and spread plated. Conidi a germination was counted 12 h rs after plating by using a plate grid and a disse ction microscope to ensure accuracy. The control plates were counted first and the percent germination was determined based off the control plates. 3.1.2 Creating UV - C Mutant Isolates Isolates were subjected to UV - C as described in the dose determination section, however the dose used was ~750 J m 2 of UVC (254 nm) to get 10 - 20% conidia survival rate . After treatment, the conidia suspension was plated onto V8 - S media (V8 strained with cheese cloth then centrifuged and the supernatant was used instead of p ure V8). V8 - S media was ultimately used for M. fructicola as the conidia germinated more uniformly and the color of the media was easier to use in the subsequent steps. After the dark period for the reduced risk of phytoremediation mentioned in Section 3.1.1, 100 µl of treated or control conidia suspension was sp read onto 10 ml V8 - S medium and allowed to dry for 30 minutes. The plates were then treated with a 10 ml overlay (1/5 V8 - S and .7% bacto agar) dosed with 10 µ g/ml b oscalid. Plates were then sealed and wrapped in aluminum foil and left for 12 hours. Plates were then examined for mycelium protruding out of the overlay. Any mycelium found growing through the overlay were harvested with sterile forceps and transferred to fresh PDAY (Potato dextrose agar with .5g yeast) amended with 10 µ g/ml b oscalid. These plates were monitored for growth, and if growth occurred, transferred to long term storage (PDAY slants stored at 4 o C). 78 3.1.3 Screening M utant I solates for Fungicide R esistance After 25 isolates were harvested from the amended 10 µ l/ml boscalid PDAY media, more rigorous fungicide testing began. These isolates were plug - plated onto 10, 15, 25, and 35 µg /ml boscalid or f luxapyroxad amended PDAY in triplicate. Growth was monitored daily and diameter measurements were taken each day until the contro l plates (un - am ended PDAY) filled the plates. The original isolate that the conidia was originally harvested from was also used as a control in these screenings. This screen was repeated twice. 3.2 V. inaequalis and B. jaapii Mutant Isolation and Screening The procedures used in the M. fructicola methods section were also subjugated upon V. inaequalis and B. jaapii with the UV - C dose determined to be 1000 and 750 J m 2 of UV - C (254 nm) respectively (results not shown) . The spore counts dosed were also lower, averaging around 1x10 4 conidia/ µ l due to the slow growth of these fungi in culture . When plating to count percent survival, Malt Dextrose agar (MMEA) and PDAY were used. Even though spores germinated and a dose was found, 10 - 15 runs, producing 10 - 20 plates a run, of UV - C followed with the 10 µg/ml boscalid overlay failed to produce any resistant mutants. Plates were even stored and checked for at least two months with no sign of germination. 4. Featured M. fructicola Results The optimal dose for M. fructicola isolate SCHM13 was determined to be 750 J/M 2 as the results on average gave a c onidial survival rate between 10 - 20% (Figure 3 - 1). After the first isolate kill curve was obtained, the other isolates were screened in a smaller window of doses closer to the doses that were effective with M. fructicola isolate SCHM13 (Figure 3 - 2). Another example of finding this kill curve was for isolate M. fructicola isolate BR36, where there was not 79 as many repeats carried out to determine the optimal dose (Figure 3 - 2). For all M. fructicola isolates, the dose of 750 J/M 2 delivered the 10 - 20% conidia survival rate. An example of the lack of finding mutants for M. fructicola is featured in Table 3 - 1. From ten rounds of UV - C completed between 8/1/13 and 9/25/13, 25 mutants were discovered (Table 3 - 1). After the first round of UV - C, only two M. fructicola isolates were treated a round due to time constraints. The rates of finding mutants were consistently low, as anywhere from 400 to 800 spores were screened at a time per isolate (Table 3 - 1). In the prelim inary screening of boscalid and f luxa pyroxad resistance in M. fructicola isolate SCHM13 and generated SCHM13 mutants there were differences comparing the mutants to the original SCHM 13 isolate , however none of the mutants were at the same growth level as the un - amended media control. (Figure 3 - 3). Two of the mutants featured in this figure, UV 3 and UV 4, were closer achieving the same colony diameter as the un - amended control, but still were statistically reduced in size (Figure 3 - 3). These same mutants were screening at higher rates of fun gicide, 15, 25, and 35 µg/ml boscalid and 15, 25, 35 µg/ml f luxapyroxad (Figure 3 - 4). These results however, indicate that while all of the mutants grew significantly larger than the original SCHM 13 isolate on boscalid amended media, they were statistically lower in diameter than the un - amended control plate (Figure 3 - 4). The high fl uxapyroxad doses were even less exciting, with many of the UV mutants having statistically similar growth as the SCHM 13 isolate (Figure 3 - 4). Though more resistant than the original SCHM13 isolate, which some mutants reaching over 200% more growth than the SCHM13 isolate on boscalid amended media, the mutants did not have the same relative growth levels as themselves or SCHM13 on un - amended media (Table 3 - 2). The hig her rates of boscalid and fl uxapyroxad had similar or even less drastic 80 relative growth levels (Tables 3 - 3 and 3 - 4). After this screening, M. fructicola isolates BROM72 and BR36 were dropped from this study as their mutants had even less growth than the SC HM13 mutants in relative growth differences (Data not shown). Figure 3 - 1: Survival rates of M. fructicola isolate SCHM13 conidia exposed to various doses of UV - C to determine a kill curve. Figure 3 - 2: Survival rates of M. fructicola isolate BR36 conidia exposed to various doses of UV - C to determine a kill curve. 0% 10% 20% 30% 40% 50% 60% 70% 80% 90% 100% 0 200 400 600 800 1000 1200 1400 Survival Rate (%) UV - C Dosage (J/M 2 ) 0% 10% 20% 30% 40% 50% 60% 70% 0 200 400 600 800 1000 1200 1400 Survival rate (%) UV - C Dosage (J/M2) 81 Table 3 - 2: The number of mutants isolated from three M. fructicola isolates (RBOM72, SCHM13, and BR36) from various rounds of UV - C treatments. Each round of UV - C used a fre sh conidial suspension of two or three strains . When the total number of screened mutants got to 25, further fungicide resistance screening was conducted on the mutants. Round of UV - C RBOM72 SCHM13 BR36 1 (8/1/13) 2 1 3 2 (8/15/13) - 1 1 3 (8/29/13) 1 1 - 4 (9/1/13) - 1 1 5 (9/14/13) 2 1 - 6 (9/15/13) 1 2 - 7 (9/21/13) - 1 0 8 (9/22/13) - 2 0 9 (9/24/13) 1 1 - 10 (9/25/13) 0 2 - Total 7 13 5 Figure 3 - 3: Preliminary screening of boscalid and fluxapyroxad resistance in M. fructicola isolate SCHM13 and generated SCHM13 mutants . B10 represents a dose of 10 µg/ml boscalid ; F10 represents a dose of 10 µg/ml f luxapyroxad; C represents an un - amended control. Error bars represent standard error, P =<.05, same letters on the same color bar re presents no significant difference. A A A A A B B A AB B C AB A A C 0 1 2 3 4 5 6 7 B10 F10 C B10 F10 C B10 F10 C B10 F10 C B10 F10 C UV 1 UV 2 UV 3 UV 4 SCHM13 COLONY DIAMETER (CM) 82 Figure 3 - 4: Higher doses of boscalid and fluxapyroxad resistance in M. fructicola isolate SCHM13 and generated SCHM13 mutants . B15, B25, B35 represents a dose of 15, 25, and 35 µg/ml boscalid ; F15, 25, 305 represents a dose of 15, 25, 35 µg/ml f luxapyroxad; C represents an un - amended control. Error bars represent standard error, P =<.05, same letters on the same color bar represents no significant difference. Table 3 - 3: Preliminary screening doses of boscalid and fluxapyroxad resistance in M. fructicola isolate SCHM13 and generated SCHM13 mutants . B10 represents a dose of 10 µg/ml boscalid ; F10 represents a dose of 10 µg/ml f luxapyroxad; C represents an un - control dose on un - amended agar. B AB CD A A BC AB C D E EF F A A A A CB B CD A B BCD B BCD F CD EF F 0 1 2 3 4 5 6 7 8 9 B15 B25 B35 F15 F25 F35 C B15 B25 B35 F15 F25 F35 C B15 B25 B35 F15 F25 F35 C B15 B25 B35 F15 F25 F35 C UV 2 UV 3 UV 4 SCHM13 COLONY DIAMETER (CM) 83 Table 3 - 4: Higher screening doses of boscalid for res istance in M. fructicola isolate SCHM13 and generated SCHM13 mutants . B15, B25, B35 represents a dose of 15, 25, and 35 µg/ml b oscalid ; C represents an un - percent of relative growth to the same mutant isolate grown on un - amended agar. Table 3 - 5: Higher screening doses of fluxapyroxad for resistance in M. fructicola isolate SCHM13 and generated SCHM13 mutants . F15, 25, 305 represe nts a dose of 15, 25, 35 µg/ml f luxapyroxad; C represents an un - dicates the percent percent of relative growth to the same mutant isolate grown on un - amended agar. 84 5. Discussion None of the mutant isolates generated by UV - C of M. fructicola had comparable growth to control isolates or plates, and therefore did not express full resistance to the fungicides. There was also no evidence of cross - resistance in the mutant isolates screened. Though kill curves were established, no mutants were obtained that had any leve l of resistance to boscalid or f luxapyroxad from B. jaapii or V. inaequalis. The UV - C treatments given were ineffective at creating 100% resistant isolates to any of t he SDHI compounds tested. 85 WORK S CITED 86 WORKS CITED 1. Avenot, H., and Michailides, T. 2010. Progress in understanding molecular mechanisms and evolution of resistance to succinate dehydrogenase inhibiting (SDHI) fungicides in phytopathogenic fungi. Crop Protection 29 :643 - 651. 2. Bachmann, F.M. 1913. The migration of bacillus amylovorus in the host tissues. Phytopathology 3 : 3 - 13. 3. Bauske, R.J. 1968. Bacterial strands: a possible role in fire blight. Iowa State Journal of Science. 43 : 119 - 124. 4. Beer, S.V., and Norelli, J.L. 1977. Fire blight epidemiology: factors affecting release of Er winia amylovora by cankers. Phytopathology 67 : 1119 - 1125. 5. Bennett, A.R. and Billing, E. 1978. Capsulation and virulence in Erwinia amylovora . Annals of Applied Biology 89 : 41 - 45. 6. Bennett, A.R. and Billing, E. 1979. Origin of the polysaccharide component of ooze from plants infected with Erwinia amylovora . Journal of General Microbiology 116 : 341 - 349. 7. Billing, E. 2011. Fire blight. Why do views on host invasion by Erwinia amylovora differ? Plant Pathology 60 : 178 - 189. 8. Blachinsky, D., Shtienburg, D., Oppenheim D., Zilberstaine M., Levi S., Zamski, E., and Shoseyow O. 2003. The role of autumn infection in the progression of fire blight symptoms in perennial pear branches. Plant Disease 87 : 1077 - 1082. 9. Boges, J . Bruchmuller, I. Erbar, C., and Geider, K. 1998. Colonization of host plants by the fire blight pathogen Erwinia amylovora, labeled with genes for bioluminescence and fluorescence. Phytopathology 88 : 416 - 421. 10. Castoria, R., De Curtis, F., Lima, G., Caput o, L., Pacifico, S., and De Cicco, V. 2001. Aureobasidium pullulans (LS - 30) an antagonist of postharvest pathogens of fruits: study on its modes of action. Postharvest Biology and Technology. 22 : 7 17. 11. Crosse, J.E., Goodman, R.N., and Shaffer, W.H. 1972. Leaf damage as predisposing factor in the infection of apple shoots by Erwinia amylovora. Phytopathology 62 : 176 - 182. 87 12. Droby, S., Cohen, L., Daus, A., Weiss, B., Horev, B., and Chalutz, E. 1998. Commercial testing of Aspire: a yeast preparation for the biological control of postharvest decay of citrus. Biological Control. 12 : 97 101. 13. Duffy, B., Vogelsanger, J., Schoch, B., and Holliger, E. 2006. Biocontrol of Erwinia amylovora using a commercial yeast strain mixture. Acta Horticulturae . (ISHS) 704: 363 - 366 . 14. Eden - Green, S.J., and Billing, E. 1972. Fireblight: Occurrence of bacterial stands on various hosts under glasshouse conditions. Plant Pathology 21 : 121 - 123. 15. Eden - G reen, S.J., and Knee, M. 1974. Bacterial exopolysaccharide in fireblight exudate. Journal of General Microbiology 81 : 509 - 512. 16. Fraaije, B.A., Bayon, C., Atkins, S., Cools, H.J., Lucas, J.A., and Fraaije, M.W. 2012. Risk assessment studies on succinate dehydrogenase inhibitors, the new weapons in the battle to control Septoria leaf blotch in wheat. Molecular Plant Pathology 13 : 263 - 275. 17. Gao, L., Berrie, A., Yang, J., and Xu, X. 2009. Within - and between - orchard variability in the sensitivity of Venturia inaequalis to myclobutanil, a DMI fungicide, in the UK.Pest Management Sci ence. 65 :11: 1241 - 1249. 18. Geider, K., Geier, G., Bellemann, P., Bernhard, F., Bugert, P., and Metzger, M. 1993. Exopolysaccharides in pathogenicity of Erwinia amylovora . Acta Horticulturae 338 : 255 - 262 19. Geier, G., and Geider, K. 1993. Characterization and influence on virulence of the levansucrase gene from the fire blight pathogen Erwinia amylovora . Physiological and Molecular Plant Pathology 42 : 387 - 404. 20. Goodman, R. N., Huang, J. S., and Huang, P - Y. 1974. Host - specific phytoto xic polysaccharide from apple tissue infected by Erwinia amylovora . Science, 183 : 1081 - 1082. 21. Gowda, S.S. and Goodman, R.N. 1970. Movement and persistence of Erwinia amylovora in shoot, stem, and root of apple. Plant Disease Reporter 54 : 576 - 580. 22. Granatstein, D . 2014. Fire blight control for organic orchards: moving beyond antibiotics. Washington State Extension. http://www.extension.org/pages/70541/fire - blight - control - for - organic - orchards: - moving - beyond - antibiotics#.VDXbDfldW70 88 23. Gross, M. G. Geier, K. Rudolp, and K. Geider . 1992. Levan and levansucrase synthesized by the fireblight pathogen Erwinia amylovora . Phy siological and Molecular Plant Pathology 40 : 371 - 381. 24. Guttman, D.S., Vinatzer B.A., Sarkar, S.F., Kettler, G., and Greenberg, J.T. 2002 A Functional Screen for the Type III (Hrp) Secretome of the Plant Pathogen Pseudomonas syringae. Science. 295 : 1722 - 1726 25. Hildebrand, E.M. 1939. Studies on fire - blight ooze. Phytopathology. 29 : 142 - 156. 26. Hildebrand, M. Dickler, E., and Geider, K. 2000. Occurrence of Erwinia amylovora on insects in a fire blight orchard. Journal of Phytopathology. 148 : 251 - 256. 27. Hockenhull, J. 1974. Some anatomical and pathological features of healthy and diseased Hawthorn ( Cataegus monogyna ) and naturally infected by the fireblight pathogen Erwinia amylovora . The Royal Veterinary and Agricultural University, Yearbook 1974: 125 - 136 . 28. Huang, P.Y., Goodman, R.N. 1977. Ultrastructural modifcations in the apple stems induced by Erwinia amylovora and the fire blight toxin. Phytopathology 66 : 269 - 276. 29. Jones, A.L. and Schnabel, E.L. 2000. The development of streptomycin - resistant strains of Erwinia amylovora . p.235 - 251. In: J.L. Vanneste (ed.), Fire blight: the disease and its causative agent, Erwinia amylovora . CABI Publishing CAB International, Wallingford, UK. 30. Keil, H.L. and van der Zwet, T. 1972. Aerial strands of Erwinia amylovora : structure and enhanced production by pesticide oil. Phytopathology 62 : 355 - 361. 31. Koczan, J. M., B. R. Lenneman, M. J. McGrath, and Sundin, G.W. 2011. Cell surface attachment structures contribute to biofilm formation and xylem colonization of Erwinia amy lovora . Applied and Environmental Biology 77 : 7031 - 7039. 32. Koczan, J.M., McGrath, M.J., Zhao, Y., and Sundin, G.W. 2009. Contribution of Erwinia amylovora exopolysaccharides amylovoran and levan to biofilm formation: implications in pathogenicity. Phytopat hology. 99 :11: 1237 - 1244. 33. Kuhn P. 1984. Mode of action of carboximides. Symposium Series of the British Mycological Society. 9 :155: 183 . 34. Kunz, S. - - a yeast preparation for the reduction of blossom infections by fire blight. In: 11 th International conference on cultivation technique and phytopathological problems in organic fruit - growing and viticulture. 35. Lewis, S. and Goodman, R.N. 1965. Mode of penetration and movement of fire blight bacteria in apple leaf and stem tissue. Phytopathology 55 : 719 - 723. 89 36. Lima, G., De Curtis, F., Castoria, R., and De Cicco, V. 2003. Integrated control of apple postharvest pathogens and survival of bio control yeasts in semi - commercial conditions. European Journal of Plant Pathology . 109 : 341 - 349. 37. Longstroth, M. 2001. The 2001 fire blight epidemic in southwest Michigan orchards. The Compact Fruit Tree. 31 : 16 - 19. 38. Luo, C.X. and Schnabel, G. 2008. Adapt ation to fungicides in Monilinia fructicola isolates with different fungicide resistance phenotypes. Phytopathology. 98: 2: 230 - 238. Norelli, J., H. Aldwinckle, T. Momol, B. Johnson, A. DeMarree ,and M.V. Bhaskara Reddy. 2000. Fire blight of apple rootstocks. New York Fruit Quarterly 8 : 2 - 5. 39. Norelli, J.L., Jones, A.L., and Aldwinkle, H.S. 2003. Fire blight management in the twenty - first Century. Plant Disease. 87 : 756 - 765. 40. Ma, Z., Proffer, T.J., Jacobs, J.L. a nd Sundin, G.W. 2006. Overexpression of the 14 - demethylase (CYP51) target gene mediates fungicide resistance in Blumeriella jaapii. Applied and Environmental Microbiology. 72 : 2581 - 2585. 41. Malnoy, M., Martens, S., Norelli, J.L., Barny, MA,. Sundin, G.W., S mits, T.H.M., and Duffy, B. 2012. Fire blight: applied genomic insights of the pathogen and the host. Annual Review of Phytopathology 50 : 475 - 94. 42. McGhee , G.C., Guasco, J. , Bellomo , L.M., Blumer - Schuette, S.E. , Shane, W.W. , Irish - Brown, A. , and Sundin, G.W. 2011. Genetic analysis of streptomycin - resistant (Sm R ) strains of Erwinia amylovora suggests that dissemination of two genotypes is responsible for the current distribution of Sm R E. amylovora in Michigan. Phytopathology. 101 : 182 - 191. 43. McGhee, G.C., Sundin, G.W. 2012 Erwinia amylovora CRISPR Elements Provide New Tools for Evaluating Strain Diversity and for Microbial Source Tracking. Plos One. 7 :7 44. McManus, P.S. and Jones, A.L. 1994. Role of Wind - driven Rain, aerosols, and contaminated budwood in incidence and special pattern of fire blight in an apple nursery. Plant Disease 87 : 1059 - 106. 45. Mills, W. 1955. Fire Blight development on apple in western New York. Plant Disease Reporter 39 : 206 - 207. 46. Momol, M.T., Ugine, J.D., Norelli, J.L., and Aldwi nkle, H.S. 1999. The effect of prohexadione calcium, SAR inducers and calcium on the control of shoot blight caused by Erwinia amylovora on apple. Acta Horticulturae. 489 : 601 - 606. 47. Morris, C.E. and Monier, J.M. 2003. The ecological significance of biofi lm formation by plant - associated bacteria. Annual Review of Phytopathology. 41 : 429 - 453. 90 48. Oh, C. - S. and Beer, S.V. 2005. Molecular genetics of Erwinia amylovora involved in the development of fire blight. Microbiology Letters. 253 : 185 - 192. 49. Pal, K. K. and McSpadden Gardener, B. 2006. Biological control of plant pathogens. The Plant Health Instructor. 1 - 20. 50. Parker, K.G., Fisher, E.G., and Mills, W.D. 1956. Fire Blight on pome fruits and its control. New York (Cornell) College of Agriculture E xtension Bulletin 966. 51. Proffer, T.J., Berardi, R., Ma, Z., Nugent, J.E., Ehret, G.R., McManus, P.S., Jones, A.L. and Sundin, G.W. 2006. Occurrence, distribution, and PCR - based detection of resistance to sterol demethylation - inhibiting fungicides in populations of Blumeriella jaapii in Michigan. Phytopathology, 96 : 709 - 717. 52. Quinones, B., Dulla, Q.,and Lindow, S.E. 2005. Quorum sensi ng regulates exopolysaccharide production, motility, and virulence in Pseudomonas syringae . Molecular Plant - Microbe Interactions. 18 : 682 - 693. 53. Renzi, M., Copini, P., Taddei, A.R., Rossetti, A., Gallipoli, L. Mazzaglia, A., and Balestra, G.M. 2012. Bacter ial canker on kiwifruit in Italy: anatomical changes in the wood and in the primary infection sites. Phytopathology. 109 : 827 - 840. 54. Rosen, H.R. 1936. Mode of penetration and progressive invasion of fire blight bacteria into apple and pear blossoms. Univers ity Arkansas College of Agriculture, Agriculture Experiment Station Bulletin No. 331. 55. Schouten, H.J. 1988. Notes on the role of water potential in the pathogenesis of fire blight, caused by Erwinia amylovora . Netherlands Journal of Plant Pathology 94 : 21 3 - 220. 56. Schouten, H.J. 1989. A possible role in pathogenesis for the swelling of extracellular slime of Erwinia amylovora at increasing water potential. Netherland Journal of Plant Pathology. 95 : 169 - 174. 57. Schouten, H.J. 1991. Simulation of potential caused by multiplication and swelling of Erwinia amylovora in intercellular space of host tissue. Netherland Journal of Plant Pathology. 97 : 139 - 141. 58. Schroth, M. N., Thompson, S.V., Hildebrand, D.C., and Moller, W.J. 1974. Epidemiology and control of fire blight. Ann. Rev. Phytopathol. 12 : 389 - 412. 59. Seemuller, E.A. and Beer, S.V. 1976. Absence of cell wall polysaccharide degradation of Erwinia amylovora. Phytopathology 66 : 433 - 436. 91 60. Southey, R.F.W. and Harper, G.J. 1971. Th e survival of Erwinia amylovora in air - borne particles: tests in the laboratory and in the open air. Journal of Applied Bacteriology 34 : 547 - 556. 61. Steiner, P.W. and Lightner, G. 1992. MARYBLYT: A predictive program for forecasting fire blight disease in apples and pears. Version 4.0 copyrighted computer program, University of Maryland, College Park, MD 20742. 62. Steinfeld, U., Sierotzki, H., Parisi, S., and Gisi, U. 2002. Comparis on of resistance mechanisms to strobilurin fungicides in venturia aequalis. Modern fungicides and antifungal compounds II 12th internation reiharsbrunn symposium. 167 - 176. 63. Studies on the overwintering and mode of infection of the fire blight organism. Mic higan Agricultural Station Bulletin 92 : 33 64. Sundin, G. W., Ehret, G. R., McManus, P. S., Nugent, J. E., Klein, W. M., Anderson, M. D. and Proffer, T. J. 2005. Evaluation of new fungicide rotations for cherry leaf spot control on Montmorency tart cherries in NW Michigan, 2004. Fungicide and Nematicide Tests, 60 : STF004. 65. Sundin, G.W., Werner, N.A., Yoder, K.S., and Aldwinckle, H.S. 2009. Field evaluation o f biological control of fire blight in the eastern United States. Plant Disease. 93 : 986 - 394. 66. Thomson, S.V. 1986. The role of the stigma in fire blight infections. Phytopathology 76 : 476 - 482. 67. United States Department of Agriculture (USDA). 2002. United State Standards for Grades of Apples. §51.302 U.S. 1 : 2. http://www.ams.usda.gov/AMSv1.0/getfile?dDocName=STELPRDC5050339. 68. van der Zwet, T.B. 1969. Study of fire blight cankers and associated bacteria in pear. Phytopathology 59 : 607 - 613. 69. van de Zwet, T.B and Beer, S.V. 1995. Fire blight - its nature, prevention, and control. Agriculture Information Bulletin No. 613. 70. Van der Zwet, T., Orolaza - Halbrendt, N., and Zeller, W. 2012 Fire blight history, biology, and management. American Phytopathological Soc iety publication. 71. Vanneste, J.L. 2000. Migration of Erwinia amylovora in host plant. In: Vanneste, J.L., ed. Fire Blight: the disease and its causative agent, Erwinia amylovora. CABI Publishing, Oxon., UK. 92 72. Vanneste, J.L. 2011. Biological control agents of fire blight: successes and challenges. ISHS Acta Horticulturae 896: XII International Workshop on Fire Blight . 409 - 416. 73. Vieira, Lelis, F.M., Czajkowski, R., M agela de Souza, R., Ribeiro, D.H., and van der Wolf, J.M. 2013. Studies on the colonization of axenically grown tomato plants by a GFP - tagged strain of Clavibacter michiganenesis subsp. michiciganensis European Journal of Plant Pathology. Plant Disease. 87 : 756 - 765. 74. Weigand, M.R., and Sundin, G.W. 2009. Long - term effects of inducible mutagenic DNA repair on relative fitness and phenotypic diversification in Pseudomonas cichorii 302959. Genetics. 181 : 199 - 208. 75. Wilson, M., Epton, H.A.S., and Sigee, D.C. 1987. Ultrastructurant studies on fire blight of hawthorn flowers. Acta Horticulturae 217 : 189 - 194. 76. Zamski, E., Shitenberg, D., and Blachinsky, D . 2006. The role of ooze exudation in the migration of Erwinia amylovora cells in pear trees infected by fir e blight. Israel Journal of Plant Sciences. 54 : 301 - 307. 77. Zhao, Y.F., Blumer, S.E. and Sundin, G.W. 2005 Identification of Erwinia amylovora genes induced during infection of immature pear tissue. Journal of Bacteriology . 187 : 8088 - 8103.