"Science, Faculty of"@en . "DSpace"@en . "UBCV"@en . "Romero Guzm\u00E1n, Atenas Sofia"@en . "2020-09-08T15:27:08Z"@en . "2020"@en . "Master of Science - MSc"@en . "University of British Columbia"@en . "Honey bees are the most critical crop pollinators worldwide and for the last decades mostly a single antibiotic has been used to protect them from American foulbrood (AFB), a lethal larval disease caused by the bacterium Paenibacillus larvae. Oxytetracycline (OTC) is the only approved prophylactic antibiotic for P. larvae infection in Canada, which has triggered resistant strains. While only the first larvae stages are vulnerable to AFB, consequences are lethal. Early microbiology, genomics and proteomics studies have identified molecular differences between the AFB resistant honey bee stages and the vulnerable larvae. Profiles of the immune factors, such as antimicrobial peptides (AMPs), and a characteristic commensal gut bacteria (CGB) community are key differences. Through this work we want to develop a new prophylactic for AFB, based on the molecular differences between the adult honey bee and the larvae. Specifically, a commensal bacterium strain that expresses and secretes naturally occurring AMPs targeting P. larvae vegetative cells. To identify AMPs candidates to be used as potential prophylactics, we pursued to uncover host-microbe specific responses of the adult honey bee. Through susceptibility assays we tested the inhibitory activity of characteristic honey bee and fruit fly naturally occurring AMPs against P. larvae, CGB (Bartonella apis, S. alvi and Bifidobacterium asteroides), and non-commensal bacteria (Bacillus subtills and Escherichia coli). To look for potential dysbiosis caused on the gut microbiota caused by the candidate AMPs we conducted an in vivo feeding experiment and proceeded with quantitative PCR (qPCR) and 16S rRNA deep amplicon sequencing, determining changes in size and composition. Jelleine and melittin (AMPs) were chosen as final candidates to attempt to be expressed and secreted by S. alvi. Then, we predicted the Sec pathway-dependent extracellular proteins and with a mass spectrometry based-proteomics experiment explored the signal peptides secretion dynamics. Finally, we suggested the signal peptides to be used for the expression and secretion of jelleine and melittin in S. alvi., offering a pioneering approach in the understanding and application of honey bee gut commensals."@en . "https://circle.library.ubc.ca/rest/handle/2429/75906?expand=metadata"@en . "A Prophylactic Probiotic to Fight Paenibacillus larvae Infection in Honey Bees by Atenas Sof\u00C3\u00ADa Romero Guzm\u00C3\u00A1n B.Sc., Monterrey Institute of Technology, 2017 A THESIS SUBMITTED IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF MASTER OF SCIENCE In THE FACULTY OF GRADUATE AND POSTDOCTORAL STUDIES (Genome Science and Technology) THE UNIVERSITY OF BRITISH COLUMBIA (Vancouver) August 2020 \u00C2\u00A9 Atenas Sof\u00C3\u00ADa Romero Guzm\u00C3\u00A1n, 2020 ii The following individuals certify that they have read, and recommend to the Faculty of Graduate and Postdoctoral Studies for acceptance, a thesis entitled: A Prophylactic Probiotic to Fight Paenibacillus larvae Infection in Honey Bees submitted by Atenas Sof\u00C3\u00ADa Romero Guzm\u00C3\u00A1n in partial fulfillment of the requirements for the degree of Master of Science in Genome Science and Technology Examining Committee: Leonard J. Foster, Biochemistry and Molecular Biology Supervisor Steven Hallam, Microbiology and Immunology Supervisory Committee Member Franck Duong, Biochemistry and Molecular Biology Supervisory Committee Member Thibault Mayor, Biochemistry and Molecular Biology Additional Examiner iii Abstract Honey bees are the most critical crop pollinators worldwide and for the last decades mostly a single antibiotic has been used to protect them from American foulbrood (AFB), a lethal larval disease caused by the bacterium Paenibacillus larvae. Oxytetracycline (OTC) is the only approved prophylactic antibiotic for P. larvae infection in Canada, which has triggered resistant strains. While only the first larvae stages are vulnerable to AFB, consequences are lethal. Early microbiology, genomics and proteomics studies have identified molecular differences between the AFB resistant honey bee stages and the vulnerable larvae. Profiles of the immune factors, such as antimicrobial peptides (AMPs), and a characteristic commensal gut bacteria (CGB) community are key differences. Through this work we want to develop a new prophylactic for AFB, based on the molecular differences between the adult honey bee and the larvae. Specifically, a commensal bacterium strain that expresses and secretes naturally occurring AMPs targeting P. larvae vegetative cells. To identify AMPs candidates to be used as potential prophylactics, we pursued to uncover host-microbe specific responses of the adult honey bee. Through susceptibility assays we tested the inhibitory activity of characteristic honey bee and fruit fly naturally occurring AMPs against P. larvae, CGB (Bartonella apis, S. alvi and Bifidobacterium asteroides), and non-commensal bacteria (Bacillus subtills and Escherichia coli). To look for potential dysbiosis caused on the gut microbiota caused by the candidate AMPs we conducted an in vivo feeding experiment and proceeded with quantitative PCR (qPCR) and 16S rRNA deep amplicon sequencing, determining changes in size and composition. Jelleine and melittin (AMPs) were chosen as final candidates to attempt to be expressed and secreted by S. alvi. Then, we predicted the Sec pathway-dependent extracellular proteins and with a mass spectrometry based-proteomics experiment iv explored the signal peptides secretion dynamics. Finally, we suggested the signal peptides to be used for the expression and secretion of jelleine and melittin in S. alvi., offering a pioneering approach in the understanding and application of honey bee gut commensals. v Lay Summary Honey bees are the most critical crop pollinators worldwide and for the last decades mostly a single effective antibiotic has been used to protect them from American foulbrood (AFB), a lethal larval disease caused by the bacteria Paenibacillus larvae. AFB endangers the progeny of the colony. Thanks to the latest advances in molecular biology, differences between the AFB-resistant adult honey bee and the larvae have been identified, highlighting immune factors known as antimicrobial peptides and commensal microorganisms that prevent pathogens from colonizing. Taking advantage of these differences, we try to develop a therapeutic probiotic composed of a commensal bacterium engineered to synthesize and secrete antimicrobial peptides naturally present in the resistant adult honey bee. We identified that these antimicrobial compounds are specific and safe and could probably be used to fight P. larvae. Tools were developed to aid the commensal bacteria genetic engineering. vi Preface I came to Foster lab with the idea to engineer the commensal bacteria of the honey bees and developed entirely the project outline and overall experimental approach. More specifically I designed all the experiments, except for the secretion ratio experiment which was mostly designed by my supervisor Leonard Foster. Technical feedback on the proteomics experiments were done by Nikolay Stoynov, Jenny Moon and Leonard Foster. A couple of paragraphs of the introduction (sections 1.1.1, 1.3.1, 1.3.2 and 1.3.3) were partially taken from the published comprehensive review I first co-authored with Alexandra Nastasa and the contribution of few others. Romero, S., Nastasa, A., Chapman, A., Kwong, W. K. & Foster, L. J. The honey bee gut microbiota: strategies for study and characterization. Insect Mol. Biol. 28, 455\u00E2\u0080\u0093472 (2019). For chapter two, all the experiments were carried by myself for most of the load of work with help of Jessica Seropian for susceptibility assays and dysbiosis experiment during the first trials. Sanger sequencing was done by the Sequencing/Bioanalysis Core Facility within UBC CMMT. Microbiome Insights was hired for the amplicon deep sequencing and microbiome analysis. Sections co 2.3.2.5 were provided by Microbiome Insights. Likewise, we took figures 2.4 and 2.17 from their microbiome analysis. The rest data analysis was done by myself, except for section 2.4.1.2, for which Greg Stacey designed the statistical model for the analysis and contributed with the figures of the same section (2.4.1.2). Mass spectrometry instruments managing and sampling running were performed Nikolay Stoynov. vii Table of Contents Abstract ......................................................................................................................................... iii Lay Summary .................................................................................................................................v Preface ........................................................................................................................................... vi Table of Contents ........................................................................................................................ vii List of Tables .............................................................................................................................. xiii List of Figures ............................................................................................................................. xiv List of Abbreviations ...................................................................................................................xx Glossary ...................................................................................................................................... xxi Acknowledgements ................................................................................................................... xxii Dedication ................................................................................................................................. xxiv Chapter 1: Introduction ................................................................................................................1 1.1 The basics of the global pollinator: The honey bee ........................................................ 1 1.1.1 Relevance ................................................................................................................ 1 1.1.2 General honey bee biology ..................................................................................... 2 The castes ............................................................................................................ 2 Development ....................................................................................................... 3 Diet ...................................................................................................................... 4 Genetics............................................................................................................... 5 1.1.3 The endangered health of the honey bee: pathogens and other threats................... 5 1.2 American foulbrood: Implications and mechanism of action of the most virulent bacterial disease of honey bees ................................................................................................... 7 viii 1.2.1 Mechanisms of infection and transmission ............................................................. 7 1.2.2 Prophylactics for P. larvae infection .................................................................... 11 Oxytetracycline resistance crisis ....................................................................... 11 Alternative strategies to fight P. larvae infection ............................................. 11 Probiotics .......................................................................................................... 13 Antimicrobial peptides ...................................................................................... 14 1.2.3 Honey bee larvae vulnerability ............................................................................. 14 1.3 Honey bee gut commensal bacteria: Description of the community ............................ 16 1.3.1 Composition .......................................................................................................... 16 1.3.2 Establishment of the microbiome ......................................................................... 17 1.3.3 Influence on bee health ......................................................................................... 19 1.4 Adult worker immune system ....................................................................................... 20 1.4.1 Social immunity .................................................................................................... 20 1.4.2 Individual immunity.............................................................................................. 21 Antimicrobial peptides ...................................................................................... 22 1.5 Hypothesis and aims ..................................................................................................... 23 1.5.1 Research goal ........................................................................................................ 23 1.5.2 Aims: ..................................................................................................................... 23 1.5.3 Hypotheses ............................................................................................................ 24 Hypothesis 1...................................................................................................... 24 Hypothesis 2...................................................................................................... 25 Hypothesis 3...................................................................................................... 26 ix Chapter 2: Exploring the molecular and biological toolbox of the adult honey bee against Paenibacillus larvae infection ......................................................................................................28 2.1 Introduction and Rationale ............................................................................................ 28 2.2 Research Goals.............................................................................................................. 30 2.3 Methods......................................................................................................................... 32 2.3.1 Susceptibility specificity assays: Zone of inhibition and MIC microdilution assays 32 Antimicrobial peptides synthesis and solubilization......................................... 32 Growth of bacterial strains and zone of inhibition assay for non-commensal bacteria 34 Broth microdilution assay on 96-well plate ...................................................... 34 Zone of inhibition for commensal gut bacteria ................................................. 35 2.3.1.4.1 Isolation of the honey bee commensal gut bacteria by enrichment culture 35 2.3.1.4.2 Taxonomy identification ............................................................................. 36 2.3.1.4.3 Growth of bacterial strains and zone of inhibition assay ............................ 37 2.3.2 In vivo effects on the gut microbiota when exposed to jelleine and melittin ....... 37 Establishing the commensal gut bacteria of newly emerged bees .................... 38 Antimicrobials feeding...................................................................................... 38 DNA extraction ................................................................................................. 38 qPCR ................................................................................................................. 39 Illumina sequencing .......................................................................................... 39 2.3.2.5.1 Quality Control ........................................................................................... 40 2.3.2.5.2 Statistical analysis ....................................................................................... 41 x 2.4 Results & Discussion .................................................................................................... 41 2.4.1 Susceptibility of non-commensal and commensal bacteria to naturally occurring antimicrobial peptides ........................................................................................................... 41 P. larvae is susceptible to jelleine and melittin ................................................ 42 Non-pathogen and non-colonizer bacteria susceptibility to honey bee and fruit fly naturally occurring AMPs ........................................................................................... 43 Susceptibility of commensal bacteria to naturally occurring AMPs ................ 44 The CGB are more resistant to the naturally occurring AMPs than non-commensal and pathogenic bacterium P. larvae .............................................................. 46 Minimum inhibitory concentration of active compounds against P. larvae ..... 51 2.4.2 Effects on the gut commensal bacteria when exposed to jelleine and melittin AMPs as potential treatments. .............................................................................................. 53 Changes in the composition: relative abundance .............................................. 53 Changes in the diversity .................................................................................... 55 Changes in size of the gut microbial community.............................................. 56 Changes in the total abundances of the core species ........................................ 59 2.5 Conclusion .................................................................................................................... 60 Chapter 3: Engineering of honey bee gut commensal bacterium Snodgrasella alvi for the secretion of naturally occurring antimicrobial peptides ..........................................................62 3.1 Introduction and rationale ............................................................................................. 62 3.2 Research goals .............................................................................................................. 66 3.3 Methods......................................................................................................................... 67 3.3.1 The Sec-dependent signal peptide library of Snodgrasella alvi ........................... 67 xi 3.3.2 Top secretion activity signal peptides experiment ................................................ 68 S. alvi protein sample preparation: secretion and whole cell fraction protein extraction68 Protein denaturation: alkylation and reduction ................................................. 69 Proteolytic digestion and peptide clean up ....................................................... 69 Liquid chromatography and tandem mass spectrometry .................................. 70 Data analysis ..................................................................................................... 70 3.3.3 Plasmid design ...................................................................................................... 72 3.4 Results & Discussion .................................................................................................... 72 3.4.1 The Sec-dependent pathway signal peptide library of Snodgrasella alvi ............. 72 3.4.2 Top secretion ratio activity signal peptides: candidates signal peptides to direct heterologous secretion .......................................................................................................... 72 3.4.3 Plasmid design ...................................................................................................... 79 Casettes of expression ....................................................................................... 79 3.5 Conclusion .................................................................................................................... 80 Chapter 4: Conclusion .................................................................................................................82 4.1 Addressing the project aims .......................................................................................... 83 4.2 Addressing the project hypothesis ................................................................................ 85 4.3 Future directions ........................................................................................................... 86 4.3.1 Efforts for the therapeutic probiotic completion ................................................... 86 4.3.2 Uncovering the mechanisms of action for the chosen antimicrobial peptides through in vitro assays. ......................................................................................................... 87 xii 4.3.3 AMPs efficacy against non-commensal bacteria: is this the reason why honey bee gut environment is very selective? ........................................................................................ 87 4.3.4 Differential protein expression profiles between AFB immune Asian honey bee and susceptible Western honey bee larvae............................................................................ 87 4.3.5 Microbiota tolerance to host AMPs validation in other insect models ................. 88 4.3.6 Chromosomal integration of our expression cassettes .......................................... 88 4.4 Closing .......................................................................................................................... 88 References .....................................................................................................................................89 xiii List of Tables Table 2.1 List of antimicrobial peptides used for the susceptibility tests ..................................... 33 Table 2.2 MIC assay dilutions ...................................................................................................... 35 Table 2.3 Zone of inhibition assay data. The average inhibition score (cm) in bold and standard error in italics. ............................................................................................................................... 42 Table 2.4 Proportions of taxonomical groups per sample. ........................................................... 54 xiv List of Figures Figure 1.1 The life stages of the worker honey bee. Egg: this stage is about three days long and finishes with egg hatching. Larva: this stage is from five to six days long and involves extreme increases in size. Larval stage ends when cells are sealed by nurses. Pupa: this stage lasts about eight to nine days and finishes when adult honey bee emerges. It involves the most important morphological changes. Adult: the life span of the adult worker honey bee depends on the season and the overall health status of the colony, but it can last from two to three weeks during summer season and several months during the winter (From left to right). ................................................. 4 Figure 1.2 Mechanism of action of Paenibacillus larvae: A) Spore germination and bacteria proliferation: Spore germination occurs in the midgut environment of the larva. Once the vegetative cells are metabolically active they are able to secrete secondary metabolites that can target any potential bacterial competitor. B) Breaching of the peritrophic matrix: Before reaching the midgut cells, P. larvae depletes the chitin present in the peritrophic matrix. C) Bacteria takes over midgut cells and invade larval cadaver: In direct contact with midgut cells P. larvae is able to disrupt them and invade the hemocoel of the larvae. ............................................................... 10 Figure 1.3 Sources of microbiota acquisition. Newly emerged bees acquire their gut microbial components through social interactions mainly, but also floral and hive environments contribute with to this event. Here are identified the four better understood sources of gut commensal bacteria A) Fecal-oral: Snodgrasella alvi, Gilliamela apicola and Frischella perrata. B) Oral trophallaxis: Lactic acid bacteria C) Hive environment grooming: Lactic acid bacteria (Lactobacillus Firm-5) D) Foraging (Flower environment): Lactic acid bacteria. ....................... 18 xv Figure 1.4 Hypothesis 1: Larvae gut has a very low load of bacterial cells, while the protein abundance for immune related proteins and peptides is also reduced (left). The adult honey bee comprises of a large and constant gut bacterial community and a robust constitutively present collection of immune related protein and peptides (right). ........................................................... 25 Figure 1.5 Hypothesis 2. A simplification of the midgut of the adult honey bee gut in the presence of P. larvae (left) vs the midgut of the honey bee larvae midgut in the presence of P. larvae (right). Adult honey bee antimicrobial peptides and commensal bacteria coexist in a delimited environment. ................................................................................................................. 26 Figure 1.6 Hypothesis 3. Sec-dependent secreted proteins carry a signal peptide that targets its secretion. The integration of homologous signal peptide (short green) to a heterologous protein (red) could help to lead its secretion (From top to bottom). ......................................................... 27 Figure 2.1 Overall rationale to identify candidate AMPs to develop prophylactic probiotics. A) Host-microbe challenge experiment. B) Naturally occurring AMPs susceptibility tests for P. larvae, commensal and non-commensal bacteria C) Microbiome dysbiosis experiment. ........... 31 Figure 2.2 Isolation of the commensal gut bacteria by enrichment cultures: the first step involves the isolation of the digestive tract; the second step is to prepare the gut homogenate in PBS; third, aliquots of the previous homogenate are used to inoculate different selective media (agar plates); After the first colonies are observed (usually 3 days), single colonies are re-streaked in fresh agar plates until reaching axenic colonies (From left to right). ........................................... 36 Figure 2.3 Axenic cultures of B. apis, B. asteroides and S. alvi (From left to right). .................. 37 Figure 2.4 Analytical Flowchart (Microbiome Insights). From raw data to qualitative and relative abundance analyses. ...................................................................................................................... 40 xvi Figure 2.5 Zone of inhibition assays raw data representative samples for P. larvae: Ampicillin, oxytetracycline, jelleine and melittin (From left to right). ............................................................ 42 Figure 2.6 Each data point represents the inhibition zone diameter caused by each AMP to P. larvae. * Pairwise significant comparison against blank. ............................................................. 43 Figure 2.7 A) Each data point represents the inhibition zone diameter caused by each AMP to E. coli. B) Each data point represents the inhibition zone diameter caused by each AMP to B. subtillis. * Pairwise significant comparison against blank. .......................................................... 44 Figure 2.8 A) Each data point represents the inhibition zone diameter caused by each AMP to B. asteroides B) Each data point represents the inhibition zone diameter caused by each AMP to S. alvi C) Each data point represents the inhibition zone diameter caused by each AMP to B. apis. * Pairwise significant comparison against blank. ............................................................................ 45 Figure 2.9 Susceptibility scores commensal vs non-commensal bacteria: Y-axis shows the average inhibition score caused by each peptide to the non-commensal bacteria cohort, while the X-axis shows the average inhibition score caused by each peptide to the commensal. Overall all the AMPs have greater inhibitory effect against non-commensal, observing a cluster of the scores above the diagonal. ....................................................................................................................... 46 Figure 2.10 The inhibition effect caused by bee derived AMPs vs fly derived AMPs. To observe if there is any pattern favoring the coexistence between honey bee CGB and humoral factors of adult honey we modeled inhibition scores with categorical variables describing whether bacteria were commensal or non-commensal, whether Gram-positive and Gram-positive and whether AMPs were bee and fly. In this figure the inhibition is clearly larger for the non-commensal for both type of AMPs and overall the honey bee derived peptides seem having stronger inhibitory activity........................................................................................................................................... 47 xvii Figure 2.11 The inhibition effect caused by bee derived AMPs vs fly derived AMPs to each species. The data points for the inhibition caused to non-commensal (Left) and the inhibition cause to commensal bacteria (right). ............................................................................................ 49 Figure 2.12 Hemolymph vs non-hemolymph AMPs. The data points caused to non-commensal (left). The data points caused to commensal bacteria (right). The inhibition effect caused by hemolymph (first boxplot) derived AMPs vs non-hemolymph (second boxplot) derived AMPs of the honey bee. ............................................................................................................................... 50 Figure 2.13 Close up to the inhibition caused by each AMP. Each panel contains the zone of inhibition data points corresponding to each AMPs to both cohorts of bacteria. ......................... 51 Figure 2.14 Close up of the inhibition caused by antibiotics. ....................................................... 51 Figure 2.15 Representative data of 96-well plate microdilution assays. Growth curves for E. coli challenged by oxytetracycline, jelleine and melittin (up). Growth curves for P. larvae challenged by oxytetracycline, jelleine and melittin (down). ......................................................................... 53 Figure 2.16 Taxonomic composition plot. Composition for each sample replicate (right) composition of the average for treatment (Left). Each color represents the corresponding taxonomic group categorized. This figure just represents the relative abundance of each group identified per sample/treatment..................................................................................................... 54 Figure 2.17 Shannon index (up) and Beta diversity (down). No significant differences in the Shannon diversity index. Beta-diversity index threw significant differences for among group factors which can be observed by a slight cluster of the datapoints. ............................................ 56 Figure 2.18 Total number of 16S rRNA gene copies. Absolute, including baseline data (left). Normalized and relative to the blank data (right). For right figure only, ANOVA pairwise significant comparison against blank: * (P < 0.05) ** (P<0.01) and *** (P<0.001). .................. 58 xviii Figure 2.19 Total number of 16S rRNA gene copies per species. Bifidobacterium species, Lactobacillus species and S. alvi (panels from left to right). ANOVA pairwise significant comparison against blank: * (P < 0.05) ** (P<0.01) and *** (P<0.001). ................................... 59 Figure 3.1 Sec-dependent signal peptides domains properties and architecture. The n-region is following the N-terminus and is positively charged; the H-region stands for helical hydrophobic region; the c-region is the terminal region of the signal peptide and is slightly polar; the cs (cleavage site) usually presents a characteristic AXA motif; the mature region domain corresponds to the exported protein. ............................................................................................. 64 Figure 3.2 Signal peptide library through SignalP 5.0 This involves retrieving type I signal peptide containing proteins list. .................................................................................................... 68 Figure 3.3 Secretion ratio experiment overview: 1. Bacterial colonies are harvested and suspended in PBS for further fractionation 2. The whole cell pellet is separated from the supernatant, washed and lysate 3. Protein sample preparation included the protein denaturation, solubilization, proteolytic digestion and peptide clean-up 4. LC MS/MS 5. Data analysis. ........ 71 Figure 3.4 Example of the protein lists obtained for each fraction A) Secreted fraction protein list. The proteins present in the secreted fraction ranked for their total intensity B) Whole fraction cell list. The proteins present in the whole cell fraction ranked for their total intensity. ............. 75 Figure 3.5 SignalP 5.0 prediction. Y-axis shows the probability of the sequence to be a signal peptide and x-axis shows the amino acids sequence that are part of that putative sequence. A) 2-oxoglutarate dehydrogenase E1 component. B) Putative membrane-associated, metal-dependent hydrolase. Each amino acid sequence shows the likelihood to belong to the corresponding motif of the signal peptide. ..................................................................................................................... 76 xix Figure 3.6 SignalP 5.0 prediction for Carbonic anhydrase (Top scored signal peptide type I containing protein). Each amino acid sequence shows the likelihood to belong to the corresponding motif of the signal peptide. ................................................................................... 80 xx List of Abbreviations CGB Commensal Gut Bacteria AMPs Antimicrobial peptides OTC Oxytetracycline AFB American Foulbrood MIC Minimal Inhibitory Concentration qPCR Real-time polymerase chain reaction AMR Antimicrobial resistance dsRNA double-stranded RNA xxi Glossary Commensal Bacteria The bacteria that benefits from the host without bringing any subsequent harm. Microbiome The collection of genomes of a microbiota. Microbiota The set of microorganisms present in a determined place. Probiotic Any live microbial culture product which beneficially influences the health and nutrition of the host when applied. xxii Acknowledgements I want to sincerely thank my supervisor Dr. Leonard Foster for giving me the greatest chance I have had lately: to become part of his unique research group and UBC. Thank you for considering me capable of being part of this group. Special thanks for letting me always explode my creativity and trusting me with every single decision I made. Thank you for being always open to listen any voice and for promoting inclusivity in STEM. Also, special thanks to my committee members, Dr. Hallam and Dr. Duong for supervising this work. I offer my gratitude to my funding sources for making my graduate pathway possible: the National Council of Science and Technology (Mexico), MITACS and the American Beekeeping Federation. I want to thank to the lab members Nik, Jenny, Greg, Dani, Anastasia, Nat, Teesha, Ali and many more that patiently helped me through my work and gave me outstanding suggestions along my journey. Particularly, I want to thank Jessica Seropian for being an enthusiastic and curious undergraduate that helped me not only with a great amount of research, but by sharing her passion for science. I also owe my gratitude to Dr. Waldan Kwong for closely supervising my ideas and mentoring me throughout this project. Thank you to my parents for raising me as a critical thinker. Thank you to my sister, Juliana, for being a role model. Thank you to my grandma, Mar\u00C3\u00ADa Esther, for taking care of me during my childhood and to the most cultured Grandpa I could ask for, for encouraging my will to learn, xxiii even when I could barely talk. Thank you to my family in Mexico who cheer me up, and thank you to my dearest friend Regina for all the love and support these last years. Lastly, thank to my most loyal furry friend, Gattusi for increasing my love to nature, living things and biology. xxiv Dedication To Jorge Antonio Mercado Alonso and Javier Francisco Arredondo Verdugo: Jorge and Javier were two outstanding graduate students of the Faculty of Engineering and Sciences of Monterrey Institute of Technology. Back in 2010, Monterrey was a city subdued by violence and organized crime, armed conflicts took over every corner. The night of March 19th, after a long day in the library, Jorge and Javier left the university close to midnight. After having some food, the students came back a few minutes later to continue studying. Tragically, they were in the wrong place at the wrong time. Army forces shot them, arguing that Jorge and Javier were mistaken for criminals. Even the media upheld that story so as not to damage the prestige and reputation of the university. It was not until years later when the truth for their innocence was revealed and their innocence and conviction for research and academy was honored. However, lately their names have been forgotten, and justice hasn\u00E2\u0080\u0099t taken place. Jorge and Javier could not finished their theses, so I am happy to dedicate my thesis to their memory. For all those that couldn\u00E2\u0080\u0099t culminate their contribution to engineering and science because of violence, for all those whose only worry is not just finishing experiments or finding funding, but to not get killed in the way home. For Latin American graduate students that risk their lives every day to research and those that would love to, but lack of opportunities to do so, I want to dedicate all my work to you. 1 Chapter 1: Introduction 1.1 The basics of the global pollinator: The honey bee 1.1.1 Relevance Apis mellifera, better known as the western honey bee, is the single most critical insect pollinator for the current food production system. Economically, honey bees contribute close to 20 billion USD and 4.6 billion CAD to the crop industry in the USA and Canada, respectively1,2. Ninety major commercial crops are tangibly dependent on honey bees for pollination, with 90% of blueberries and cherries in the USA pollinated by these bees1,3. Products such as honey, royal jelly, propolis and wax add up to the honey bee\u00E2\u0080\u0099s economic importance, supplementing the billion-dollar pollination industry. For the scientific community, the value of the honey bee is incalculable. It is considered a model organism for eusociality, social behavior and behavioural genetics4,5. The sophistication reached through millions of years of parallel evolution with flowering plants has been an object of study for evolutionists, ecologists and entomologists6,7. Recent advances in molecular technologies has markedly increased the study of honey bees\u00E2\u0080\u0099 genetics, producing a provoking landscape. The uniqueness of honey bee genome stands out when contrasted to Drosophila (fruit fly) and Anopheles (mosquito), being that some traits are markedly different from more canonical models of insect genetics which is particularly challenging for comparative genomics approaches8. Additionally, the study of the honey bee gut microbiota has revolutionized the field turning out the interest of researchers in the understanding of the microbiome contribution to honey bee\u00E2\u0080\u0099s health, social behaviours and evolution9,10. Moreover, due to the relative simplicity of the honey bee gut microbial composition, it has become a model community for understanding microbiome 2 interactions and co-metabolic processes extensible to more complex gut communities such as the human microbiota 11,12. 1.1.2 General honey bee biology The castes When mentioning the honey bee, it is commonly referring to the most abundant member of the colony, the worker honey bee, one of the three castes of this species. The colony is the operational unit of this social insect and comprises of commonly 50,000-60,000 individuals, although sometimes colonies can reach 100,000 or more individuals13, in which just one of them is a fertile female (the queen); few thousands are fertile males (drones); and the rest of the colony is made up of sterile females, previously mentioned (workers). Honey bee\u00E2\u0080\u0099s advanced social organization and diet triggered caste differentiation have been partially understood by their genome assembly and by years of studying their social interactions and behaviours14. However, some of their most fine mechanisms are still to be understood, such as diet-induced methylation changes and altered gene expression inducing caste differentiation control15. Each caste has a specific goal for the stability and survivorship of the colony. Drones mate with the queen, enabling her to fertilize eggs. The queen stores the semen separate from the eggs for the rest of her life, and when she lays an egg a sperm can be released from the spermatheca to fertilize it (if demanded)13. The unfertilized eggs are most of the time drones and the diploid are females, either workers or the next queen. In addition to laying eggs, the queen is responsible for the synthesis of pheromones that guide the operation of the hive 16. Worker bees can be subclassified into two categories depending on their functional task, the nurses and the foragers. 3 The nurses are the younger adults which are mostly in the colony grooming after the potential threats (i.e. pathogens) and feeding the larvae and queen. Older adults, known as foragers, are responsible on collecting pollen, nectar and propolis. This role involves leaving the hive and flying long distances13. Development The lifespan of the honey bees varies heavily depending on the caste and seasonal variability17. For example, the adult workers bees, which are active in the summer season, normally survive just for around three weeks, while they can survive several months in the inactive winter season. The queen lifespan comprises of up to three years, but the drone dies after mating, when it is no more than few weeks old13. Even if lifespan changes drastically between castes, all of them comprise four major developmental stages: egg, larvae, pupae and adult (Figure 1.1). The starting point of the life cycle is an egg, relaying on the queen which is the only member able to lay eggs, usually one per cell (referring to one of the holes (cells) in honey comb); the second stage is the larvae, which doesn\u00E2\u0080\u0099t involve important morphological changes but a considerable change in the size by feeding from the adults. The larvae lacks most of the distinct organs, such as eyes, legs and external carcasses. However, it is a very critical stage for the caste differentiation of queen and workers by nutrition18. Being an almost completed differentiated individual, the pupae develops the cuticle and starts acquiring its characteristics dark color13. A couple of days after emerged, the adult completes its distinct characteristics. The acquisition of a stable gut commensal community usually takes five days for the worker bees (post hatching)19. 4 Figure 1.1 The life stages of the worker honey bee. Egg: this stage is about three days long and finishes with egg hatching. Larva: this stage is from five to six days long and involves extreme increases in size. Larval stage ends when cells are sealed by nurses. Pupa: this stage lasts about eight to nine days and finishes when adult honey bee emerges. It involves the most important morphological changes. Adult: the life span of the adult worker honey bee depends on the season and the overall health status of the colony, but it can last from two to three weeks during summer season and several months during the winter (From left to right). Diet Even if each caste relies on different nutritional needs and feeding mechanisms to satisfy their requirements the staples of any honey bee diet include nectar/honey and pollen13. Nectar, which is from 20 to 60% w/w sugars (mainly sucrose, fructose and glucose) provides the carbohydrates, while pollen provides for the source of proteins, lipids and vitamins20. In fact, pollen is the sole protein source that honey bees have, and can vary from 10-60% of protein content (dry mass)21. The differences in diet are clear starting with the early larval stage. The quantity and quality of the food fed to female larvae determines whether those larvae will develop into workers or 5 queens. Worker larvae are fed primarily brood food produced by the hypopharyngeal and mandibular glands of nurse bees, although some pollen is fed directly to larvae on the fourth and fifth day of larval development13. The food of the queen larvae is called royal jelly and it has a higher composition of mandibular gland secretions and sugars than the worker larval food. Even feeding habits are dramatically different between castes, adults queens are mostly fed by adult workers and rarely by themselves13. Genetics Genetically, the honey bee has some remarkable features. Depending of the nutritional and hormonal stimulation previously mentioned, females (diploid eggs) can develop into a queen or a worker, which is a classic example of the environment impacting the phenotype. After the assembly of Apis mellifera genome by the The Honeybee Genome Sequencing Consortium (2006), more of the honey bee biology was revealed. With a total of 236 million bases pairs, this genome opened a new genome architecture landscape for insects. The unique distribution of nucleic acid motifs, such as the spatial heterogeneity of A+T content, high CpG content and the absence of most major families of transposons were considered some highlights. Interestingly, there are some evidences of the greater similarity to vertebrate genomes for DNA methylation patterns than to insects15. Finally, the lower evolutionary rate for all the insects currently sequenced has been attributed to honey bees8. 1.1.3 The endangered health of the honey bee: pathogens and other threats A misconception triggered by the worldwide increase in the total number of honey bee hives, mainly by the recent joining of the China, Australia and Argentina to beekeeping industry22, 6 hides the trend that honey bee numbers drops precipitously in North America and Europe each winter. In general pollinators, whatever they are wild or managed, face a challenging scenario derived from our current food production system and the changes to the ecosystem by human exploitation of the environment. The species richness of wild pollinators has become limited, considerably triggered by habitat loss23. Climate change also has played a substantial role affecting directly the fertility of honey bees and related species, which thermoregulation is limited24. Western honey bees are managed pollinators, meaning that hives are supervised with the specific goal to commercialize the pollination services. Even if the human economy and food supply depend highly on the pollination services of honey bees, the last few decades have been particularly challenging due to the increase of the number of threats: a wider diversity of pathogens, agrochemical exposure in crops; poor nutrition from monocultures pollination, just to mention a few 22,23. Honey bees were introduced to North America from Europe when explorers settled the continent25, bringing with them their natural pathogens, which play an important role in the population dynamics. Nonetheless, after the intensification of the agricultural practices in North America, the landscape dramatically changed. The high demand of pollination services, principally for almond and berry production, made it necessary to export honey bees from Asia and Australia, leading to the spread of non-natural pathogens23. The introduction of pests, poor nutrition and presence of pesticides residues act synergistically making honey bees more susceptible to disease23. 7 Among all the pathogenic agents, viruses are the most diverse. The most common virus of the honey bees is deformed wing virus (DWV) followed by the Dicistroviruses, Israeli acute paralysis virus, Kashmir bee virus, etc. 26. Parasites commonly contribute to the spread of viruses27. Currently the most important economic losses of bees across the world have been driven by the Varroa destructor mite, an invasive parasite that transferred from its native host, the Asian honey bees (Apis cerana), when both species were in close proximity. Today this mite is a major vector for the spreading of DWV28. Meanwhile, bacterial diseases in honey bees have lost the current attention, since can be relatively controlled by the use of antibiotics as prophylactics, which is very alarming, considering the recent increase on antibiotic resistant bacteria worldwide. Historically, American foulbrood (AFB), a bacterial caused disease, has been considered the most catastrophic disease for the beekeeping community29. Although that changed since the introduction of Varroa mite, focusing efforts on controlling this last one. Efforts to improve overall bee health to prevent this parasite are very broad an include strategies for developing integrated pest management guidelines, understanding the mechanisms of actions of the disease, developing active compounds, designing molecular tools to fight against pathogens and selective breeding strategies30\u00E2\u0080\u009332. 1.2 American foulbrood: Implications and mechanism of action of the most virulent bacterial disease of honey bees 1.2.1 Mechanisms of infection and transmission American foulbrood is caused by the infection of the Gram-positive, spore forming bacteria Paenibacillus larvae. As suggested by its name it only affects the honey bee brood, particularly at the larval stage. However, as being a lethal and highly virulent disease with no treatment, once 8 the colony is infected it should be burned, otherwise the adults will contaminate the rest of the progeny, risking the future of the colony and leading to the contamination of the nearby colonies and apiaries that would be followed by the collapse of all the proximate populations of honey bees. According to Ebeling, J., et al. (2016), the pathobiology of AFB can be simply described as two stages of activities that eliminate biological and physical barriers present in the larvae midgut until the bacteria hijacks the midgut cells and invades the hemocoel33. Figure 1.2 depicts the mechanism of action of P. larvae. Initially, infection is spread by adult workers feeding larvae with spores contaminated food. Once the spores germinate in the larvae midgut lumen, the disruption starts. First, P. larvae vegetative bacteria releases secondary metabolites, including antibiotics such as polyketides and non-ribosomal peptides that potentially kill any bacterial competitors present in the larvae gut environment34,35. In fact, even gut bacterial composition of the adult honey bee (vector of disease but immune) has been reported altered by the disruptive pathogen36. Secondly, P. larvae releases chitin degrading enzymes that eliminate the peritrophic matrix, removing any physical barrier protecting the midgut cells37. Finally, once P. larvae is in direct contact with the midgut cells it releases toxins that target DNA, RNA and cytoskeleton38. Thereafter, the bacteria permeates the hemocoel (analogous to the bloodstream). By then, the larvae is dead. However P. larvae will continue its activity until degrades the larvae cadaver that will eventually become a ropey mass. When the nutrient availability is low, the bacteria goes back to its dormant state38. One additional element that increases the lethality of the pathogen is the presence of S-layer proteins that likely contribute to the attachment of P. larvae to the midgut surface39. Overall, the above described mechanism is highly conserved in the two most recurrent 9 genotypes of P. larvae, ERIC I and ERIC II (classified according enterobacterial repetitive intergenic consensus sequences). However, the ERIC II is considered more destructive resulting in a faster disease progression29. At this point there is no actual treatment for AFB: it can only be prevented or eradicated by also destroying the hive. For its prevention, the use of antibiotics is the only formal strategy and is not permitted world-wide40. In North America, the use of mostly two antibiotics, oxytetracycline (OTC) and tylosin, has been systemic for the last seven decades, leading to antimicrobial resistance (AMR)41,42. In Canada the only antibiotic approved for the prevention of AFB is OTC40. On the other side, there are few options for the eradication. First, and very dramatic, it is to burn the complete hive after evidence of infections. Also, the burning of any tool used during the beekeeping activities, followed by burying the burned remains. A second option is the use of radiation for the non-living hive materials, but this is considerably exclusive and not something that most beekeepers can access easily. A final strategy, only available if the infection is moderate without clinical symptoms, is the \u00E2\u0080\u009Cshook swarm methods\u00E2\u0080\u009D. This technique was used before the advent of antibiotics and consists of the transferred of the adults to a new hive with new foundations, to start a new cycle of reproduction with healthy progeny43. 10 Figure 1.2 Mechanism of action of Paenibacillus larvae: A) Spore germination and bacteria proliferation: Spore germination occurs in the midgut environment of the larva. Once the vegetative cells are metabolically active they are able to secrete secondary metabolites that can target any potential bacterial competitor. B) Breaching of the peritrophic matrix: Before reaching the midgut cells, P. larvae depletes the chitin present in the peritrophic matrix. C) Bacteria takes over midgut cells and invade larval cadaver: In direct contact with midgut cells P. larvae is able to disrupt them and invade the hemocoel of the larvae. 11 1.2.2 Prophylactics for P. larvae infection Oxytetracycline resistance crisis Antimicrobial resistance threatens global stability, human health and the food production system44. Considering that pharmaceutical companies get a low revenue through the development and commercialization of antibiotics by investing great amounts of money on a product that will face with microbial adaptation in a relatively fast timeframe, it is not surprising that few new options have been designed for AFB44. As just mentioned, Canadian beekeepers rely on one single antibiotic to prevent AFB (OTC), while the USA also relies on tylosin, but it has more dosage restrictions40. Alarming, the evaluation of more than a hundred different bacterial isolates suggested that there is no correlation between haplotypes and the degree of OTC resistance, providing insights into the mechanisms by which the bacterial strains have developed resistance (non-genomic manners) and evidencing that resistance to OTC has evolved multiple times in P. larvae and will continue this way 41. The potential horizontal transfer of plasmid DNA also involves the commensal gut bacteria CGB of honey bees. Plasmids conferring antibiotic resistance have been found among the core commensal bacteria. Tian B., et al. (2012) surveyed for the presence of eight tetracycline resistance loci and two ribosome protection genes in commensal bacteria genomes. Colonies that had history of OTC treatment showed highly abundance of these genes compared to the colonies that were OTC free for more than two decades42. Alternative strategies to fight P. larvae infection There has been increasing interest in finding active compounds or strategies to be used as alternatives for dealing with AFB. Most of the current efforts are focused on finding compounds 12 that inhibit the growth of the vegetative form of P. larvae. Following this, studies on oral toxicity to honey bees must also be conducted, and in some outstanding studies, the validation of the compound\u00E2\u0080\u0099s activity during an infection. Here are described some of the most remarkable attempts on the exploration of natural compounds, such as essential oils, propolis and antimicrobial peptides. Essential oils are mostly volatile secondary metabolites of plants. Kuzy\u00C5\u00A1inov\u00C3\u00A1, et al. (2013) evaluated the in vitro activity of worldwide known essential oils against P. larvae45. However, the inhibition assays tested a range of volumes, ignoring the concentration of the active compound, as a rough estimation of doses. Among the strongest inhibitory compounds, thyme, oregano and clove were reported. Even if the availability of active compounds against P. larvae is highly optimistic, more studies are needed to conclude that these compounds are compatible with honey bees for several reasons: first, essential oils usage has been explored to prevent and treat several diseases in vivo and in vitro, such is the case of certain types of cancers or infections, suggesting a non-specific cytotoxic mechanism. Commonly the doses that completely inhibit bacterial growth are found as cytotoxic46. Second, and linked to the former disclaimer, the commensal bacteria could also be affected by the essential oils active compounds. Third, essential oils can potentially alter bees\u00E2\u0080\u0099 behavior by binding odorant receptors. Another plant-derived product is propolis, which mostly consists of tree resins and is collected by honey bees to be used as a protective antimicrobial layer in the hive environment47. Among the more abundant active compounds in propolis are polyphenols, terpenoids, steroids, and aminoacids48. Inhibition assays, toxicity assays and in vivo studies have been performed to look 13 into the activity of propolis ethanolic extract49. A temporary control of the infection was achieved, but if the treatment is discontinued the disease reappears and it is not preventive. The application of propolis could be complementary but not as main strategy (since is not definitive), and also it should be determined if could potentially cause selective pressure of more aggressive P. larvae strains. Probiotics A probiotic is any live microbial culture product which beneficially influences the health and nutrition of the host when applied50. The mechanisms through probiotics prevent infections of pathogenic bacteria include the stimulation of humoral and cellular immunity, decrease of unfavorable metabolites, secretion of beneficial secondary metabolites and through competition for nutrients and other resources51. Since previously studies have demonstrated that the honey bee gut environment is rarely friendly to non-commensal bacteria11, the options for probiotic bacteria are limited. In general attempts to provide honey bees with beneficial bacteria include in its majority the application of lactic acid bacteria. Daisley, B. A., et al. (2019) reported an evaluation of the nutritional supplement \u00E2\u0080\u009CBioPatty\u00E2\u0080\u009D, a mixture of nutrients and three different lactobacillus species. According to the author\u00E2\u0080\u0099s, previous in vitro studies for the three used species against P. larvae shown inhibitory activity. Additionally, by feeding the bees with the supplement containing the bacteria cocktail before a natural AFB outbreak the treated bees survived better52. However, it is worth-noting that Lactobacillus sp. are mostly residents of the crop (foregut), while P. larvae germination and growth is favored in the midgut, so likely the effect is caused by indirect manners but the authors missed including a suggested MOA53 which limits the discussion. One limitation to further investigate is the dysbiosis caused (if there is any) 14 considering that the biological niches of the commensal bacteria could be replaced by these close relative species, and bring potential side effects. Moreover, the composition of the patty should be revised, considering the fact that the mixture itself increased P. larvae loads, suggesting that promotes the overgrowth of antagonistic bacteria. At the end this approach is encouraging and relatively safe by including the in vivo validation during a natural outbreak and the use of hive naturally present bacteria, respectively. Antimicrobial peptides Moving from live organisms to active compounds, Khilnani and Wing (2015) explored the naturally occurring antimicrobial peptides (AMPs) of the adult honey bee against P. larvae vegetative cells. Under their defined conditions, defensin I proved to be effective against the target pathogen54. Their survey included five of the honey bee AMPs (apidaecin, abeacin, hymenoapteacin, defensin I and defensin II), which represents the four families of AMPs present in the adult honey bee hemolymph. Attempts to replicate the observed by Khilnani, or even trying few other AMPs from others organisms are expected to be done and validate the reproducibility of the inhibition assays, considering the likely bias caused by set up conditions (chosen media, peptide solubilization and P. larvae strain). Also the survey of more tissue specific honey bee peptides such as melittin and jelleine is missing. 1.2.3 Honey bee larvae vulnerability P. larvae is considered an exquisitely specialized pathogen, not only because of having a single species target (the Western honey bee), but because of its time window of infection, which is extraordinarily limited to the early stage of the larvae29. Only first and second instars (from a 15 total of five instars) are considered highly vulnerable to AFB55. Nonetheless, because of its extreme contagiousness and inevitably lethality, it is a serious threat to the honey bee. Therefore, the biological factors that permit the more advanced larval stages and adult bees to resist this infection have been attempted to be uncovered56\u00E2\u0080\u009358. The closely related Asian honey bee is immune to AFB at any of its developmental stages. Interestingly, one of the first attempts to understand the biological reason of this immunity, back before the \u00E2\u0080\u009Comics\u00E2\u0080\u009D boom, was to look into the direct activity of the Asian honey gut bacteria isolates against P. larvae cultures. With more than four isolates being active against AFB causative agent, the importance of the gut commensals in the protection of infection was suggested 57. Aside from commensal bacteria, some of the first attempts to determine intrinsic differences between immune and non-immune AFB honey bees involved to identify genes/proteins that were differentially expressed during P. larvae infection and at basal state along the larval stages. It was suggested that abaecin and defensin I were expressed in similar levels along the five instars constitutively, but as a response to the infection, the peptides were slightly up-regulated in the older bees59. The constitutive expression of peptides at very early stages suggests that larvae is not devoid of an immunological response, although the targeting of only two AMPs is a quite limited approach and is far from the understanding of the complete dynamics of the host defense system. It can also be argued that the transcript levels that were measured in that study were not necessarily correlated to protein abundance or activity. 16 Later on, Chang QWT., et al. (2009) looked into the differentially expressed proteins for the AFB-resistant fifth instar larvae, constitutively and under infection. Among the up-regulated immune factors, hymenoptaecin, prophenoloxidase (proPO) and lysozyme were identified when larvae were challenged. Lysozyme is responsible for the degradation of Gram-positive bacteria cell membranes, suggesting an explanation for the resistance of the older stages of the larvae. At the second instar larvae, expression of prophenoloxidase was very low compared to the following stages, which also fits with the sensitivity to the pathogen at these various stages58. This was one of the most complete and integrative approaches to understand the larvae vulnerability to the disease, considering the complete survey of the proteome and not only a set of proteins, which also leaded to understand the metabolic implications involved in the disease response. Although some physical components and cellular mechanisms non-yet explored could add up to the understanding of larvae susceptibility to AFB, commensal bacteria and immune factors are relatively well characterized. 1.3 Honey bee gut commensal bacteria: Description of the community 1.3.1 Composition As part of the efforts made to investigate Colony Collapse Disorder (CCD) in 2006, scientists surveyed, for the first time, the commensal gut bacteria of honey bees using the recently feasible metagenomics approaches available. With the goal of finding any microbial indicator involved in the mysterious decline of colonies, Cox-foster et al. (2007) screened CCD- positive and CCD negative colonies widely distributed in the United States. Many microbes whose presences were not correlated with CCD status denoted the likely resilient and characteristic honey bee gut microbiota60. 17 The honey bee gut microbiota has been described as a relatively simple environment. Particularly regarding bacteria, only nine species clusters make up 95-99% of the total bacterial composition of a healthy gut. From those, five species are ubiquitous to all adult honey bees and closely related bee species (Bombus sp. and stingless bees): the two Gram negative Snodgrassella alvi and Gilliamella apicola and the three Gram- positive Lactobacillus Firm-4 and Lactobacillus Firm 5 and Bifidobacteirum asteroides. In lower numbers Frischella perrara, Bartonella apis, Parasaccharibacter apium and gluconobacter related species are found occupying highly specific niches of the honey bee61. Regarding eukaryotic contributors, Martha Gilliam, pioneer in the study of the A. mellifera gut microbiota, reported four fungal species isolated from honey bee guts: Penicillium frequentans, Penicillium cyclopium, Aspergillus flavus and Aspergillus niger (Gilliam, 1997). In 2007, the first metagenomic survey of the bee gut found additional eukaryotes, including Pandora delphacis, Mucor spp., Nosema ceranae, Nosema apis and Leishmania/Leptomonas sp.60 Some of these, especially the Nosema spp., are known honey bee parasites and were initially not identified likely to difficulties on the culture models available for microsporidians as obligate intracellular organisms. 1.3.2 Establishment of the microbiome Newly emerged honey bees and larvae lack the characteristic commensal gut microbiota of adult bees 62. This community of bacteria is acquired in the first 4 to 6 days of interactions occurring outside but mainly inside the hive. Faecal-oral, trophallaxis (oral-oral), hive grooming and foraging, are the normal routes of transmission19 (Figure 1.3). Martinson, et al. (2012) suggested that larvae and newly emerged bees contain none or very few bacterial cells (no more than a 18 thousand cells compared to the billion bacterial cells of the adults), most of them being erratic bacteria. Although more studies are required for the full characterization of the contribution of these microorganisms to resist P. larvae infection, it is preliminarily suggested that they first act as a physical barrier avoiding the direct contact to the gut cells. The contribution of these microorganisms to the production of antibacterial compounds is still to be characterized but correlation has been identified63. Figure 1.3 Sources of microbiota acquisition. Newly emerged bees acquire their gut microbial components through social interactions mainly, but also floral and hive environments contribute with to this event. Here are identified the four better understood sources of gut commensal bacteria A) Fecal-oral: Snodgrasella alvi, Gilliamela apicola and Frischella perrata. B) Oral trophallaxis: Lactic acid bacteria C) Hive environment grooming: Lactic acid bacteria (Lactobacillus Firm-5) D) Foraging (Flower environment): Lactic acid bacteria. 19 1.3.3 Influence on bee health Authors in the field define microbiota as the collection of microorganisms present in a determined environment, while the microbiome is the repertoire of the genes of these microorganisms12 . The distinction between the microbiota and the functional microbiome becomes more important when considering organisms that lack a consistent, re-establishing core microbiota, such as most mammals 64,65 and Drosophila sp.64. Although common taxonomic units are too inconsistent to be considered \u00E2\u0080\u0098core\u00E2\u0080\u0099 in Drosophila flies, evidence shows that the function provided to the host fly by its microbiota is more consistent 64,66. This suggests that a very different collection of bacteria (a different microbiota) can provide a very similar set of functions (a similar functional microbiome) 64,67. The arisal of a core functional microbiome in an individual is similar to the arisal of a core microbiota, simply allowing the first successful species to colonize every functional niche rather than showing one consistent species preference. Many of the evolutionary arguments for the development of core microbiota also apply to core functional microbiomes. However, a long evolutionary history of host\u00E2\u0080\u0093microbe codiversification, as in honey bees, probably leads to increasing specialization of microbes and hosts toward each other68,69. Some honey bees commensal species are not found in any other species, not even other bees61. These intimate symbiotic relationships lead to more rewarding partnerships than in less specialized microenvironments, but they also result in considerably more codependence10,70,71. Regarding individual contribution of each species, there has been an important advance after the culture methods description for most of the CGB. For example, tracking of secondary metabolites in gnotobiotic bees (germ-free or specifically inoculated animals) helped to detangle the individual contribution to metabolites fermentation of host derived nutrients made by 20 commensal organisms. Bifidobacterium asteroides, is tentatively involved in stimulating the production of host hormones72. As for humoral factors, higher expression of apidaecin and hymenoptaecin was found in the gut tissue when the CGB were present63, likely playing an as yet unexplored role in the constitutive production of these peptides. Finally, N. apis infections of bees were more severe when the absence of commensal bacteria was first induced by antibiotics treatments, suggesting either an antagonist interaction or simply implying the physical barrier contribution of the non-pathogenic bacterial community73. Pairing survivorship experiments with proteogenomics approaches could help to increase the understanding of this antagonistic effect. 1.4 Adult worker immune system 1.4.1 Social immunity Living in very dense populations implies the continuous risk of disease due to individuals interacting constantly and sharing common resources. Nonetheless, the eusocial nature of bees helped to provide a defense to meet this challenge, social immunity. Social insects perform social behaviors that can help them prevent or combat disease. Most of these behaviors are focused on avoiding the introduction of parasites as a first prophylactic strategy, but when parasites do enter, the efforts are focused on avoiding the establishment and proliferation. Finally, if the infection was inevitable for some individuals, attempts to control it are initiated through individual manners74. Social immunity in honey bees is complex and involves several activities: lining the hive with pieces of propolis (resin with antimicrobial properties)47; worker bees grooming after the cells to get rid of potential threats; feeding antimicrobial substances to the brood74; and there is even 21 evidence that sick foragers do not return to the hive to avoid infecting their nestmates74. All these colony efforts do not guarantee that few, some, or all the colony members will have to fight an infection once in a while relying on their individual molecular toolbox: cellular and humoral immunity. 1.4.2 Individual immunity Individually, honey bees rely on their innate immune system, which is an integrated collection of cellular and humoral responses. One category of these involves the cellular responses of the infected cells, such as the encapsulation and melanization responses. The second is the humoral activity, comprising the release of AMPs and other immune factors. Unlike mammals, insects lack of adaptative immune system. However, like fruit flies and mosquitoes, the honey bee innate defenses involve the four conserved pathways: Toll, Imd, Janus kinase (JAK)/STAT and JNK75. The entire understanding of honey bees\u00E2\u0080\u0099 immunological responses has been buried by their unique genome architecture, making difficult to attribute an identity to all the genetic repertoire. There is also a need for functional validation of mechanisms. While is true that honey bees have fewer immune related genes, Evans et al. (2006) argued that limited repertoire is the result of fewer paralogue genes but that all of the canonical systems are still active. According to their analysis, Toll pathway in honey bees regulates AMPs, phenoloxidase (melanizing agent) and three lysozymes; while the Imd pathway can provide both positive and negative feedback for the expression of the AMPs 75. 22 Antimicrobial peptides Antimicrobial peptides are mostly cationic chains of roughly 9 to 70 amino acids. The mechanism of action of each AMP depends on the amino acid composition, however for the AMPs with considerable hydrophobicity, permeating the cell outer membrane is the most observed mechanism. Initial interaction with the negatively charged cell membrane is feasible through the net positive charge of the peptide76. Fat bodies and the hemocytic cells are the factories of these compounds, and their synthesis can either be constitutive, stimulated by environmental stressors, or dependent on life stage and sex77. Four families of AMPs are present in the hemolymph of honey bees: abaecin, apidaecin, defensin and hymenoptaecin, with each one having a different spectrum of activity. Abaecin has a broad spectrum, as a 34 long proline rich peptide78. Apidaecin Ia, Ib and II are the three closely related peptides that make up the apidacecin family, which is characterized by being proline rich and highly active against G-negative bacteria, in a bacteriostatic manner79. Defensins are part of the cysteine-rich peptide family with two isoforms, defensin I and defensin II, the former one being synthesized in the salivary glands and the latter one in the fat bodies. Both peptides are made up for 51 amino acids with 6 cysteine residues forming three disulfide bonds77,80. Hymenoptaecin is the largest cationic polypeptide present in the honey bees, with a total of 93 amino acids and including a 2-pyrrolidone-5-carboxylic acid at the N-terminus81. Few other AMPs are more site specific such as the case of those found in the bee venom and the royal jelly. The bee venom is composed of up 60% dry weight of melittin, which is a 26 residue polypeptide82. The jelleine family includes four peptides jelleine I-IV, all constitutively synthesized by the adult workers and secreted in the royal jelly83. 23 1.5 Hypothesis and aims 1.5.1 Research goal Efforts to counter OTC resistance have been focused on identifying active compounds that act directly against P. Larvae vegetative cells, largely disregarding the existing but complex mechanisms of infection resistance present in the adult honey bee. Our main goal is to develop a new prophylactic strategy for AFB based on the molecular differences between the adult and larvae. Specifically, I aim to develop a commensal bacterium strain that expresses and secretes naturally occurring antimicrobial peptides targeting P. larvae. In order to develop this ambitious goal, we need some intermediate aims. First, we need to identify humoral factors, specifically AMPs, of the honey bee that are not threats to the commensal bacteria but are effective against P. larvae proliferation. Second, we need to develop a robust strategy for the engineering of the commensal bacteria that includes not only the heterologous expression of the adult honey bee\u00E2\u0080\u0099s antimicrobial peptides, but also their secretion. 1.5.2 Aims: 1. The isolation and culture of the characteristic honey bee CGB from adult honey bees. 2. Determine the degree of specificity of immune factors against P. larvae. 3. The identification of AMPs that inhibits P. larvae but minimally affects the commensal gut bacteria, as well as their active concentration. 4. Survey the response of the microbiota to the use of the candidate antimicrobial peptides. 5. Identify the molecular machinery present in S. alvi for delivery and secretion of the antimicrobial peptide chosen previously. 24 6. Validate that the therapeutic probiotic colonizes honey bee larvae and that it is biologically active. 1.5.3 Hypotheses Hypothesis 1 P. larvae has evolved to act on the life stages of bees that do not have any defenses against it. Likely the costly mechanisms involved in the synthesis of immune factors is not yet initiated in 1st and 2nd instar larvae, making larvae to rely on the adult bee for grooming and protection. Conversely, the adult honey bee is sophisticatedly equipped to eradicate this highly co-evolved pathogen. AMPs and CGB are two of the most prominent mechanisms of defense of the adult honey bee and low expression of the AMPs and lacking of the robust CGB make the early larval gut an optimal environment for P. larvae proliferation. We hypothesized that by providing an engineered probiotic composed of a commensal bacterium able to express and secrete adult honey bee AMPs, larvae will resist the infection (Figure 1.4). 25 Figure 1.4 Hypothesis 1: Larvae gut has a very low load of bacterial cells, while the protein abundance for immune related proteins and peptides is also reduced (left). The adult honey bee comprises of a large and constant gut bacterial community and a robust constitutively present collection of immune related protein and peptides (right). Hypothesis 2 We hypothesized that in the limited AMPs repertoire of honey bees some of them are highly specific against P. larvae or invasive pathogens. Also, likely to the historical coexistence of these intrinsic immune factors and the core microbiota, most of the bacterial species are particularly adapted, which suggests that most of them are resistant to the intrinsic AMPs of honey bee, individually and as a community (Figure 1.5). 26 Figure 1.5 Hypothesis 2. A simplification of the midgut of the adult honey bee gut in the presence of P. larvae (left) vs the midgut of the honey bee larvae midgut in the presence of P. larvae (right). Adult honey bee antimicrobial peptides and commensal bacteria coexist in a delimited environment. Hypothesis 3 Heterologous secretion of recombinant peptides in the commensal bacterium S. alvi can be conducted through the implementation of homologous signal peptides present in its secretome (collection of secreted proteins) to the heterologous target antimicrobial peptide gene. Since the translocation depends on the interaction between the signal peptide, protein target and secretion machinery, we hypothesize that those signal peptides that show higher secretion activity for their native targets have higher chances to work with the heterologous target to induce secretion (Figure 1.6). 27 Figure 1.6 Hypothesis 3. Sec-dependent secreted proteins carry a signal peptide that targets its secretion. The integration of homologous signal peptide (short green) to a heterologous protein (red) could help to lead its secretion (From top to bottom). 28 Chapter 2: Exploring the molecular and biological toolbox of the adult honey bee against Paenibacillus larvae infection 2.1 Introduction and Rationale The characterization of different protein expression profiles as a function of a cellular state through mass spectrometry has increased the understanding of the dynamics in the living systems 84. Pathogenic challenge is a particularly intriguing state of the cell\u00E2\u0080\u0099s biology. Changes in the expression of proteins triggered by infection or disease can be revealing when it comes to understanding the mechanism(s) of pathogenesis, as well as the host strategies available to fight disease. Early proteomics studies of the honey bee explored the protein expression changes along developmental stages85, providing information on the molecular differences that likely increase the propensity of adult bees to resist certain diseases better than the developing brood. Additional experiments focused on honey bee larvae challenged with P. larvae brought further insight into mechanisms used by developing honey bees to fight infection 58. Immune related protein changes are a common theme in similar studies involving infection 86. However, the relevance of the results depends on the experimental design; the source tissue for the protein extraction selected is key, as well as the method of infection. Injection could trigger other responses that would interfere in the results87,88. The analysis of tissues in which the immune factors are synthetized or transported, such is the case of fat bodies and hemolymph, increases the resolution of the profiling. In Chan et al. (2009) lysozyme, prophenoloxidase and the AMP hymenoptaecin were overexpressed in challenged honey bee larvae specifically looking at hemolymph58. 29 The host immune system responses to non-pathogenic bacteria could aid to a better understanding of both immunity and mechanisms of colonization. The stimulation of the immune system by oral inoculation of honey bees with non-pathogenic CGB triggers the up-regulation of apidaecin and hymenoptaecin63. One intriguing question we were trying to address is, how specific are the humoral responses (innate immunity) available for bacterial stimulation by challenging adult bees with bacteria from three representative biological niches (pathogen, non-colonizer, and commensal)? What are the gaps between commensalism and infection? The answer to these questions could lead us to identify humoral factors that can be used for the treatment and prevention of P. larvae, and also to determine the role of the commensal bacteria in the constitutive expression of humoral factors. We had planned and performed a set of experiments that included to challenge adult honey bees with three different bacterial isolates (E. coli, P. larvae and S. alvi) and to further analyze changes on the protein expression profiles involved in the specific host-microbe responses. However, because of the shutdown started in March 2020 these results were not ready for this work. The gold standard for further validation of the direct effectiveness and degree of activity of any antimicrobial agent against a pathogen, such as P. larvae, is susceptibility assays. For an initial and qualitative survey, the use of zone of inhibition assays reveal whether or not a compound impedes the growth of the pathogen or bacteria of interest. Usually the concentrations used for this test are high and further validation is required to identify the minimal inhibitory concentration (MIC). MIC tests the lowest concentration that will block the growth of a determined number of microbial cells through serial and geometrically distributed dilutions89. Previous efforts have been made to adapt these assays to evaluate the effect against P. larvae and 30 CGB considering the challenges involved when working with fastidious organisms48,54,63. For this work we performed zone of inhibition assays to surveyed effects of naturally AMPs against P. larvae to later determine MIC. In recent years there has been an important set of studies determining the influence and importance that the gut microbiota confers to the overall health of the honey bee9,11,12. The use of antibiotics and other therapeutics affects the composition and dynamics of the CGB, making honey bees more susceptible to disease and death. Recent studies have surveyed the irremediable damage caused to the CGB community when exposed to antibiotics and herbicides90,91. Considering the increasing attention that gut microbiome dysbiosis of the honey bee demands, the validation of any antibiotic or prophylactic must be evaluated. For this purpose, microbial community structure and dynamics associated with AMPs challenge were determine using small subunit ribosomal RNA (SSU or 16S SSU rRNA) gene amplicon sequencing. Total 16S rRNA copy numbers were determine using qPCR. 2.2 Research Goals My overarching research goal of this set of experiments was to identify AMP candidates that could then be engineered into a prophylactic probiotic. The criteria for candidate selection included that i) AMPs are active against P. larvae vegetative cells in a highly specific manner and ii) the overall CGB are not disrupted. The approach to achieving this goal is summarized as series of aims visually depicted in Figure 2.1. The first aim was to identify the extent of specificity of the responses driven by the specialized pathogen P. larvae through mass spectrometry proteomics, particularly interested in the up-regulation of humoral responses such as AMPs when compared against the responses to S. alvi and E. coli. The second aim was to 31 validate the discriminatory effect of AMPs against pathogens, commensal, and non-commensal bacteria through susceptibility tests. The MICs of AMPs active against P. larvae were also be determined. Finally, the third aim was to evaluate the effects of these candidates on the CGB community, surveying for dysbiosis. This last aim was aided with 16s rRNA amplicon based metagenomic studies and qPCR to determine potential shifts in microbial community structure and abundance. Figure 2.1 Overall rationale to identify candidate AMPs to develop prophylactic probiotics. A) Host-microbe challenge experiment. B) Naturally occurring AMPs susceptibility tests for P. larvae, commensal and non-commensal bacteria C) Microbiome dysbiosis experiment. 32 2.3 Methods 2.3.1 Susceptibility specificity assays: Zone of inhibition and MIC microdilution assays To identify honey bee AMPs that discriminatorily inhibit the growth of the P. larvae and to evaluate their effect on the commensal bacteria we did zone of inhibition assays on agar plates. We tested the activity of five honey bee derived AMPs and three fruit fly AMPs to increase the understanding of specificity. The bacteria here challenged included two cohorts categorized as non-commensals (Paenibacillus larvae, Escherichia coli and Bacillus subtilis) and commensals (Bartonella apis, Bifodobacterium asteroides and Bartonella apis). Once AMPs that discriminatorily target P. larvae are identified, MIC for P. larvae should be surveyed through microdilution assays. Antimicrobial peptides synthesis and solubilization For the purposes of this work, we only considered representative AMPs of the honey bee and the fruit fly. From the former organism, we chose abaecin, apidaecin, defensin I, jelleine and melittin; while from the latter we chose andropin, cecropin and defensin I, that just for practical reasons will be referred as defensin II after this section (to differentiate it from honey bee defensin I). The AMPs sequences were retrieved from APD3 (Antimicrobial peptide database) and synthetized by Peptide 2.0 (Virginia, USA) (Table 2.1). The peptides were solubilized in milli-Q water to a final concentration of 4 mg/ml and stored at -20o C until used. 33 Table 2.1 List of antimicrobial peptides used for the susceptibility tests Antimicrobial peptide (Organism of source) Sequence Abaecin (Apis mellifera) YVPLPNVPQPGRRPFPTFPGQGPFNPKIKWPQGY Apidaecin (Apis mellifera) GNNRPVYIPQPRPPHPRI Defensin I (Apis mellifera) VTCDLLSFKGQVNDSACAANCLSLGKAGGHCEKVGCICRKTSFKDLWDKRF Jelleine (Apis mellifera) PFKLSLHL Melittin (Apis mellifera) GIGAVLKVLTTGLPALISWIKRKRQQ Andropin (Drosophila melanogaster) VFIDILDKVENAIHNAAQVGIGFAKPFEKLINPK Cecropin (Drosophila melanogaster) GWLKKIGKKIERVGQHTRDATIQGLGIAQQAANVAATAR Defensin I (Drosophila melanogaster) ATCDLLSKWNWNHTACAGHCIAKGFKGGYCNDKAVCVCRN 34 Growth of bacterial strains and zone of inhibition assay for non-commensal bacteria Zone of inhibition assays were performed as described by Khilnani and Wing (2005)54. In short, cultures of E. coli, B. subtills and P. larvae were grown overnight in 3x concentration R2B media at 37\u00CB\u009AC and 225 rpm agitation. The bacterial growth was then subcultured (1:20) and grown for an additional 16 h. OD600 readings for each culture were adjusted as needed. A total of 500 \u00C2\u00B5l of P. larvae and 250\u00C2\u00B5l of E. coli and B. subtills cultures then were pelleted and washed with PBS. The supernatant was resuspended into 150\u00C2\u00B5l of fresh 3x R2B media. Agar plates (1.5 % agar composition) of 3x RB2 media were inoculated with 150 \u00C2\u00B5l of bacterial suspension and spread using a sterile cell spreader. A hole was punched in the agar with a sterile tip and the AMP or antibiotic was added to this hole. After 48 h. the inhibition zone around each punched hole was measured. Broth microdilution assay on 96-well plate Bacterial cultures for E. coli and P. larvae were prepared by selecting a few colonies with the same morphological appearance and inoculating 5 ml of 3x RB2 broth and incubated at 37o C in a shaker at 225 rpm until it reached visible turbidity that is equal to Mc Farland standard of 0.5 (2 h for E. coli, B. subtillis and 6 h for P. larvae). Antibiotic dilutions were prepared according to the concentration range (Table 2.2). The antibiotic dilution (5\u00C2\u00B5l per well) and the bacterial inoculum (95\u00C2\u00B5l per well) were added to each well. The plate was incubated in a Tecan200 plate reader in orbital mode at 336 rpm for 5 min every 15 min and (amplitude of 1.5 mm) at 37o C. 35 Table 2.2 MIC assay dilutions Stage Concentration (\u00C2\u00B5g/ml) 1 100 2 50 3 25 4 12.5 5 6.25 Zone of inhibition for commensal gut bacteria To work with commensal bacteria it was necessary to first isolate these from honey bee gut homogenates and then validate their taxonomic identity. Also different media was tested before defining which was more adequate to work with for the zone of inhibition assays. 2.3.1.4.1 Isolation of the honey bee commensal gut bacteria by enrichment culture Adult honey bees were chilled for 5 min. at -20o C. Once bees were anesthetized were immersed in 70% ethanol to clean surfaces. Guts were pulled out with sterile forceps. After being resuspended in sterile PBS, guts were homogenized using a sterile pestle until reached a uniform suspension as described by Kwong & Moran (2013)92. Heart infusion (Difco), trypticase soy with 5% sheep\u00E2\u0080\u0099s blood (VWR), MRS (Sigma) and tomato juicer (Sigma) agar plates were streaked with the homogenate. After 48-72 h. visible colonies were picked and re-streaked onto 36 fresh agar plates. This was repeated until visually axenic cultures were obtained (Figure 2.3). Colonies from axenic final cultures were used for taxonomy identification. The overall description of this process in presented in figure 2.2. Figure 2.2 Isolation of the commensal gut bacteria by enrichment cultures: the first step involves the isolation of the digestive tract; the second step is to prepare the gut homogenate in PBS; third, aliquots of the previous homogenate are used to inoculate different selective media (agar plates); After the first colonies are observed (usually 3 days), single colonies are re-streaked in fresh agar plates until reaching axenic colonies (From left to right). 2.3.1.4.2 Taxonomy identification The amplification of the 16s rRNA gene (1465 bp) was conducted with the 16S universal primers 27F (5\u00E2\u0080\u0099-AGAGTTTGATCCTGGCTCAG-3\u00E2\u0080\u0099) and 1492R (5\u00E2\u0080\u0099-GGTTACCTTGTTACGACTT-3\u00E2\u0080\u0099) with 35 cycles of amplification (95\u00CB\u009AC for 20 s, 52\u00CB\u009AC for 30 s and 72\u00CB\u009AC for 40 s) after an initial incubation for 10 min at 95\u00CB\u009AC. PCR products were cleaned using AMPure XP PCR purification beads (Beckman coulter). Sanger sequencing of the amplicons was carried using 3130xl Genetic analyzer (Applied BioSystems) by the Sequencing/Bioanalysis Core Facility within UBC CMMT (Vancouver, Canada). Three different 37 hits were obtained for the 16S rRNA BLAST using NCBI 16S ribosomal RNA (Bacteria and Archaea) database: Snodgrassella alvi WKB2, Bartonella apis and Bifidobacterium asteroides. 2.3.1.4.3 Growth of bacterial strains and zone of inhibition assay The zone of inhibition assay described in section 2.3.1.4 was performed on S.alvi, B. asteroides and B. apis. Cultures were grown overnight in adequate media (TSB pH 6.0-6.5 (S.alvi), MRS pH 6.2 (B. asteroides) and S10 pH 7.0 (B. apis)). The solid agar used was 1.5% agar Mueller Hinton (Difco) with 5% sheep\u00E2\u0080\u0099s blood (Cedarlane) for S. alvi and B. apis and MRS with 5% blood for B. asteroides. 2.3.2 In vivo effects on the gut microbiota when exposed to jelleine and melittin Through this set of experiments we want to evaluate the effect on the size and composition of the gut microbiota when honey bees are challenge with AMPs and oxytetracycline at the real use dose. Here we include the establishment of the commensal gut bacteria to newly emerged bees, Figure 2.3 Axenic cultures of B. apis, B. asteroides and S. alvi (From left to right). 38 followed by the antimicrobials challenge, DNA extraction and the molecular analysis (qPCR and illumina sequencing). Establishing the commensal gut bacteria of newly emerged bees A brood frame was isolated from the hive and set at 37\u00CB\u009AC and 90% humidity. The day after bees emerged they were designated to a specific treatment cage (10 bees were chosen for each cage). For five days bees were fed with hindgut solution and bee bread, which was freshly prepared every day. The hindgut solution consisted of 10 hindguts homogenized in equal parts of PBS and 0.5M sucrose solution. The bee bread consisted of 50% v/v of autoclaved pollen and water to provide a solid food source. Two nurse bees were placed in each cage to ensure the newly emerged bees had a source of lactic acid-producing bacteria. Antimicrobials feeding Upon establishing the CGB after 5 d, bees were fed with a syrup solution containing the previously determined MIC for jelleine and melittin and the beekeeping prophylactic dose of OTC40 (jelleine 100\u00C2\u00B5g/ml; melittin 6.5 \u00C2\u00B5g/ml; OTC 6.8 \u00C2\u00B5g/ml). After three days bees were collected in 90% ethanol and stored at 4o C to proceed with DNA extraction. DNA extraction Three bees were randomly sampled from each replicate after the end of the experiment. Hindguts were extracted and pooled with 500 \u00C2\u00B5l of disruption beads (Fisher), 500 \u00C2\u00B5l of RLT buffer (Qiagen), and 5\u00C2\u00B5l of 2-mercaptoethanol. Each tube was placed in an ice bucket to start the sonication. After 5 min. of sonication (amplitude 60%, 5 min. Sonication probe (Brason, 39 Connecticut, USA). Samples were spun to separate buffer and beads. 100 \u00C2\u00B5l of the supernatant was transferred to 1.5 mL Eppendorf tube containing 100 \u00C2\u00B5l Ethanol, then gently mixed with a vortex. For the following DNA recovery steps, QIAamp DNA mini kit (#51304 Qiagen) was used starting from step 6. The DNA concentration was measured using Qubit (Thermo Fisher Scientific) and then normalized to 10 ug/\u00C2\u00B5l. qPCR Real-time qPCR amplification was performed using the ABI 7500 fast real-time qPCR system (Thermo Fischer Scientific). For reactions preparation iTaq Universal SYBR Green supermix (Biorad) was used as previously described by Raymann et al. (2017)90. The forward primer was 27F (5\u00E2\u0080\u0099-AGAGTTTGATCCTGGCTCAG-3\u00E2\u0080\u0099) and the reverse primer was 355R (5\u00E2\u0080\u0099-CTGCTGCCTCCCGTAGGAGT-3).To build a standard curve, a single copy of the V4 16S rRNA gene was cloned into pGEM-T vector93. Dilutions from 1x103 to 1x1012 copies were adjusted. The amplification cycle conditions were set to an initial 95\u00CB\u009AC for 20 sec and 40 rounds of 95\u00CB\u009AC for 3 sec and 60\u00CB\u009AC for 30 sec. Illumina sequencing This section was performed and described by Microbiome Insights. (https://microbiomeinsights.com/). First, bacterial 16S rRNA gene were PCR-amplified with dual-barcoded primers targeting the V4 region (515F 5\u00E2\u0080\u0099-GTGCCAGCMGCCGCGGTAA-3\u00E2\u0080\u0099, and 806R 5\u00E2\u0080\u0099-GGACTACHVGGGTWTCTAAT-3\u00E2\u0080\u0099) (Note that V, H and W are codes of bases for degenerative primers), as per the protocol of Kozich et al. (2013). Amplicons were sequenced with an Illumina MiSeq using the 300-bp paired-end kit (v.3). Sequences were denoised, 40 taxonomically classified using Greengenes (v.13_8) as the reference database, and clustered into 97%-similarity operational taxonomic units (OTUs) with the mothur software package (v.1.39.5) (Schloss et al. 2009), following the recommended procedure (https://www.mothur.org/wiki/MiSeq_SOP; accessed Nov 2017). The general pipeline is better depicted on figure 2.4. Figure 2.4 Analytical Flowchart (Microbiome Insights). From raw data to qualitative and relative abundance analyses. 2.3.2.5.1 Quality Control This section was performed and described by Microbiome Insights. The potential for contamination was addressed by co-sequencing DNA amplified from specimens and from each of the template-free controls and extraction kit reagents processed the same way as the specimens. Operational taxonomic units were considered putative contaminants 41 (and were removed) if their mean abundance in controls reached or exceeded 25 % of their mean abundance in specimens. 2.3.2.5.2 Statistical analysis This section was performed and described by Microbiome Insights. Alpha diversity was estimated with the Shannon index on raw OTU abundance tables after filtering out contaminants. The significance of diversity differences was tested with ANOVA. To estimate beta diversity across samples, we excluded OTUs occurring with a count of less than 3 in at least 10% of the samples and then Bray-Curtis indices were computed. We visualized beta diversity, emphasizing differences across samples, using Principal Coordinate Analysis (PCoA) ordination. Variation in community structure was assessed with permutational multivariate analyses of variance (PERMANOVA) with treatment group as the main fixed factor and using 9999 permutations for significance testing. All analyses were conducted in the R environment. 2.4 Results & Discussion 2.4.1 Susceptibility of non-commensal and commensal bacteria to naturally occurring antimicrobial peptides For the zone of inhibition assays, for both commensal and non-commensal bacteria, the zone of inhibition diameter caused by the tested AMPs was measured 48 h. after the experiment setting. Table 2.3 shows the average of the inhibition scores (cm) and standard error. Examples of the raw data obtained for P. larvae are shown in figure 2.5. 42 Table 2.3 Zone of inhibition assay data. The average inhibition score (cm) in bold and standard error in italics. Figure 2.5 Zone of inhibition assays raw data representative samples for P. larvae: Ampicillin, oxytetracycline, jelleine and melittin (From left to right). P. larvae is susceptible to jelleine and melittin According to the zone of inhibition diameter and based on statistical significance (One-way ANOVA, post-hoc Tukey test p-value < 0.05), jelleine (naturally present in royal jelly AMP; P= 0.002) and melittin (naturally present in honey bee venom AMP; P= 0.026) caused significant inhibition on the growth of P. larvae. Ampicillin and OTC caused inhibition is considered as positive control. Figure 2.6 shows the data points of each inhibition zone (cm) replicate for each AMP, as well as for the positive controls. Treatment Water 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00Ampicillin 1.17 0.03 4.50 0.30 0.80 0.40 2.10 - 3.10 - 6.17 -Oxytetracycline 2.57 0.07 4.73 0.20 2.63 0.15 2.40 - 2.10 - 3.73 -Abaecin 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00Apidaecin 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.33 0.18Defensin I 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 1.27 0.07 0.00 0.00Jelleine 0.47 0.07 0.33 0.03 0.30 0.15 0.43 0.03 0.33 0.17 0.87 0.14Melittin 0.80 0.10 0.40 0.00 0.70 0.20 0.97 0.19 1.17 0.52 0.70 0.08Andropin 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.43 0.26Cecropin 0.33 0.18 0.00 0.00 0.00 0.00 0.50 0.00 0.57 0.03 0.44 0.02Defensin II 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 1.47 0.13 0.00 0.00P. LarvaeB. apis B. asteroides S. alvi B. subtilis E. coli43 Figure 2.6 Each data point represents the inhibition zone diameter caused by each AMP to P. larvae. * Pairwise significant comparison against blank. Non-pathogen and non-colonizer bacteria susceptibility to honey bee and fruit fly naturally occurring AMPs E.coli and B. subtillis, two non-colonizers of the honey bee gut microbiota, were challenged to the same AMPs as P. larvae. The growth for the Gram-negative bacteria, E.coli was inhibited by Defensin I (Apis mellifera; P= 0.006), Defensin II (Defensin I of Drosophila melanogaster; P= 0.001), and melittin (P=0.013). The growth of the Gram-positive B. subtills was inhibited by cecropin (Drosophila melanogaster; P= 0.001), jelleine (P=0.004) and melitin (P=1.0x10-7) (Figure 2.7). * * * * 44 A) B) Figure 2.7 A) Each data point represents the inhibition zone diameter caused by each AMP to E. coli. B) Each data point represents the inhibition zone diameter caused by each AMP to B. subtillis. * Pairwise significant comparison against blank. Susceptibility of commensal bacteria to naturally occurring AMPs Three commensal bacteria species were isolated through enrichment cultures and used to survey their susceptibility to naturally occurring AMPs, from both, fruit fly and honey bees: The Gram-negative B. apis; the core commensal and Gram-negative S. alvi and the Gram-positve and lactic acid bacteria B. asteroides. Any group of AMPs treatments performed on B. asteroides and S. alvi weren\u00E2\u0080\u0099t significantly different from the blank (One-way ANOVA, post-hoc Tukey test, pair-wise comparison against blank p-value >0.05). B. apis was significantly inhibited by both * * * * * * * * * * 45 jelleine (P= .005) and mellitin (5. 0x10-6), and more susceptible to OTC than to ampicillin (Figure 2.8). A) B) C) Figure 2.8 A) Each data point represents the inhibition zone diameter caused by each AMP to B. asteroides B) Each data point represents the inhibition zone diameter caused by each AMP to S. alvi C) Each data point represents the inhibition zone diameter caused by each AMP to B. apis. * Pairwise significant comparison against blank. * * * * * * * * 46 The CGB are more resistant to the naturally occurring AMPs than non-commensal and pathogenic bacterium P. larvae After analyzing the score of inhibition for the two groups of bacteria we worked with (commensal and non-commensal), it was observed that generally, the average of the inhibition score caused by each individual peptide is larger against the non-commensal as depicted in figure 2.9. Figure 2.9 Susceptibility scores commensal vs non-commensal bacteria: Y-axis shows the average inhibition score caused by each peptide to the non-commensal bacteria cohort, while the X-axis shows the average inhibition score caused by each peptide to the commensal. Overall all the AMPs have greater inhibitory effect against non-commensal, observing a cluster of the scores above the diagonal. 47 Figure 2.10 The inhibition effect caused by bee derived AMPs vs fly derived AMPs. To observe if there is any pattern favoring the coexistence between honey bee CGB and humoral factors of adult honey we modeled inhibition scores with categorical variables describing whether bacteria were commensal or non-commensal, whether Gram-positive and Gram-positive and whether AMPs were bee and fly. In this figure the inhibition is clearly larger for the non-commensal for both type of AMPs and overall the honey bee derived peptides seem having stronger inhibitory activity. By having two different sources of AMPs (honey bee and fly), as well as a broad collection of bacteria tested (commensal and non-commensal representatives), we aimed to observe if there is any pattern favoring the concurrence between honey bee CGB and humoral factors of the adult honey bee due to historical coexistence. We modeled inhibition scores with categorical variables describing whether bacteria were commensal or non-commensal, whether bacteria were gram-positive or gram-negative, and whether peptides were bee or fly (Figure 2.9). This model was significant (P=6e-5) and had an adjusted R-squared of 0.15 (F=6.6 on 145 degrees of freedom). However, we acknowledge that this simplified model does not considered the chemical nature of each peptide. 48 Regarding the inhibition exclusively caused to honey bee commensal bacteria (CGB), and contrary to expected, we found that their growth appear to be more inhibited in the presence of honey derived AMPs than fly derived AMPs. When analyzing the inhibition caused to non-commensal bacteria the effect of bee derived honey bee AMPs seems to be also higher for that caused by the fly ones, having honey bee AMPs greater inhibitory effect against the chosen organisms for these assays that fly AMPs (P= 0.04). Another highlight is the fact that the non-commensal bacteria is more sensitive to both groups of peptides (P=0.003), suggesting that in general, the commensal bacteria here tested are better adapted to the host and host-like AMPs mechanisms of action, likely due to being acclimated to their presence (Figure 2.11). Finally, Gram-negative bacteria are significantly more resistant to the chosen peptides than Gram-positive bacteria (P = 0.005). However to validate the previous mentioned, the biophysical properties that confer this hypothetical adaptation of the CGB, additional experiments are required, such as membrane disruption and membrane interaction, which could point out shared response patterns of the two bacterial groups to the selected AMPs. One potential explanation for the less inhibitory effect observed for fly AMPs could be as simple as the limited number of AMPs surveyed. However, it can also be related to the more inconsistent microbial repertoire, meaning that fly AMPs must control to highly specific pathogens to permit the broad range of variable bacterial components94. To validate this last, fruit fly AMPs used could be used to challenge a large cohort of drosophila related bacteria. 49 There is a tissue specific AMP concentration in honey bees. Jelleine and Apisimin (defensin I isoform) are important components of the royal jelly, which is secreted through the salivary glands; melittin is a venom component secreted through the stinger; abaecin, apidaecin, defensing I and hymenoptaecin (this last one not included in this work) are constitutively present in the hemolymph. We decided to further explore if the tissue specific concentration of bee derived AMPs causes bias in terms of higher or lower sensitivity of the bacterial species. This based on the rationale that we are working with gut microbiota bacteria that likely coexist with AMPs common in the hemolymph and throughout the body, but not necessary with more specialized ones. To answer whether there is a significant difference between hemolymph (abaecin, apidaecin and defensin) and non-hemolymph (jelleine and melittin) peptides generally, we used a final regression model in where inhibition scores were modeled with hemolymph status, commensal status and gram-status terms. This model was significant (P= 1e-12, R-squared=0.47, F=27 on 92 DF). We found a strong trend in which non-hemolymph peptides have Figure 2.11 The inhibition effect caused by bee derived AMPs vs fly derived AMPs to each species. The data points for the inhibition caused to non-commensal (Left) and the inhibition cause to commensal bacteria (right). 50 significantly greater inhibitory effect against the bacteria we challenged that hemolymph peptides (P=1e-12). Figure 2.11 shows the data points of inhibition score obtained for either hemolymph or non-hemolymph AMPs for both groups of bacteria. Lastly, we compared the effect of each individual AMP to collectively commensals vs non-commensals. Blank shows the baseline of 0.0 cm inhibition caused to both groups. Apidaecin and andropin show a slightly negative effect (but not statistically significant) against non-commensals; melittin is inhibitor for both groups as expected for its reported mechanism of action of indistinctively pore formation82. However, slightly higher for non-commensal. The actions of jelleine were similar to those observed for melittin. Both defensins are similar to this last pattern, while cecropin seems to be more effective against non-commensal (figure 2.13). Comparing the effect of the positive control antibiotics used we can see a clear higher sensitivity Figure 2.12 Hemolymph vs non-hemolymph AMPs. The data points caused to non-commensal (left). The data points caused to commensal bacteria (right). The inhibition effect caused by hemolymph (first boxplot) derived AMPs vs non-hemolymph (second boxplot) derived AMPs of the honey bee. 51 of the commensal to OTC than to AMP and the opposite for non-commensal, which are more sensitive to ampicillin (figure 2.14). Figure 2.13 Close up to the inhibition caused by each AMP. Each panel contains the zone of inhibition data points corresponding to each AMPs to both cohorts of bacteria. Figure 2.14 Close up of the inhibition caused by antibiotics. Minimum inhibitory concentration of active compounds against P. larvae After the susceptibility assays, we found that only jelleine and melittin were statistically significant inhibitors of P. larvae growth, therefore we proceed to look at their MIC. We 52 surveyed the MIC for jelleine and melittin through 96 well plate microdilution assay. Figure 2.15 shows the 24 h. growth curves obtained for each bacterial isolate challenged at different concentration ranges of the AMPs surveyed, as well as for the positive control (X-axis equals to time (hours) and Y-axis equals to OD600). A flat curve means no growth along time while having a slightly s-shaped curve means normal growth; considering the saturation of growth expected for E. coli we reached a plateau when growth is optimal (blank and likely low concentration of antibiotics or AMPs); P. larvae duplicates in a lower rate so reaching the plateau at optimal conditions (blank and likely low concentration of antibiotics or AMPs) is expected only after 24 h. Each panel of figure 2.15 shows the representative data obtained for each case: E.coli exposed to positive control OTC; E. coli exposed to jelleine; E. coli exposed to melittin; P. larvae exposed to OTC (positive control); P. larvae exposed to jelleine; P. larvae exposed to melittin. The effective concentration for jelleine and melittin at which P. larvae growth is inhibited is 100 \u00C2\u00B5g/ml and 6.25 \u00C2\u00B5g/ml respectively, which highlights that probably just melittin is a realistic option as prophylactic. 53 Figure 2.15 Representative data of 96-well plate microdilution assays. Growth curves for E. coli challenged by oxytetracycline, jelleine and melittin (up). Growth curves for P. larvae challenged by oxytetracycline, jelleine and melittin (down). 2.4.2 Effects on the gut commensal bacteria when exposed to jelleine and melittin AMPs as potential treatments. Changes in the composition: relative abundance After five days of establishing CGB communities in the newly emerged bees, we fed them with sterile sucrose syrup supplemented with AMPs at MIC (found from previous experiment) and OTC prophylactic dose concentration. After 5 d of antimicrobial challenge the bacterial community DNA was extracted and used for qPCR and 16S rRNA V4 amplicon deep sequencing. The sequencing data quality permitted only limited scale for most of the taxonomic groups categorized. The sequencing data provided proportions for the finest of the reached scales summarized in figure 2.15. S. alvi is the only species shown, the rest is at the genus level (unclassified Lactobacillus and unclassified Bifidobacterium), family (Acetobacteracea and Lactobacillacea), order (Lactobaccillales and Rhizobiales) and class (gammaproteobacteria). 54 For those groups that showed less than 1% abundance we grouped as others (Neisseriaceae (family) Gluconacetobacter (genus) Bifidobacteriaceae (family) SUP05 (species)). For those with the lowest resolution (eg. unclassified bacteria or proteobacteria (phylum, very general)) we decided just to group them as unclassified. Table 2.4 shows the percentage obtained for each group. For a simplification of the overall distribution of these groups along the treatments replicates and average we show figure 2.16. Table 2.4 Proportions of taxonomical groups per sample. Sample S. alvi (species) Lactobacillus (genus) Bifidobacterium (genus) Acetobacteraceae (family) Lactobacillaceae (family) Lactobacillales (order) Rhizobiales (order) Gammaproteobacteria (class) others unclassfied totalB1 25.56% 24.34% 4.02% 25.27% 1.58% 0.43% 11.90% 5.97% 0.35% 0.37% 99.81%B2 33.22% 15.87% 11.84% 14.07% 3.12% 0.32% 15.74% 4.79% 0.48% 0.38% 99.84%B3 18.25% 35.47% 6.40% 21.55% 3.38% 0.49% 0.72% 12.44% 0.60% 0.48% 99.78%O1 6.73% 34.92% 18.87% 10.52% 3.27% 0.58% 18.83% 4.59% 0.76% 0.57% 99.63%O2 15.41% 30.67% 16.93% 14.75% 4.83% 0.55% 9.29% 5.49% 1.03% 0.70% 99.66%O3 26.22% 26.58% 12.93% 14.70% 3.17% 0.57% 0.66% 13.61% 0.78% 0.59% 99.80%J1 20.44% 27.49% 20.31% 9.19% 3.27% 4.00% 1.51% 11.41% 1.18% 0.96% 99.76%J2 18.35% 41.32% 20.95% 4.01% 3.80% 0.48% 2.39% 7.40% 0.57% 0.60% 99.86%J3 23.99% 25.04% 12.28% 1.24% 2.62% 1.11% 26.05% 6.52% 0.43% 0.50% 99.77%M1 11.85% 21.40% 19.32% 3.90% 3.25% 0.71% 32.04% 5.85% 0.70% 0.69% 99.71%M2 7.72% 30.71% 10.43% 4.08% 3.30% 1.97% 36.85% 3.78% 0.42% 0.48% 99.73%M3 15.32% 27.87% 12.95% 19.49% 3.92% 1.17% 3.33% 14.06% 0.98% 0.71% 99.80%BL 66.07% 0.77% 1.74% 0.04% 0.06% 0.00% 0.19% 30.80% 0.33% 0.00% 100.00%Figure 2.16 Taxonomic composition plot. Composition for each sample replicate (right) composition of the average for treatment (Left). Each color represents the corresponding taxonomic group categorized. This figure just represents the relative abundance of each group identified per sample/treatment. 55 Changes in the diversity The results here reported were provided by Microbiome insights: Differences in alpha diversity or Shannon diversity index (measure of richness and evenness in the sample) and beta diversity (similarity among microbes) relative to the blank are shown in figure 2.17. ANOVA determined no significant differences in the Shannon diversity index. The following model was used: value ~ Group. Permutational analysis of variance (adonis R function, or Permanova) determined significant differences in beta-diversity among Group factor(s). The following adonis model was used: t(otu) ~ Group. 56 Changes in size of the gut microbial community The qPCR data was analyzed by aid of the standard curve built with the Ct values for the plasmid dilutions and their known DNA concentration. From the linear equation obtained we estimated number of copies of the 16S rRNA gene for each sample. We did a total of 4 trials with three technical replicates for each biological replicate (three). Figure 2.18 (left) shows the raw number Figure 2.17 Shannon index (up) and Beta diversity (down). No significant differences in the Shannon diversity index. Beta-diversity index threw significant differences for among group factors which can be observed by a slight cluster of the datapoints. 57 of 16S rRNA copies (average per treatment), including the number of copies for BL (baseline). The baseline number of copies was expected to be minimal since it represents the 16S rRNA gene copies of the CGB present just after bees emerged, and in theory they emerge gnotobiotic 19,62. Having the baseline is a good proof of principle since indicates that the commensal community was successfully set (when comparing size and composition of baseline to blank)62. Then, we have the total number of 16s rRNA gene copies obtained for each of the treatments: blank (B) (fed exclusively with sterile syrup) jelleine (J) (fed with 100 \u00C2\u00B5g/ml jelleine in syrup solution) melittin (M) (fed with 6.25 \u00C2\u00B5g/ml melittin in syrup solution) and OTC (O) (fed with 6.8 \u00C2\u00B5g/ml OTC in syrup solution). For different reasons, including lack of defecation and a considerable lower amount of social interactions, honey bees in captivity can\u00E2\u0080\u0099t match exact size and composition of the hive present honey bees. However, the blank was expected to show the most similar patterns for both size and composition to hive bees microbiomes, and therefore take it as a status of health reference (non-dysbiotic gut). Unexpectedly, we obtained higher number of 16S rRNA gene copies for the antibiotics/AMPs treatments than for the blank (Figure 2.18), which was lower to a million copies of this gene, contrasted to the hundred million total reported for blanks treated in a similar manner90. We could expect that the microbiota size would not be affected by the treatments, but having a larger size in the treatments than the blank size (no perturbation) can indicate a few things: One reason could be the presence of pathogens, which is likely considering newly emerged bees were fed with hindguts from adult bees of the hive that are not exempt to be infected by pathogenic or opportunistic bacteria (or trophallaxis between them and the nurses provided for the same goal). In fact between the taxonomic groups classified as others, we found bacilli in the blank samples, which could be P. larvae (cross contamination), and also a considerable amount of Gammaproteobacteria, which either could be G. apicola, core 58 but commonly correlated to disease , or Serratia mancerens (opportunistic bacteria of several insects including the honey bee)90. This last suggests that feeding infected bees with antibiotics/AMPs actually aids to kill these pathogens and flourished a more healthy commensal community, which should be in size close the characteristic billion of bacterial cells that bees have in the open environment61. Another reason to explain this observed increase of total DNA is the fact that qPCR data reflects the total bacterial DNA for either alive or dead bacterial cells. The application of antibiotics/AMPs at MIC doesn\u00E2\u0080\u0099t interrupt cell replication completely. The number of living cells is thus replenished and stays constant, therefore we observed an accumulation of dead cell DNA95. A clearer picture of the size patterns is shown in figure 2.18 (right) in which is shown the 16S rRNA gene copies number relative to the blank. *** *** * Figure 2.18 Total number of 16S rRNA gene copies. Absolute, including baseline data (left). Normalized and relative to the blank data (right). For right figure only, ANOVA pairwise significant comparison against blank: * (P < 0.05) ** (P<0.01) and *** (P<0.001). 59 Changes in the total abundances of the core species Finally we determined the total number of bacterial cells of the most relevant species (CGB identified): S. alvi, lactobacillus species and Bifidobacterium species. The total number of copies for each sample was multiplied by its corresponding proportion, and then divided by the number of 16S rRNA operons per genome for each bacteria (Bifidobacterium=2, Lactobacillus= 4 and S. alvi = 4)90. Figure 2.19 shows the total number of cells per bacterial group relative to the blank. Bifidobacterium species and Lactobacillus species size for jelleine, melittin and OTC treatments were statistically significant different to the blank, while S. alvi was not statistically significant different to the blank for any of the treatments highlighting its resiliency. Figure 2.19 Total number of 16S rRNA gene copies per species. Bifidobacterium species, Lactobacillus species and S. alvi (panels from left to right). ANOVA pairwise significant comparison against blank: * (P < 0.05) ** (P<0.01) and *** (P<0.001). * ** ** *** *** *** 60 2.5 Conclusion Through this body of work the primary goal was to identify naturally occurring honey bee AMPs as candidates to be expressed in a therapeutic probiotic. The criteria for the selection of the candidates include that these AMPs are active against P. larvae in a highly specific manner and that the overall CGB are not disrupted. As an initial approach we had planned host-microbiome challenge experiments, in which we were trying to observe changes in the humoral factors expression profiles according to the ecological niche of the bacterial challenge. However, the results for this set-up weren\u00E2\u0080\u0099t included in this work because of the March shutdown. Moving to the susceptibility of specificity assays we considered safer to attempt to use naturally occurring AMPs of the honey bee for biological and ethical reasons, but tested three fruit fly AMPs to increase the understanding of specificity. We can conclude that there exist intrinsic mechanisms that evolutionary adapted both, the adult honey bee and its commensal bacteria, to coexist and keep a functional status. As an evidence, commensal bacteria are significantly more resistant to both fly and bee-derived AMPs than non-commensal bacteria. In fact our data also suggests that commensal bacteria did well against AMPs of fly, likely because of the homology between insect derived AMPs, at least among the commonest families of these humoral factors. Interestingly, non-hemolymph AMPs show higher inhibitory activity against CGB markedly, which was not the case for hemolymph AMPs, evidencing that the adaptation derived of historical coexistences. Additionally, bee derived peptides seems to be a very efficient repertoire of collection, considering the limited amount of them present in the honey bee, but still more inhibiting to both 61 cohorts (commensal and non-commensal) that the fruit fly ones. Moreover, B. asteroides was the most resilient bacterium when challenged by AMPs of both sources (bee and fly). B. asteroides is broadly used as probiotic, suggesting it has the ability to colonize different animal systems, and actually because it is acquired through a more unpredictable approach (floral environment) demands high resilience. We also concluded that the effect caused by jelleine and melittin to the overall community is mild since the total size do not show decrement. However, more validation is needed such as the use of an additional positive control of OTC, since previously much higher concentration for this antibiotic has shown a harsh decrement in the overall size of the CGB (450 \u00C2\u00B5g/ml). Likely an additional experiment to explain the blank smaller size is required, such as screening for a particular pathogen or antibiotics resistance genes, or in a more drastic approach to repeat the experiment increasing the used antimicrobials concentration. Based on the preliminary results presented here, we conclude that the use of jelleine and melittin is safe since the ratios of abundance for the core components are preserved. Adding up to the fact that are realistically usable as prophylactics for biological and ethical reasons. 62 Chapter 3: Engineering of honey bee gut commensal bacterium Snodgrasella alvi for the secretion of naturally occurring antimicrobial peptides 3.1 Introduction and rationale There is still a very conservative stance around the world regarding the application of genetic engineering in living organisms96. Despite this concern, considering the great influence that microbiomes have on their host\u00E2\u0080\u0099s phenotypes, the potential benefits of engineering core or temporary resident microbes to stimulate health are difficult to ignore. Some bacterial strains have been previously enhanced with the particular purpose of fighting disease through heterologous protein expression and secretion approach. For example, Forkus et al. (2017) engineered Escherichia coli to express and secrete the antimicrobial peptide microcin J25, decreasing the recurrence of Salmonella enterica in turkey gastrointestinal tracts up to 97%97. Hwang et al. (2017) engineered E. coli to secrete an anti-biofilm enzyme against the pathogen Pseudomonas aeruginosa, preventing its colonization of human gut98. The development of engineered probiotics that express and secrete antimicrobial compounds is highly rewarding since the bacterium provides a renewable source of these compounds, analogous to a host constitutive component, and also offers a biological backbone for delivery. Antimicrobial peptides are part of the first line of defense employed by multicellular organisms against pathogens. These short peptide chains are composed of between 12 to 100 amino acids, typically with an overall positive charge (net charge of +2 to +9)99. Their antimicrobial nature is primarily due to the interactions with the negatively charged cell membrane of bacterial cells, leading to direct membrane breach, in most of cases. This mechanism likely prevents the 63 development of resistance against them; contrasting with the antibiotics mode of action that triggers SOS responses leading to fast mutations100. Also, bacteria can be intrinsically resistant to certain antibiotics and share genes by horizontal transfer, such as those involved in the modification of the antimicrobial target, drug efflux and inactivation of the antibiotic by hydrolysis or modification101. However, some AMPs are also able to exert intracellular inhibitory activities either as a complementary or main mode of disruption102, which highlights its diversity. Nonetheless, there are some challenges involved in the expression of AMPs in prokaryotes: First, when expressing an eukaryotic protein into a prokaryote cell, not all the correct molecular machinery may be available (e.g., modification enzymes), and the codon usage may be incompatible103; second, in this particular case, the antimicrobial peptide could directly kill the host; and third, the host cells quality control machinery can target the peptide and degrade it104. Few strategies have been developed to overcome these challenges103,104. For the eukaryote to prokaryote transition, codon optimization is one general suggestion, but it requires understanding the codon bias for the host organism. Focusing on the stability of the peptide in the host cell, fusion/chaperon proteins can prevent the degradation by the host cellular regulation mechanisms104. A convenient strategy to prevent the degradation of the peptide inside the host is to secrete it in the unfolded state, which can be achieved by targeting the protein to a translocation pathway104,105, although secretion of a heterologous protein is at the same time another complete challenge. Most of the efforts to develop heterologous secretors strains have been for industrial purposes106, but there is a great potential in the employment of similar paths for design of therapeutics. 64 The Sec and Tat secretion pathways are the two main secretory pathways in bacterial cells. Both rely on signal peptides that provide allosteric properties to the targeted proteins to interact with specific protein channels known collectively as secretion machinery107. Approximately 96% of E. coli exportome (non-cytoplasmic proteins) uses the Sec-dependent pathway108, therefore most of the current efforts directed to export proteins are based on the exploiting its machinery106. Signal peptides of this pathway are unique to each protein but share physicochemical properties such as three characteristic regions as depicted in figure 3.1. The N-terminal region is positively charged, the intermediate region is hydrophobic, finally, the C-terminal region has a polar nature 109. Figure 3.1 Sec-dependent signal peptides domains properties and architecture. The n-region is following the N-terminus and is positively charged; the H-region stands for helical hydrophobic region; the c-region is the terminal region of the signal peptide and is slightly polar; the cs (cleavage site) usually presents a characteristic AXA motif; the mature region domain corresponds to the exported protein. Attaching a putative signal peptide to the protein target could potentially direct it to secretion106. However, there is no way to predict which candidate signal peptide will successfully secrete a target protein, or at which rate. Validation of the individual signal peptide is needed anytime one of the elements of a secretion system is replaced (expression organism and peptide or protein to secrete)110. Different authors have found that there is not sequence homology between the signal peptides of determined families of secreted proteins and little correlation between protein 65 sequence and signal peptide sequence for prediction106. The best way to identify candidates for the secretion of heterologous proteins is the screening of a large set of signal peptides or a library of either homologous (of the organism being used) or heterologous (of other organisms). Although it has been reported that homologous peptides generally have higher yields111. Testing the ability to drive protein secretion of putative homologous signal peptides in Lactobacillus plantarum WCFS1 was very illustrative reaffirming that the secretion rate is not only dependent on the candidate signal peptide but of the target protein even when compared with gold standard heterologous signal peptides (used for secretion system previously with high reported yields)111 . For organisms which have not had any previous characterization of their extracellular components, it is recommended to survey their genome repertoire for the identification of the homologous signal peptides. Boekhorst, J., et al. (2006) predicted the secretome of Lactobacillus plantarum WCFS1 by identifying the proteins with Sec and Tat pathways signal peptides, parallel to the annotation and characterization of these proteins112. They provided very valuable information of the likely microbe-environment interactions by identifying secreted enzymes, transporters, regulators and adherence proteins. Another way to determine the secreted components involves proteomics experiments for the analysis of secreted fractions; either analyzing cell-free supernatant or shaving cells with trypsin to recover cell surface proteins and peptides113,114. The advantage of the in silico approaches include the prediction of the signal peptide sequence and cleavage site. Although in vivo experiments allow the identification of the dynamics of these peptides and their secretion ratios. However, using in silico and in vivo approaches paired can be complementary, particularly to discern between attached extra-cellular 66 proteins and actually secreted proteins. Furthermore the identification of non-classical signal peptide-dependent secreted proteins can be surveyed when comparing different approaches. After designing the secretion strategy, another challenge would be the genetic manipulation of CGB. For the engineering of S. alvi there are already some available tools including the bee toolkit (BTK), a broad-host-range plasmid developed by Leonard et al. (2018) for modification of bee commensal Proteobacteria. The transfer of the BTK plasmid to a recipient cell from a donor cell is achieved by conjugation. This molecular toolbox allows expression of heterologous proteins (such as antibiotic\u00E2\u0080\u0090resistant genes and fluorescent markers) and repression of chromosomal genes via a CRISPRi approach. BTK contains 11 promoters, both inducible and constitutive. The authors validated the BTK on different strains of the following species including S. alvi, G. apicola, Serratia marcescens and Bartonella apis115. Additionally, S. alvi was recently used as dsRNA vector for targeting deforming wing virus and Varroa destructor essential genes32. Some highlights include the stability of the genetic engineered strain that could remain in the gut environment for several days; even managing colonize honey bees co-housed with the recipient bees with no negative effects on the bees\u00E2\u0080\u0099 survival were reported32. While certainly S. alvi can be challenging to engineer when compared to more well-studied microorganisms such as E. coli, there are some tools now available that can facilitate these attempts. 3.2 Research goals The main goal for the experiments here described is to design a strategy for engineering S. alvi to express and secrete the AMPs jelleine and melittin, both with demonstrated activity against the 67 causative bacteria of America Foulbrood, P. larvae (Chapter 2). The general aim is to develop a secretion strategy which consists of two minor aims: The first one involves the identification of the Sec-dependent signal peptides available for homologous secretion of proteins; and the second involves the design of mass spectrometry proteomics experiments to identify higher secretion activity signal peptides. We considered this approach will also indirectly aid in the annotation of S. alvi genome. The second general aim is to design the expression system to transform S. alvi bacterial cells. Our final project goal is to validate expression, secretion dynamics and the activity of the AMPs as part of the probiotic. 3.3 Methods 3.3.1 The Sec-dependent signal peptide library of Snodgrasella alvi The translated genome of S. alvi wkB2 was retrieved from Uniport (IUP00019668) and it contained a total of 2,295 proteins. All the sequences were used as input for SignalP 5.0116. As selection criteria for predicting the Sec pathway dependent extra cellular proteins of S. alvi we filtered for proteins that scored more than 0.7 in the SPI-type likelihood score, which tries to identify signal peptides that can be cleaved by SPase I (Sec/SPI) (Figure 3.2).68 Figure 3.2 Signal peptide library through SignalP 5.0 This involves retrieving type I signal peptide containing proteins list. 3.3.2 Top secretion activity signal peptides experiment This set of experiments have the goal to identify candidates signal peptides to direct heterologous secretion of the honey bee AMPs in S. alvi. First we obtained the whole cell and secretion fraction and the proceeded with protein extraction. Protein sample preparation included the protein solubilization, proteolytic digestion and the peptide clean up. Then we proceeded with LC-MS/MS and data analysis. S. alvi protein sample preparation: secretion and whole cell fraction protein extraction The overall outline of these experiments can be observed in figure 3.3. First, one individual axenic culture of S. alvi was harvested (picking all the colonies) and resuspended in 50 \u00C2\u00B5l of PBS (each plate is one replicate). The resuspended bacterial colonies were centrifuge at 3000 x g for Uniprot\u00E2\u0080\u00A2 translated genome: 2295 proteins SignalP 5.0 \u00E2\u0080\u00A2 putative signal peptidesFilter for Sec/SPI score > 0.7\u00E2\u0080\u00A2 218 putative Sec-I type signal peptide -containing sequences 69 15 min at room temperature. The supernatant (secretion fraction) was moved to a clean 1.5 ml tube and the pellet (whole cell fraction) was washed twice with PBS. Whole cell fraction was resuspended in lysis buffer (100 mM ammonium bicarbonate, 2% SDS), vortex roughly for 30 sec and boiled for 5 min in a thermomixer (Eppendorf). Total protein content of both fractions was measured based on its absorbance at 280 nm using NanoDrop One (Thermo Fisher Scientific). An aliquot of each sample was taken according to the volume needed to work with 20 \u00C2\u00B5g of protein. To precipitate the proteins 4 volumes of cold 100% acetone were added to each sample, vortex roughly and incubated at -20o C overnight. Then samples were centrifuged for 10 minutes at 13,000 x g. The visible pellet was washed with cold 80% acetone three times by centrifuging at the same previous conditions. After last centrifuge round the acetone was discarded and the samples were dried in the vacuum centrifuge (SpeedVac, ThermoFisher). The pellet was resuspended with equal volumes of water and trifluoroethanol. Protein denaturation: alkylation and reduction 20 \u00C2\u00B5l of protein solution were mixed with one volume of TRIS buffer (pH 8.5) and 5 \u00C2\u00B5l of 0.1M TCEP (tris(2-carboxyethyl)phosphine). The mixture was incubated for 20 min at room temperature. Later on, 5 \u00C2\u00B5l of 0.4 M chloroacetamide was added and then incubated at 95\u00CB\u009AC for 10 minutes. 40 \u00C2\u00B5l of 50 mM ammonium bicarbonate were added and mixed roughly. Proteolytic digestion and peptide clean up An initial 2 h digestion was set by adding 0.4 \u00C2\u00B5g of Lys C. Following overnight trypsin digestion was set up adding 0.4 \u00C2\u00B5g of trypsin. The sample was desalted and purified on STAGE-tips 84. The columns were moved to another tube for elution by adding 80 \u00C2\u00B5l of buffer B (80% 70 Acetonitrile, 0.1% formic acid) twice. Samples were vacuum centrifuged (SpeedVac, ThermoFisher). Liquid chromatography and tandem mass spectrometry Purified peptides were analyzed using a quadrupole \u00E2\u0080\u0093 time of flight mass spectrometer (Impact II; Bruker Daltonics) coupled on-line to an Easy nano LC 1000 HPLC (ThermoFisher Scientific), using a Captive spray nanospray ionization source (Bruker Daltonics). The fused silica analytical column was custom made and packed with 1.9 \u00CE\u00BCm-diameter Reprosil-Pur C-18-AQ beads (Dr. Maisch, www.Dr-Maisch.com) and it was heated to 50\u00C2\u00B0C using an in-house build temperature controller. Buffer A consisted of aqueous 0.1% formic acid and 2 % acetonitrile, and buffer B was 0.1% formic acid and 90 % acetonitrile. Samples were resuspended and loaded in buffer A. The Impact II was set to acquire in a data-dependent auto-MS/MS mode with inactive focus fragmenting the 20 most abundant ions after each full-range scan from m/z 200 Th to m/z 2000 Th. The isolation window for MS/MS was 2 to 3 Th, depending on parent ion mass to charge ratio; and the collision energy ranged from 23 to 65 eV depending on ion mass and charge 4. Parent ions were then excluded from MS/MS for the next 0.3 min and reconsidered if their intensity increased more than 5 times. Singly charged ions were excluded. Mass accuracy was not allowed to exceed 10 ppm. Data analysis ByonicTM (v3.4.0 \u00E2\u0080\u0093 ProteinMetrics Inc.) was used for peptide and protein identification using the reference proteome of S. alvi (target proteome) and Bos taurus (to identify contaminants present in the samples as consequence of the bacteria growth media), followed by 1% FDR 71 filtering. For a total of six proteins lists (secreted fraction/whole cell fraction and three replicates each), proteins were ranked by their total intensity values (Figure 3.3). To obtain the list of top secreted proteins we attempted to calculate the secretion ratio of all the proteins in the secreted fraction for each protein list. The secretion ratio was defined as the position value in the total intensity rank for the whole cell fraction list, minus the position value in the rank for the secreted fraction list (each secretion protein list is paired to a whole cell protein list). Then the average of the secretion ratio was calculated using the three replicates. Because of lack of reproducibility between the replicates we explored each one individually. Figure 3.3 Secretion ratio experiment overview: 1. Bacterial colonies are harvested and suspended in PBS for further fractionation 2. The whole cell pellet is separated from the supernatant, washed and lysate 3. Protein sample preparation included the protein denaturation, solubilization, proteolytic digestion and peptide clean-up 4. LC MS/MS 5. Data analysis. 72 3.3.3 Plasmid design The peptide coding sequence was designed based on the deposited sequence in the National center of Biotechnology Information (NCBI) of A. mellifera\u00E2\u0080\u0099s jelleine and melittin precursors. Promotor CP25 and terminator pBP-BBa_B0015 sequences were retrieved from igem (http://parts.igem.org/Part:BBa_K1509003; http://parts.igem.org/sequencing/part_analysis.cgi?part=BBa_B0015). The sequences of the signal peptide chosen was predicted through SignalP 5.0 server (http://www.cbs.dtu.dk/services/SignalP/). The restriction enzymes were chosen according to the available restrictions sites of the plasmid pBTK503, part of the bee toolkit115. The cassette synthesis was made by GeneArt (Thermo Fisher Scientific) adding the feature of codon optimization for Proteobacteria (E. coli). 3.4 Results & Discussion 3.4.1 The Sec-dependent pathway signal peptide library of Snodgrasella alvi The translated genome of S. alvi (2,295 protein sequences) was used as input for SignalP 5.0 for the prediction of extracellular proteins containing Sec-dependent type I signal peptides (cleaved by SPase I) with a score higher than 0.7. A total of 218 sequences passed this criteria. 3.4.2 Top secretion ratio activity signal peptides: candidates signal peptides to direct heterologous secretion Mass spectrometry-based proteomics was used to analyze the protein composition of S. alvi secreted and whole cell fractions. After the identification of the proteins present in both fractions (Byonic search and 1% FDR filter) and the removal of the Bos Taurus hits, we ranked the 73 remaining proteins by their total intensity values. Figure 3.5 shows a single replicate example of the protein lists for both fraction obtained after ranking them by total intensity. The three replicates of the secretion fraction protein lists have a similar number of proteins (50 for replicate 1 and 2, and 57 for replicate 3), while the whole cell protein lists have more heterogeneous numbers (100 for replicate 1, 27 for replicate 2 and 34 for replicate 3). Ideally, the secretion ratio would be determined as the average of the difference between the rank values in the whole cell list minus the rank value of the secreted protein list for each protein. Despite our efforts to prioritize reproducibility, attempting to average the secretion ratio from the three replicates, only two proteins were constantly present in all of them, so those are what we considered to proceed with analysis. Theoretically, a protein that shows high intensity in the secreted protein list would show low intensity (or not be even detectable) in the whole cell protein list. However, the two chosen candidate proteins were not present in any of the whole cell protein lists. For this scenario we decided to assume that they were undetectable because a very low intra-cellular concentration, categorizing as \u00E2\u0080\u009Chighly secreted\u00E2\u0080\u009D. Disappointingly, none of the two proteins carries a signal peptide according to SignalP 5.0 prediction. Figure 3.5 shows the prediction obtained for 2-oxoglutarate dehydrogenase E1 component and a putative membrane-associated, metal-dependent hydrolase. 74 A) Secreted fraction Rank for intensity DescriptionTotalIntensityProteinDB number1 >tr|X2HDU7|X2HDU7_9NEIS Uncharacterized protein OS=Snodgrassella alvi wkB2 OX=1196094 GN=SALWKB2_1885 PE=4 SV=1 308836.0 15302 >tr|X2H8E4|X2H8E4_9NEIS Uncharacterized protein OS=Snodgrassella alvi wkB2 OX=1196094 GN=SALWKB2_2277 PE=4 SV=1 185198.0 10003 >tr|X2H8J6|X2H8J6_9NEIS 2-oxoglutarate dehydrogenase E1 component OS=Snodgrassella alvi wkB2 OX=1196094 GN=SALWKB2_1293 PE=4 SV=1 183950.0 18924 >tr|X2H5L2|X2H5L2_9NEIS Peptidyl-prolyl cis-trans isomerase OS=Snodgrassella alvi wkB2 OX=1196094 GN=SALWKB2_1272 PE=3 SV=1 179712.0 5555 >tr|X2H8Y5|X2H8Y5_9NEIS Uncharacterized protein OS=Snodgrassella alvi wkB2 OX=1196094 GN=SALWKB2_0221 PE=4 SV=1 172910.0 3686 >tr|X2HJA8|X2HJA8_9NEIS Phosphoadenylyl-sulfate reductase [thioredoxin] / Adenylyl-sulfate reductase [thioredoxin] OS=Snodgrassella alvi wkB2 OX=1196094 GN=SALWKB2_2224 PE=3 SV=1153632.0 14197 >tr|X2H7E9|X2H7E9_9NEIS Glutamate racemase OS=Snodgrassella alvi wkB2 OX=1196094 GN=murI PE=3 SV=1 129272.0 12268 >tr|X2H521|X2H521_9NEIS MFS permease OS=Snodgrassella alvi wkB2 OX=1196094 GN=SALWKB2_0158 PE=4 SV=1 94480.0 15079 >tr|X2HCI3|X2HCI3_9NEIS Diaminopimelate epimerase OS=Snodgrassella alvi wkB2 OX=1196094 GN=dapF PE=3 SV=1 93874.0 145110 >tr|X2HE18|X2HE18_9NEIS Outer-membrane lipoprotein carrier protein OS=Snodgrassella alvi wkB2 OX=1196094 GN=SALWKB2_0434 PE=3 SV=1 90624.0 199211 >tr|X2H924|X2H924_9NEIS ANK_REP_REGION domain-containing protein OS=Snodgrassella alvi wkB2 OX=1196094 GN=SALWKB2_1518 PE=4 SV=1 79916.0 71512 >tr|X2H3G9|X2H3G9_9NEIS Transcriptional regulator, AraC family OS=Snodgrassella alvi wkB2 OX=1196094 GN=SALWKB2_0562 PE=4 SV=1 77522.0 183013 >tr|X2HEM1|X2HEM1_9NEIS YciO family OS=Snodgrassella alvi wkB2 OX=1196094 GN=SALWKB2_2201 PE=3 SV=1 75666.0 166814 >tr|X2H365|X2H365_9NEIS Uncharacterized protein OS=Snodgrassella alvi wkB2 OX=1196094 GN=SALWKB2_0437 PE=4 SV=1 67630.0 23915 >tr|X2H5P3|X2H5P3_9NEIS UvrABC system protein B OS=Snodgrassella alvi wkB2 OX=1196094 GN=uvrB PE=3 SV=1 66728.0 44216 >tr|X2H6S6|X2H6S6_9NEIS Uncharacterized protein OS=Snodgrassella alvi wkB2 OX=1196094 GN=SALWKB2_1702 PE=4 SV=1 62232.0 163617 >tr|X2H334|X2H334_9NEIS Uncharacterized protein OS=Snodgrassella alvi wkB2 OX=1196094 GN=SALWKB2_0402 PE=4 SV=1 49980.0 78218 >tr|X2H8Q9|X2H8Q9_9NEIS Putative RNA polymerase OS=Snodgrassella alvi wkB2 OX=1196094 GN=SALWKB2_1373 PE=3 SV=1 38872.0 79819 >tr|X2HAB7|X2HAB7_9NEIS Valine--tRNA ligase OS=Snodgrassella alvi wkB2 OX=1196094 GN=valS PE=3 SV=1 38522.0 171720 >tr|X2HGD4|X2HGD4_9NEIS O-succinylhomoserine sulfhydrylase OS=Snodgrassella alvi wkB2 OX=1196094 GN=metZ PE=3 SV=1 37376.0 28521 >tr|X2H7M6|X2H7M6_9NEIS Omega-amino acid--pyruvate aminotransferase OS=Snodgrassella alvi wkB2 OX=1196094 GN=SALWKB2_1043 PE=3 SV=1 37362.0 97522 >tr|X2H4W9|X2H4W9_9NEIS Replicative DNA helicase OS=Snodgrassella alvi wkB2 OX=1196094 GN=SALWKB2_1077 PE=3 SV=1 37044.0 8623 >tr|X2HEG9|X2HEG9_9NEIS 1-deoxy-D-xylulose-5-phosphate synthase OS=Snodgrassella alvi wkB2 OX=1196094 GN=dxs PE=3 SV=1 35660.0 227924 >tr|X2HHA4|X2HHA4_9NEIS Ribonuclease H OS=Snodgrassella alvi wkB2 OX=1196094 GN=rnhA PE=3 SV=1 35202.0 205325 >tr|X2H5F0|X2H5F0_9NEIS Cytidylate kinase OS=Snodgrassella alvi wkB2 OX=1196094 GN=cmk PE=3 SV=1 35078.0 118526 >tr|X2HDY9|X2HDY9_9NEIS L-Proline/Glycine betaine transporter ProP OS=Snodgrassella alvi wkB2 OX=1196094 GN=SALWKB2_1945 PE=4 SV=1 34560.0 145227 >tr|X2HDB9|X2HDB9_9NEIS 50S ribosomal protein L33 OS=Snodgrassella alvi wkB2 OX=1196094 GN=rpmG PE=3 SV=1 33748.0 165928 >tr|X2H971|X2H971_9NEIS Uncharacterized protein OS=Snodgrassella alvi wkB2 OX=1196094 GN=SALWKB2_0326 PE=4 SV=1 33564.0 68629 >tr|X2HDQ9|X2HDQ9_9NEIS Outer membrane component of tripartite multidrug resistance system OS=Snodgrassella alvi wkB2 OX=1196094 GN=SALWKB2_0294 PE=3 SV=133178.0 161030 >tr|X2HAK9|X2HAK9_9NEIS UDP-N-acetylenolpyruvoylglucosamine reductase OS=Snodgrassella alvi wkB2 OX=1196094 GN=murB PE=3 SV=1 31454.0 14331 >tr|X2H9Z9|X2H9Z9_9NEIS Chaperone protein HscA homolog OS=Snodgrassella alvi wkB2 OX=1196094 GN=hscA PE=3 SV=1 30786.0 64132 >tr|X2HCK3|X2HCK3_9NEIS Transcription termination/antitermination protein NusA OS=Snodgrassella alvi wkB2 OX=1196094 GN=nusA PE=3 SV=1 30050.0 99233 >tr|X2HD97|X2HD97_9NEIS PPK2 domain-containing protein OS=Snodgrassella alvi wkB2 OX=1196094 GN=SALWKB2_0159 PE=4 SV=1 29810.0 211734 >tr|X2HBS4|X2HBS4_9NEIS Adenylate kinase OS=Snodgrassella alvi wkB2 OX=1196094 GN=adk PE=3 SV=1 29222.0 214535 >tr|X2HB53|X2HB53_9NEIS Uncharacterized protein OS=Snodgrassella alvi wkB2 OX=1196094 GN=SALWKB2_2213 PE=4 SV=1 28088.0 54936 >tr|X2H5V4|X2H5V4_9NEIS Transporter, EamA family OS=Snodgrassella alvi wkB2 OX=1196094 GN=SALWKB2_1382 PE=4 SV=1 27916.0 165037 >tr|X2HD89|X2HD89_9NEIS Putative ABC transporter, periplasmic component YrbD OS=Snodgrassella alvi wkB2 OX=1196094 GN=SALWKB2_1655 PE=4 SV=1 26868.0 47738 >tr|X2H3H5|X2H3H5_9NEIS Putative membrane-associated, metal-dependent hydrolase OS=Snodgrassella alvi wkB2 OX=1196094 GN=SALWKB2_0567 PE=4 SV=126268.0 29039 >tr|X2HA68|X2HA68_9NEIS Uncharacterized protein OS=Snodgrassella alvi wkB2 OX=1196094 GN=SALWKB2_1888 PE=4 SV=1 25936.0 200640 >tr|X2HBY5|X2HBY5_9NEIS 2-keto-3-deoxy-D-arabino-heptulosonate-7-phosphate synthase I beta OS=Snodgrassella alvi wkB2 OX=1196094 GN=SALWKB2_1185 PE=4 SV=124900.0 125841 >tr|X2HET7|X2HET7_9NEIS Polysaccharide biosynthesis protein CpsM OS=Snodgrassella alvi wkB2 OX=1196094 GN=SALWKB2_2240 PE=4 SV=1 22934.0 223842 >tr|X2H634|X2H634_9NEIS RNA polymerase sigma factor RpoD OS=Snodgrassella alvi wkB2 OX=1196094 GN=rpoD PE=3 SV=1 22824.0 174343 >tr|X2HBR9|X2HBR9_9NEIS Uncharacterized protein OS=Snodgrassella alvi wkB2 OX=1196094 GN=SALWKB2_1236 PE=4 SV=1 22510.0 25544 >tr|X2H8K5|X2H8K5_9NEIS Protein of avirulence locus ImpE OS=Snodgrassella alvi wkB2 OX=1196094 GN=SALWKB2_0101 PE=4 SV=1 22212.0 110445 >tr|X2HA02|X2HA02_9NEIS D-amino acid dehydrogenase small subunit OS=Snodgrassella alvi wkB2 OX=1196094 GN=SALWKB2_1798 PE=4 SV=1 18936.0 52746 >tr|X2HEB5|X2HEB5_9NEIS 4-hydroxy-3-methylbut-2-enyl diphosphate reductase OS=Snodgrassella alvi wkB2 OX=1196094 GN=ispH PE=3 SV=1 18396.0 46347 >tr|X2HA77|X2HA77_9NEIS Molybdenum ABC transporter, periplasmic molybdenum-binding protein ModA OS=Snodgrassella alvi wkB2 OX=1196094 GN=SALWKB2_0615 PE=4 SV=11824 . 149448 >tr|X2HCQ4|X2HCQ4_9NEIS CTP synthase OS=Snodgrassella alvi wkB2 OX=1196094 GN=pyrG PE=3 SV=1 17566.0 949 >tr|X2HBM2|X2HBM2_9NEIS GMP synthase [glutamine-hydrolyzing] OS=Snodgrassella alvi wkB2 OX=1196094 GN=guaA PE=3 SV=1 14824.0 138150 >tr|X2HF25|X2HF25_9NEIS ATP-dependent DNA helicase RecG OS=Snodgrassella alvi wkB2 OX=1196094 GN=recG PE=3 SV=1 14564.0 43575 B) Whole cell fraction Figure 3.4 Example of the protein lists obtained for each fraction A) Secreted fraction protein list. The proteins present in the secreted fraction ranked for their total intensity B) Whole fraction cell list. The proteins present in the whole cell fraction ranked for their total intensity. Rank for intensity DescriptionTotalIntensityProteinDB number1 X2HI07_9NEIS Outer membrane protein class 1 OS=Snodgrassella alvi wkB2 OX=1196094 GN=SALWKB2_1759 PE=4 SV=1 8084960.0 5352 X2H718_9NEIS Outer membrane protein A OS=Snodgrassella alvi wkB2 OX=1196094 GN=SALWKB2_1777 PE=4 SV=1 5597212.0 13283 X2H3G6_9NEIS Uncharacterized protein OS=Snodgrassella alvi wkB2 OX=1196094 GN=SALWKB2_0557 PE=4 SV=1 5412524.0 9054 X2H9D9_9NEIS Chaperone protein DnaK OS=Snodgrassella alvi wkB2 OX=1196094 GN=dnaK PE=2 SV=1 5208962.0 7165 X2H826_9NEIS DNA-binding protein HU-beta OS=Snodgrassella alvi wkB2 OX=1196094 GN=SALWKB2_1158 PE=3 SV=1 3887862.0 4976 X2H979_9NEIS 60 kDa chaperonin OS=Snodgrassella alvi wkB2 OX=1196094 GN=groL PE=3 SV=1 3813850.0 15337 X2H853_9NEIS Aminotransferase OS=Snodgrassella alvi wkB2 OX=1196094 GN=SALWKB2_1178 PE=3 SV=1 3554796.0 768 X2H8F7_9NEIS YadA-like protein OS=Snodgrassella alvi wkB2 OX=1196094 GN=SALWKB2_2297 PE=4 SV=1 3134494.0 16359 X2H8Z4_9NEIS Autotransporter domain-containing protein OS=Snodgrassella alvi wkB2 OX=1196094 GN=SALWKB2_1478 PE=4 SV=1 3016492.0 204910 X2HER9_9NEIS Putative oxidoreductase OS=Snodgrassella alvi wkB2 OX=1196094 GN=SALWKB2_2230 PE=4 SV=1 2666952.0 57611 X2HAX1_9NEIS Uncharacterized protein OS=Snodgrassella alvi wkB2 OX=1196094 GN=SALWKB2_0845 PE=4 SV=1 2597534.0 191412 X2HDM6_9NEIS 50S ribosomal protein L7/L12 OS=Snodgrassella alvi wkB2 OX=1196094 GN=rplL PE=3 SV=1 2505900.0 94413 X2HFC5_9NEIS Transcriptional regulator OS=Snodgrassella alvi wkB2 OX=1196094 GN=SALWKB2_0884 PE=4 SV=1 2232250.0 97914 X2HCR4_9NEIS YadA-like protein OS=Snodgrassella alvi wkB2 OX=1196094 GN=SALWKB2_0009 PE=4 SV=1 2172956.0 227215 X2H5L9_9NEIS Uncharacterized protein OS=Snodgrassella alvi wkB2 OX=1196094 GN=SALWKB2_0343 PE=4 SV=1 1886684.0 109916 X2HGA7_9NEIS RNA-binding protein Hfq OS=Snodgrassella alvi wkB2 OX=1196094 GN=hfq PE=3 SV=1 1639686.0 61117 X2HAD7_9NEIS Succinate-semialdehyde dehydrogenase [NAD] Succinate-semialdehyde dehydrogenase [NADP+] OS=Snodgrassella alvi wkB2 OX=1196094 GN=SALWKB2_1973 PE=4 SV=11570406. 157918 X2H475_9NEIS Pyruvate dehydrogenase E1 component OS=Snodgrassella alvi wkB2 OX=1196094 GN=SALWKB2_0787 PE=4 SV=1 1528898.0 173419 X2H8D1_9NEIS Thioredoxin OS=Snodgrassella alvi wkB2 OX=1196094 GN=SALWKB2_1233 PE=3 SV=1 1377632.0 32620 X2HB30_9NEIS HTH cro/C1-type domain-containing protein OS=Snodgrassella alvi wkB2 OX=1196094 GN=SALWKB2_1016 PE=4 SV=1 1359182.0 73221 X2HDR1_9NEIS Acyl carrier protein OS=Snodgrassella alvi wkB2 OX=1196094 GN=acpP PE=3 SV=1 1346306.0 9922 X2H3K0_9NEIS Uncharacterized protein OS=Snodgrassella alvi wkB2 OX=1196094 GN=SALWKB2_0597 PE=4 SV=1 1156194.0 70023 X2H756_9NEIS PpiC domain-containing protein OS=Snodgrassella alvi wkB2 OX=1196094 GN=SALWKB2_1827 PE=4 SV=1 1155926.0 4324 X2H5I1_9NEIS Chaperone protein ClpB OS=Snodgrassella alvi wkB2 OX=1196094 GN=clpB PE=3 SV=1 1106902.0 63525 X2H8Y3_9NEIS Type IV pilus biogenesis protein PilQ OS=Snodgrassella alvi wkB2 OX=1196094 GN=SALWKB2_1463 PE=3 SV=1 1069762.0 211626 X2HCE7_9NEIS Alkyl hydroperoxide reductase C OS=Snodgrassella alvi wkB2 OX=1196094 GN=SALWKB2_1481 PE=3 SV=1 1053092.0 85527 X2H6Y5_9NEIS Acetyltransferase component of pyruvate dehydrogenase complex OS=Snodgrassella alvi wkB2 OX=1196094 GN=SALWKB2_0788 PE=3 SV=11046692.0 84528 X2H998_9NEIS Elongation factor Tu OS=Snodgrassella alvi wkB2 OX=1196094 GN=tuf PE=3 SV=1 1036608.0 2929 X2HCJ2_9NEIS Uncharacterized protein OS=Snodgrassella alvi wkB2 OX=1196094 GN=SALWKB2_1390 PE=4 SV=1 1015258.0 208630 X2HHB7_9NEIS 10 kDa chaperonin OS=Snodgrassella alvi wkB2 OX=1196094 GN=groS PE=3 SV=1 984924.0 219731 X2H2R7_9NEIS 50S ribosomal protein L1 OS=Snodgrassella alvi wkB2 OX=1196094 GN=rplA PE=3 SV=1 931850.0 183932 X2HGU1_9NEIS Uncharacterized protein OS=Snodgrassella alvi wkB2 OX=1196094 GN=SALWKB2_1389 PE=4 SV=1 920358.0 43633 X2H4P5_9NEIS HTH OST-type domain-containing protein OS=Snodgrassella alvi wkB2 OX=1196094 GN=SALWKB2_0033 PE=4 SV=1 917124.0 19734 X2HD87_9NEIS LysM domain-containing protein OS=Snodgrassella alvi wkB2 OX=1196094 GN=SALWKB2_0149 PE=4 SV=1 801394.0 202435 X2H921_9NEIS 50S ribosomal protein L11 OS=Snodgrassella alvi wkB2 OX=1196094 GN=rplK PE=3 SV=1 643390.0 6936 X2HC30_9NEIS Uncharacterized protein ImpB OS=Snodgrassella alvi wkB2 OX=1196094 GN=SALWKB2_1366 PE=4 SV=1 633972.0 182837 X2H993_9NEIS DNA-directed RNA polymerase subunit beta OS=Snodgrassella alvi wkB2 OX=1196094 GN=rpoB PE=3 SV=1 600654.0 164638 X2H8B5_9NEIS Nucleoside diphosphate kinase OS=Snodgrassella alvi wkB2 OX=1196094 GN=ndk PE=3 SV=1 599364.0 127739 X2H5L2_9NEIS Peptidyl-prolyl cis-trans isomerase OS=Snodgrassella alvi wkB2 OX=1196094 GN=SALWKB2_1272 PE=3 SV=1 587664.0 55540 X2H6B6_9NEIS LemA family protein OS=Snodgrassella alvi wkB2 OX=1196094 GN=SALWKB2_0608 PE=4 SV=1 576308.0 35341 X2H7Y9_9NEIS 30S ribosomal protein S5 OS=Snodgrassella alvi wkB2 OX=1196094 GN=rpsE PE=3 SV=1 524166.0 228742 X2HE62_9NEIS YadA-like protein OS=Snodgrassella alvi wkB2 OX=1196094 GN=SALWKB2_2030 PE=4 SV=1 517970.0 43143 X2H4T6_9NEIS Phosphoenolpyruvate synthase OS=Snodgrassella alvi wkB2 OX=1196094 GN=SALWKB2_1047 PE=3 SV=1 514418.0 142844 X2H4K1_9NEIS Phage tail sheath monomer OS=Snodgrassella alvi wkB2 OX=1196094 GN=SALWKB2_0947 PE=4 SV=1 495282.0 67045 X2HB82_9NEIS Superoxide dismutase OS=Snodgrassella alvi wkB2 OX=1196094 GN=SALWKB2_1076 PE=3 SV=1 485416.0 60646 X2HEI6_9NEIS 50S ribosomal protein L15 OS=Snodgrassella alvi wkB2 OX=1196094 GN=rplO PE=3 SV=1 461692.0 100547 X2H621_9NEIS Type IV pilus biogenesis protein PilP OS=Snodgrassella alvi wkB2 OX=1196094 GN=SALWKB2_1462 PE=4 SV=1 457478.0 133648 X2HAE6_9NEIS OmpA-like domain-containing protein OS=Snodgrassella alvi wkB2 OX=1196094 GN=SALWKB2_1983 PE=4 SV=1 447700.0 14849 X2H4C7_9NEIS Putative transmembrane protein OS=Snodgrassella alvi wkB2 OX=1196094 GN=SALWKB2_0852 PE=4 SV=1 444380.0 154350 X2H2I3_9NEIS Protein GrpE OS=Snodgrassella alvi wkB2 OX=1196094 GN=grpE PE=3 SV=1 437886.0 53651 X2HAL5_9NEIS 50S ribosomal protein L9 OS=Snodgrassella alvi wkB2 OX=1196094 GN=rplI PE=3 SV=1 432096.0 203852 X2H6V9_9NEIS NADP-dependent malic enzyme OS=Snodgrassella alvi wkB2 OX=1196094 GN=SALWKB2_0748 PE=4 SV=1 422206.0 105153 X2HDN0_9NEIS Elongation factor G OS=Snodgrassella alvi wkB2 OX=1196094 GN=fusA PE=3 SV=1 420066.0 216554 X2HDQ1_9NEIS Methionine ABC transporter substrate-binding protein OS=Snodgrassella alvi wkB2 OX=1196094 GN=SALWKB2_0284 PE=4 SV=14 6394.0 23455 X2HEH8_9NEIS 50S ribosomal protein L2 OS=Snodgrassella alvi wkB2 OX=1196094 GN=rplB PE=3 SV=1 414996.0 91056 X2HA54_9NEIS Putative periplasmic protein OS=Snodgrassella alvi wkB2 OX=1196094 GN=SALWKB2_0656 PE=4 SV=1 394850.0 73657 X2H2S2_9NEIS 30S ribosomal protein S12 OS=Snodgrassella alvi wkB2 OX=1196094 GN=rpsL PE=3 SV=1 391618.0 228958 X2H9K0_9NEIS Uncharacterized protein OS=Snodgrassella alvi wkB2 OX=1196094 GN=SALWKB2_0385 PE=4 SV=1 384198.0 190359 X2HBP6_9NEIS Trigger factor OS=Snodgrassella alvi wkB2 OX=1196094 GN=tig PE=3 SV=1 359724.0 214060 X2HCV1_9NEIS 30S ribosomal protein S2 OS=Snodgrassella alvi wkB2 OX=1196094 GN=rpsB PE=3 SV=1 338452.0 70661 X2HDP2_9NEIS Protein RecA OS=Snodgrassella alvi wkB2 OX=1196094 GN=recA PE=3 SV=1 334916.0 115262 X2HBQ7_9NEIS 30S ribosomal protein S1 OS=Snodgrassella alvi wkB2 OX=1196094 GN=SALWKB2_1226 PE=3 SV=1 317248.0 104563 X2HDG5_9NEIS Peptidylprolyl isomerase OS=Snodgrassella alvi wkB2 OX=1196094 GN=SALWKB2_1826 PE=4 SV=1 315496.0 155664 X2H6X9_9NEIS Glutamine synthetase OS=Snodgrassella alvi wkB2 OX=1196094 GN=SALWKB2_0778 PE=3 SV=1 306386.0 109365 X2HIY2_9NEIS DNA-directed RNA polymerase subunit alpha OS=Snodgrassella alvi wkB2 OX=1196094 GN=rpoA PE=3 SV=1 292932.0 215466 X2HFC8_9NEIS Glutamine ABC transporter, periplasmic glutamine-binding protein OS=Snodgrassella alvi wkB2 OX=1196094 GN=SALWKB2_0889 PE=3 SV=1292810.0 205667 X2HCC2_9NEIS Type IV pilus biogenesis protein PilO OS=Snodgrassella alvi wkB2 OX=1196094 GN=SALWKB2_1461 PE=4 SV=1 282008.0 57568 X2H685_9NEIS Type IV pilus biogenesis protein PilE OS=Snodgrassella alvi wkB2 OX=1196094 GN=SALWKB2_1542 PE=3 SV=1 261560.0 31669 X2H984_9NEIS 30S ribosomal protein S16 OS=Snodgrassella alvi wkB2 OX=1196094 GN=rpsP PE=3 SV=1 252970.0 85170 X2H752_9NEIS Neisseria-specific antigen protein, TspA OS=Snodgrassella alvi wkB2 OX=1196094 GN=SALWKB2_1822 PE=4 SV=1 242592.0 103671 X2H7Z7_9NEIS 50S ribosomal protein L5 OS=Snodgrassella alvi wkB2 OX=1196094 GN=rplE PE=3 SV=1 241824.0 31272 X2H6Q2_9NEIS Chaperone protein HtpG OS=Snodgrassella alvi wkB2 OX=1196094 GN=htpG PE=3 SV=1 241408.0 11173 X2HHU6_9NEIS Cell processes adaptation osmotic adaptation OS=Snodgrassella alvi wkB2 OX=1196094 GN=SALWKB2_1714 PE=4 SV=1 230850.0 152074 X2H996_9NEIS Transmembrane protein OS=Snodgrassella alvi wkB2 OX=1196094 GN=SALWKB2_1593 PE=4 SV=1 225394.0 78575 X2HB50_9NEIS Lipoprotein OS=Snodgrassella alvi wkB2 OX=1196094 GN=SALWKB2_1036 PE=3 SV=1 216136.0 218676 X2HAY9_9NEIS 50S ribosomal protein L24 OS=Snodgrassella alvi wkB2 OX=1196094 GN=rplX PE=3 SV=1 194834.0 118377 X2H803_9NEIS 50S ribosomal protein L16 OS=Snodgrassella alvi wkB2 OX=1196094 GN=rplP PE=3 SV=1 192894.0 66878 X2H3K8_9NEIS Uncharacterized protein OS=Snodgrassella alvi wkB2 OX=1196094 GN=SALWKB2_0607 PE=4 SV=1 180242.0 201479 X2H927_9NEIS 30S ribosomal protein S10 OS=Snodgrassella alvi wkB2 OX=1196094 GN=rpsJ PE=3 SV=1 174940.0 228880 X2H2V8_9NEIS L-lactate dehydrogenase OS=Snodgrassella alvi wkB2 OX=1196094 GN=SALWKB2_0307 PE=3 SV=1 171716.0 95581 X2HAV8_9NEIS Uncharacterized protein OS=Snodgrassella alvi wkB2 OX=1196094 GN=SALWKB2_0830 PE=4 SV=1 161276.0 174982 X2H866_9NEIS Ketol-acid reductoisomerase (NADP(+)) OS=Snodgrassella alvi wkB2 OX=1196094 GN=ilvC PE=3 SV=1 154524.0 1383 X2HB75_9NEIS ABM domain-containing protein OS=Snodgrassella alvi wkB2 OX=1196094 GN=SALWKB2_1066 PE=4 SV=1 128306.0 104684 X2H617_9NEIS Polyphosphate kinase 2 OS=Snodgrassella alvi wkB2 OX=1196094 GN=SALWKB2_1457 PE=4 SV=1 126444.0 96485 X2HEH1_9NEIS 50S ribosomal protein L29 OS=Snodgrassella alvi wkB2 OX=1196094 GN=rpmC PE=3 SV=1 122690.0 75386 X2HAP0_9NEIS ABC-type phosphate/phosphonate transport system periplasmic component OS=Snodgrassella alvi wkB2 OX=1196094 GN=SALWKB2_0856 PE=4 SV=1119496.0 42487 X2HF86_9NEIS S-ribosylhomocysteine lyase OS=Snodgrassella alvi wkB2 OX=1196094 GN=luxS PE=3 SV=1 113370.0 176988 X2H682_9NEIS HMA domain-containing protein OS=Snodgrassella alvi wkB2 OX=1196094 GN=SALWKB2_1537 PE=4 SV=1 109504.0 115489 X2HEK3_9NEIS Putative ATP/GTP-binding protein OS=Snodgrassella alvi wkB2 OX=1196094 GN=SALWKB2_2186 PE=4 SV=1 97124.0 141390 X2H4C5_9NEIS Endoribonuclease L-PSP OS=Snodgrassella alvi wkB2 OX=1196094 GN=SALWKB2_0847 PE=4 SV=1 94380.0 219891 X2HAX0_9NEIS 50S ribosomal protein L17 OS=Snodgrassella alvi wkB2 OX=1196094 GN=rplQ PE=3 SV=1 90336.0 148992 X2H2M8_9NEIS Twitching motility protein PilT OS=Snodgrassella alvi wkB2 OX=1196094 GN=SALWKB2_0232 PE=4 SV=1 89784.0 148893 X2HBY6_9NEIS Type IV fimbrial biogenesis protein PilY1 OS=Snodgrassella alvi wkB2 OX=1196094 GN=SALWKB2_1306 PE=4 SV=1 86628.0 128794 X2H7N6_9NEIS Uncharacterized protein OS=Snodgrassella alvi wkB2 OX=1196094 GN=SALWKB2_1058 PE=4 SV=1 57168.0 187195 X2HEI0_9NEIS 30S ribosomal protein S4 OS=Snodgrassella alvi wkB2 OX=1196094 GN=rpsD PE=3 SV=1 55830.0 13296 X2H4D1_9NEIS Aerotaxis sensor receptor protein OS=Snodgrassella alvi wkB2 OX=1196094 GN=SALWKB2_0857 PE=4 SV=1 53898.0 78497 X2HFS3_9NEIS Peptidyl-prolyl cis-trans isomerase OS=Snodgrassella alvi wkB2 OX=1196094 GN=SALWKB2_1054 PE=3 SV=1 52322.0 194498 X2H9M1_9NEIS NADH:flavin oxidoreductase OS=Snodgrassella alvi wkB2 OX=1196094 GN=SALWKB2_1683 PE=4 SV=1 48020.0 48299 X2HFH7_9NEIS Uncharacterized protein OS=Snodgrassella alvi wkB2 OX=1196094 GN=SALWKB2_0949 PE=4 SV=1 47732.0 1741100 X2HAN1_9NEIS Nitrogen regulatory protein P-II OS=Snodgrassella alvi wkB2 OX=1196094 GN=SALWKB2_0841 PE=3 SV=1 45850.0 699101 X2HC81_9NEIS Thioredoxin reductase OS=Snodgrassella alvi wkB2 OX=1196094 GN=SALWKB2_1250 PE=3 SV=1 39160.0 129076 Figure 3.5 SignalP 5.0 prediction. Y-axis shows the probability of the sequence to be a signal peptide and x-axis shows the amino acids sequence that are part of that putative sequence. A) 2-oxoglutarate dehydrogenase E1 component. B) Putative membrane-associated, metal-dependent hydrolase. Each amino acid sequence shows the likelihood to belong to the corresponding motif of the signal peptide. Thus we move to analyze each replicate individually (pair of protein lists). Surprisingly there were only 5 to 6 proteins containing signal peptides (5 for replicate 1; 6 for replicate 2; and 5 for replicate 3) from a total of 50 to 57 identified proteins (50 for replicate 1 and 2, and 57 for replicate 3). Also, and as mentioned earlier, all these proteins containing signal peptide were 77 different between each replicate. The Sec-dependent type I signal peptide carrying proteins found are: Outer-membrane lipoprotein carrier protein, D-amino acid dehydrogenase, CTP synthase, Cyclohexadienyl dehydrogenase, type IV fimbrial biogenesis protein PilT1, blue copper oxidase CueO, autotransporter domain-containing protein, TonB-dependent receptor and outer membrane protein assembly factor BamA. The rest of the proteins were uncharacterized proteins. Regarding to whole cell fraction protein lists, none of them contains any of these proteins carrying signal peptide. This last can suggest that these proteins are mostly secreted and their in-cell presence is low and no detectable for our limits. A first guess to explain the low number of signal peptide carrying proteins in the secretion fraction could be linked to cell lysis during sample preparation, however that would also suggests that those proteins are highly present in the whole cell fraction, or at least detectable. Nevertheless the proteins present in the whole cell fraction and the secreted fraction are substantially different and don\u00E2\u0080\u0099t even overlap. A second explanation to our results could be the presence of a high number of non-classical secreted proteins, however, most of the bacterial systems show secretion patters governed by signal peptide-dependent108. From a more technical approach and regarding lack or reproducibility, our experimental set up includes harvesting cells from agar plates rather than liquid culture. Samples are likely contaminated with considerable amount of bovine proteins that limit the yield of actual S. alvi secreted proteins recovery. At this point, S. alvi is only known to grow in enriched liquid broths which contain proteins from other sources, making an analysis of its extracellular proteins in liquid culture currently impossible. If we or others could identify a minimal liquid media that 78 supports grown then this experiment could be repeated in much the same way that it was done here. The inconsistency observed in the replicates for the secreted fraction suggests that the secretion of proteins on solid media is unreliable, at best. Although surface attached peptides can probably be studied by complementing this approach with cell washes followed of cell trypsin shaving117. Additional concerns are related to the solubilization efficiency we obtained, the lack of a chaotrope agent in our methods is likely affecting the peptide downstreaming. Membrane attached and periplasm proteins could likely being poorly recovered due to their hydrophobic cell compartmentalization. The combined used of urea and thiourea have been used successfully for the preparation of membrane and integral membrane proteins118. Also to improve the reliability of our approach (concentration equals total intensity), iBAQ algorithm, in where the total protein intensity is divided by the number of tryptic peptides, or the top three method, in where the three peptides with highest intensity are used for quantitation should be employed for standardize the scale of the intensity119. This last being particular useful and decisive with larger protein lists that we were expecting in the whole cell fraction. SignalP 5.0 prediction algorithm has proven high accuracy for the identification of Gram-negative signal peptides (0.938 correlation coefficient)116. In general, SignalP 5.0 prediction algorithms commonly match the in vivo validated signal peptides but overestimate their presence; such is the case of E. coli that with a total of 137 signal peptides experimentally validated, 136 were predicted using SingalP 5.0; however, Uniprot reports 498 proteins with a signal peptide, while SignalP 5.0 detects 612, suggesting an overestimation in most of the cases, highlighting the importance of pairing in vivo and in silico approaches116,120. For our experiment there are some issues to address before considering this data as reference for S.alvi, particularly 79 because of the lack of reproducibility between replicates. Although, If we compare the number of experimentally validated S. alvi vs SignalP 5.0 predcited ratio against E.coli, we have that 10% of the signal peptide predicted for S.alvi have been validated through this experiment while for E. coli is 22% along different experimental set ups116. 3.4.3 Plasmid design Casettes of expression At this point, we then wished to create expression cassettes that could be tested in S. alvi for their ability to express and secrete active AMPs. As mentioned in Ch. 2, we decided to focus on jelleine and melitin. We used the following melittin DNA sequence \u00E2\u0080\u0098ggcattggcgcggtgctgaaagtgctgaccaccggcctg\u00E2\u0080\u0099, and for jelleine it was \u00E2\u0080\u0098ccgtttaaactgagcctgcatctg\u00E2\u0080\u0099. The promoter used was that of cp25 (ctttggcagtttattcttgacatgtagtgagggggctggtataatcacatagtactgtt) from igem. The terminator used was pBP-BBa_B0015 (ccaggcatcaaataaaacgaaaggctcagtcgaaagactgggcctttcgttttatctgttgtttgtcggtgaacgctctctactagagtcacactggctcaccttcgggtgggcctttctgcgtttata), also from igem (https://igem.org/Main_Page). For cloning purposes, the sequence was flanked with the restriction enzymes EcoRI (gaattc) PciI (acatgt). Since we were unable to validate the in vivo activity of any of the signal peptides recovered through SignalP5, we decided to propose a preliminary signal peptide candidate based on the highest Sec/SPI likelihood score previously mentioned (Figure 3.6), which was the signal peptide of carbonic anhydrase (X2H8Y7), whose sequence is \u00E2\u0080\u0098atgcgcaaaaacattctggcgctgtgcctgagcctgggcctggcgaccagcgcgatggcggaa\u00E2\u0080\u0099. These elements were assembled into two different cassettes, ready for synthesis and insertion into a plasmid such as pBTK503 from the bee toolkit32. 80 Figure 3.6 SignalP 5.0 prediction for Carbonic anhydrase (Top scored signal peptide type I containing protein). Each amino acid sequence shows the likelihood to belong to the corresponding motif of the signal peptide. 3.5 Conclusion For this chapter we attempted to identify Sec-dependent signal peptide candidates that could drive secretion of heterologous AMPs in S. alvi, and then validate its activity with an expression vector coding for the signal peptide and AMPs of interest. Although designing the heterologous secretion from a Gram-negative bacteria could be challenging, it is known that generally Gram-negative secrete fewer native proteins, suggesting that the heterologous protein could be favored. In fact, S. alvi is known to have a strong metabolic cooperation relationship with G. apicola limiting the number of secreted CAZy enzymes present121. Furthermore, S. alvi has shown to be 81 a highly ubiquitous and resilient microorganism, robust enough to be a vector of heterologous AMPs. The rationale behind our secretion strategy is that the most active signal peptides, in terms of secretion activity for the host organism (S. alvi), seem most likely to be the best signal peptides for driving heterologous secretion. Our experimental procedure was designed to identify the proteins that are more secreted by identifying the total intensity difference between the whole cell and the secreted fraction for each protein. However, the final steps in being able to validate and test these predictions ended up being delayed and then, finally, cancelled due to the COVID-19 shutdown of 2020. We were unable to troubleshoot the set-up of the experiment in order to increase the resolution of the proteins in the samples, which likely will include strategies for cleaner samples or minimal media design for S. alvi. We had hoped that this would increase the reproducibility of secreted proteins identified between replicates, as well as to help us identify less abundant proteins, which could still have strong secretion activity. At the moment, we have a total of 218 Sec-dependent type I signal peptides available in S. alvi for the validation of their activity, however, choosing the candidates for screening would have to be based on in silico criteria such as the scores of likelihood produced by SignalP 5.0. Further testing of the expression cassettes suggested in here will also have to follow. 82 Chapter 4: Conclusion The honey bee microbiome field has been developing at a rapid pace ever since the early 2000s 12. Both culture and culture-independent methods of microbial study have served to build a robust understanding of the role that the different taxa play in the gut environment. The contribution of the bacterial components to the bee health, derived from a profound host-microbe co-dependence, is highly remarkable. Through the characterization of metabolic contributions, the relationship between gut bacterial species and metabolites \u00E2\u0080\u0093 both consumed and synthesized \u00E2\u0080\u0093 has been uncovered. However, some other kinds of contributions, such as deep immune related mechanisms, are still to be better understood. Throughout the past decade, there has been a great deal of advancement in the development of molecular tools for monitoring the response of the microbial community and understanding each species individually. For the former of the two, metagenomics has served as powerful approach for surveying the fungal and bacterial composition of the microbiota overall, while culture methods have allowed to understand more deeply individual components of the honey bee microbiota. Nowadays is possible to cultivate all of the eight bacterial core taxa. The development of a molecular toolbox for genetically engineering the bacterial components has also been a milestone for the field. Through this project we wanted to apply these available tools for the development of a therapeutic probiotic, consisting of the commensal bacterium S. alvi for the expression and secretion of naturally occurring antimicrobial peptides (AMPs) to fight against P. larvae infection. We step on decades of work of other research groups; however, we also contributed with unique features to the field. We explored the feasibility of using naturally occurring AMPs to be expressed in our prophylactic probiotic, trying to ask a broad group of 83 questions such as what is the level of specificity of the adult responses to bacteria within different ecological niches? How do commensal and non-commensal bacteria respond to the same group of antimicrobial peptides? What adult honey bee antimicrobial peptides have direct activity against P. larvae? In addition, we also pioneered on the characterization of the first proteomic experiment carried on S. alvi for a better understanding of its secretion dynamics. 4.1 Addressing the project aims Through two chapters we set two general goals: first the identification of AMPs to be expressed in the prophylactic probiotic, and second, the design of a strategy to genetically engineer S. alvi to express and secrete heterologous peptides (the honey bee AMPs). For the former goal, the aims included the isolation and culture of the characteristic honey bee CGB from adult honey bees; uncovering the specificity of immune factors against P. larvae; identification antimicrobial peptides that inhibits P. larvae vegetative cells growth, but minimally affects the commensal gut bacteria, as well as their active concentration; and to evaluate the response of the core microbiota (in vivo) to the candidate AMPs. The idea behind isolating as many CGB members as possible was to survey the individual responses to the AMPs, and thus to have a clearer idea of the resolution of the modulation that these immune factors have on the commensal bacteria. The isolation of these unique bacteria wasn\u00E2\u0080\u0099t the only challenge, since even if the methods are published for isolation by several authors, inhibition assays require a more intense culture strategies (using minimal media and to reach high cell densities). Unfortunately, it wasn\u00E2\u0080\u0099t possible to work with all CGB members for the susceptibility assays. We isolated five CGB : G. apicola, Lactobacillus Kunkeii, Bifidobacterium asteroides, Bartonella apis and S. alvi but ended 84 up working with the last three. Carrying susceptibility experiments in most of CGB species will also aid in the understanding of the microbiome reshape role that the host AMPs have. Because of unpreceded conditions (two months of UBC shutdown due to COVID-19) all the samples of the host-microbe experiment weren\u00E2\u0080\u0099t run on time for this work (LC-MS/MS on hold). However we expect our aim will be eventually achieved since it complements the understanding on the responses triggered by specific microorganisms to the host immune system. The identification of the AMPs candidates to be expressed in S. alvi was successful, meaning that we were able to identify relatively specific AMPs as well as MICs. More interestingly we observed some slight patterns on the selectivity of naturally occurring AMPs against commensal and non-commensal bacteria. Finally, the overall effect on the microbiota was depicted, although we would like to repeat the experiment with a positive control of OTC at the concentration at which dysbiosis was reported by previous authors, since we couldn\u00E2\u0080\u0099t conclude with a single explanation for what we observed: overall increase of bacterial cells for all the treatments and positive control. For the second goal the first aim was to identify the molecular machinery present in S. alvi for the delivery and secretion of the chosen AMPs. Then we aimed to engineer S. alvi cells to later validate that the therapeutic probiotic colonizes larvae gut and that it is biologically active (active AMPs). For the identification of signal peptides we partially validated our strategy, which consisted on the identification of the whole cell proteins fractions versus the secreted fraction to determine ratios of secretion for the homologous signal peptides. Despite of the lack of reference proteomics experiments for S. alvi we developed an interesting approach from solid culture colonies. Both fractions (whole cell and secretions) were analyzed and shown clear differences, 85 although more of resolution was needed to identify Sec-dependent secreted proteins. In summary, for this first aim we couldn\u00E2\u0080\u0099t understand the Sec dependent secreted protein dynamics through our first trial, and even if we repeated the trials with technical improvements, we were not able to run them on time for this work (due to the unprecedented shutdown). Finally, engineering S. alvi was not possible in the timeframe for the previously mentioned event. However we showed in this work the likely end assembly of the expression cassette we will employ for this aim. 4.2 Addressing the project hypothesis We started this study framework with three hypotheses to test: first, we hypothesized that by providing an engineered probiotic composed of a commensal bacterium able to express and secrete adult honey bee AMPs, larvae would resist the P. larvae infection; second, we hypothesized that the historical coexistence of the intrinsic immune factors of the honey bee and its commensal bacteria leads to highly adapted commensals, which suggests that most of them are resistant to the intrinsic AMPs of honey bees, individually and as a community; third we hypothesized that heterologous secretion of recombinant peptides in the commensal bacterium S. alvi can be conducted through the attachment of homologous signal peptides present in its secretome to the heterologous target AMP gene, and that also those signal peptides that show higher secretion activity for their native targets have higher chances to work with the heterologous target. The first hypothesis required the ending of all the experiments scheduled for this work, meaning genetic engineering of S. alvi, followed of inoculation experiments and P. larvae infection to S.alvi inoculated and non-inoculated larvae. The second hypothesis was successfully addressed and validated: Through relatively simple experiments (inhibition assays) 86 accompanied by the metagenomic approach, we validated the resistance that most commensal bacteria have against the overall repertoire of AMPs and also the inhibitory activity of the same AMPs against P. larvae and non-commensal bacteria. The third hypothesis couldn\u00E2\u0080\u0099t be answered without the aid of the transformation of S. alvi, since we needed to test a variety of expression cassettes representing a different repertoire of homologues signal peptides, including those identified in the secretion ratio experiment (ideally) and few others heterologous signal peptides as reference. 4.3 Future directions 4.3.1 Efforts for the therapeutic probiotic completion Even though all the experiments needed for the completion of this goal were planned, the circumstances surrounding COVID-19 didn\u00E2\u0080\u0099t allow us to finish all of them. Therefore, a short-term future direction is to follow up on the missing experiments for the completion of our goal. The minimal experiments required include: first, screening of a collection of homologous signal peptides as part of the expression cassette for S. alvi, which also involves to determine the yield of secretion of each signal peptide candidate; second, and once expression and secretion of either jelleine or melittin in S. alvi is achieved, the probiotic has to be tested in isolated larvae under laboratory conditions to determine the colonization patters and any potential side effects; third, the validation of the probiotic would need to test the prophylactic activity as part as a pre-treatment to a P. larvae infection experiment, to later compare efficacy with naked AMPs and OTC. 87 4.3.2 Uncovering the mechanisms of action for the chosen antimicrobial peptides through in vitro assays. Characterization of jelleine\u00E2\u0080\u0099s and melittin\u00E2\u0080\u0099s mode of action through biophysical techniques in vitro will offer a deep close up look into the implications of these AMPs, not only in P. larvae, but in the whole microbial community. Membrane disruption, membrane interaction, and DNA or RNA interaction assays would give a better insight of any risk in the open environment usage. Furthermore, this will aid to understand more about the susceptibly specificity observed. 4.3.3 AMPs efficacy against non-commensal bacteria: is this the reason why honey bee gut environment is very selective? It would be very revealing to validate the efficacy of the honey bee AMPs against more non-commensal bacteria species, such as those present in the floral environment or in the microbiota of related insects. It is known that environmental bacteria is very rarely able to colonize the bee gut, therefore validating the selectivity of the AMPs would be a very interesting follow up. 4.3.4 Differential protein expression profiles between AFB immune Asian honey bee and susceptible Western honey bee larvae The remarkably specific vulnerability of only first and second instar larvae of the Western honey bee to P. larvae infection is very intriguing. Very closely related species such as the Asian honey bees are not vulnerable. (at any developmental stage). An interesting set of experiments could involve the comparison of the immune-related protein expression profiles of the Asian honey bee versus the Western honey bee. 88 4.3.5 Microbiota tolerance to host AMPs validation in other insect models An interesting follow up could be to survey the level of specificity of a variety of AMPs by performing similar assays with the corresponding host-microbiota pairing. For example, to test the same fruit fly AMPs (or more) against fruit fly gut commensals. However, we considered high throughput methods for this same goal could enlarge the scope of understanding and the sensibility of the results. 4.3.6 Chromosomal integration of our expression cassettes To integrate the AMPs secretion expression cassette into S. alvi chromosomal DNA for creating a stable strain. This aim would not only make a stable strain for long term usage but also would avoids the gene horizontal transfer between non-targets CGB or pathogens. 4.4 Closing The bee microbiome is an approachable and rewarding subject of study that offers available platforms to answer a great variety of biological questions. Furthermore, the importance of the honeybees for the global economy and ecological equilibrium are strong reasons to study and manipulate them. The use of the commensal bacteria as modulators of health is expected to become commonplace in the coming years. Also, making bacteria able to secrete heterologous proteins is a very promising venue in biotechnology and pharmaceuticals, but could be exploited not only for industrial approaches but for medicinal ones, such as drug delivery and pathogen modulation. This project offered the one of the very first attempts for this purpose in the agrobiotechnology sector. 89 References 1. Pollination Facts - American Beekeeping Federation. Abfnet.org (2020) at . 2. Bixby, C., Guarna, M. M. & Hoover, S. E. Canadian Honey Bee Queen Breeders\u00E2\u0080\u0099 Reference Guide. Canadian Association of Professional Apiculturists Publication 55 (2019). 3. National Research Council. Status of Pollinators in North America. The National Academies Press 326 (2007). 4. Oldroyd, B. P. & Thompson, G. J. Behavioural Genetics of the Honey Bee Apis mellifera. Advances in Insect Physiology 33 (2006). 5. Galizia, C. G., Eisenhardt, D. & Giurfa, M. Honey Bee Neurobiology and Behaviour. A Tribute to Randolf Menzel. Honeybee Neurobiology and Behavior (2012). doi:10.1007/978-94-007-2099-2. 6. D\u00C3\u00B6tterl, S. & Vereecken, N. J. The chemical ecology and evolution of bee-flower interactions: A review and perspectives. Can. J. Zool. 88, 668\u00E2\u0080\u0093697 (2010). 7. Kwong, W. K. et al. Dynamic microbiome evolution in social bees. Sci. Adv. 3, 1\u00E2\u0080\u009317 (2017). 8. The Honeybee Genome Sequencing Cosortium. Insights into social insects from the genome of the honeybee Apis mellifera. Nature 443, 932\u00E2\u0080\u0093949 (2006). 9. Cornman, R. S. et al. The Bee Microbiome: Impact on Bee Health and Model for Evolution and Ecology of Host-Microbe Interactions. MBio 7, 1\u00E2\u0080\u00939 (2016). 10. Raymann, K. & Moran, N. A. The role of the gut microbiome in health and disease of adult honey bee workers. Curr. Opin. Insect Sci. 26, 97\u00E2\u0080\u0093104 (2018). 90 11. Zheng, H., Steele, M. I., Leonard, S. P., Motta, E. V. S. & Moran, N. A. Honey bees as models for gut microbiota research. Lab Animal vol. 47 317\u00E2\u0080\u0093325 (2018). 12. Romero, S., Nastasa, A., Chapman, A., Kwong, W. K. & Foster, L. J. The honey bee gut microbiota: strategies for study and characterization. Insect Mol. Biol. 28, 455\u00E2\u0080\u0093472 (2019). 13. Winston, M. L. The Biology of the Honey Bee. Harvard University Press 281 (1991). 14. Shi, Y. Y. et al. Diet and cell size both affect queen-worker differentiation through DNA methylation in honey bees (apis mellifera, apidae). PLoS One 6, 2\u00E2\u0080\u00937 (2011). 15. Kucharski, R., Maleszka, J., Foret, S. & Maleszka, R. Nutritional control of reproductive status in honeybees via DNA methylation. Science. 319, 1827\u00E2\u0080\u00931830 (2008). 16. Bortolotti, L. & Costa, C. Chemical communication in the honey bee society. in Neurobiology of Chemical Communication 147\u00E2\u0080\u0093210 (2014). doi:10.1201/b16511. 17. Ilyasov, R. A. & Kwon, H. W. Phylogenetics of Bees. Phylogenetics of Bees (2019). doi:10.1201/b22405. 18. Slater, G. P., Yocum, G. D. & Bowsher, J. H. Diet quantity influences caste determination in honeybees (Apis mellifera). Proceedings. Biol. Sci. 287, 20200614 (2020). 19. Powell, J. E., Martinson, V. G., Urban-Mead, K. & Moran, N. A. Routes of acquisition of the gut microbiota of the honey bee Apis mellifera. Applied and Environmental Microbiology 80 7378\u00E2\u0080\u00937387 (2014). 20. Nicolson, S. W. Bee food: the chemistry and nutritional value of nectar, pollen and mixtures of the two. African Zool. 46, 197\u00E2\u0080\u0093204 (2011). 21. Wright, G. A., Nicolson, S. W. & Shafir, S. Nutritional Physiology and Ecology of Honey Bees. Annual Review of Entomology 63 327\u00E2\u0080\u0093344 (2018). 91 22. Goulson, D., Nicholls, E., Bot\u00C3\u00ADas, C. & Rotheray, E. L. Bee declines driven by combined Stress from parasites, pesticides, and lack of flowers. Science. 347, (2015). 23. Potts, S. G. et al. Global pollinator declines: Trends, impacts and drivers. Trends Ecol. Evol. 25, 345\u00E2\u0080\u0093353 (2010). 24. McAfee, A. et al. Vulnerability of honey bee queens to heat-induced loss of fertility. Nat. Sustain. 3, 367\u00E2\u0080\u0093376 (2020). 25. Oldroyd, B. P. Domestication of honey bees was associated with expansion of genetic diversity. Mol. Ecol. 21, 4409\u00E2\u0080\u00934411 (2012). 26. Brutscher, L. M., McMenamin, A. J. & Flenniken, M. L. The Buzz about Honey Bee Viruses. PLoS Pathog. 12, 1\u00E2\u0080\u00937 (2016). 27. Genersch, E. & Aubert, M. Emerging and re-emerging viruses of the honey bee (Apis mellifera L.). Vet. Res. 41, (2010). 28. Wilfert, L. et al. Honeybee disease: Deformed wing virus is a recent global epidemic in honeybees driven by Varroa mites. Science. 351, 594\u00E2\u0080\u0093597 (2016). 29. Genersch, E. Paenibacillus larvae and American Foulbrood - Long since known and still surprising. J. fur Verbraucherschutz und Leb. 3, 429\u00E2\u0080\u0093434 (2008). 30. Brosi, B. J., Delaplane, K. S., Boots, M. & De Roode, J. C. Ecological and evolutionary approaches to managing honeybee disease. Nat. Ecol. Evol. 1, 1250\u00E2\u0080\u00931262 (2017). 31. Guarna, M. M. et al. Peptide biomarkers used for the selective breeding of a complex polygenic trait in honey bees. Sci. Rep. 7, 1\u00E2\u0080\u009310 (2017). 32. Leonard, S. P. et al. Engineered symbionts activate honey bee immunity and limit pathogens. Science (80-. ). 576, 573\u00E2\u0080\u0093576 (2020). 33. Ebeling, J., Knispel, H., Hertlein, G., F\u00C3\u00BCnfhaus, A. & Genersch, E. Biology of 92 Paenibacillus larvae, a deadly pathogen of honey bee larvae. Appl. Microbiol. Biotechnol. 100, 7387\u00E2\u0080\u00937395 (2016). 34. Garcia-Gonzalez, E. et al. Biological effects of paenilamicin, a secondary metabolite antibiotic produced by the honey bee pathogenic bacterium Paenibacillus larvae. Microbiologyopen 3, 642\u00E2\u0080\u0093656 (2014). 35. M\u00C3\u00BCller, S., Garcia-Gonzalez, E., Genersch, E. & S\u00C3\u00BCssmuth, R. D. Involvement of secondary metabolites in the pathogenesis of the American foulbrood of honey bees caused by Paenibacillus larvae. Nat. Prod. Rep. 32, 765\u00E2\u0080\u0093778 (2015). 36. Erban, T. et al. Honeybee (Apis mellifera)-associated bacterial community affected by American foulbrood: Detection of Paenibacillus larvae via microbiome analysis. Sci. Rep. 7, 1\u00E2\u0080\u009310 (2017). 37. Garcia-Gonzalez, E. et al. Paenibacillus larvae Chitin-Degrading Protein PlCBP49 Is a Key Virulence Factor in American Foulbrood of Honey Bees. PLoS Pathog. 10, (2014). 38. Yue, D., Nordhoff, M., Wieler, L. H. & Genersch, E. Fluorescence in situ hybridization (FISH) analysis of the interactions between honeybee larvae and Paenibacillus larvae, the causative agent of American foulbrood of honeybees (Apis mellifera). Environ. Microbiol. 10, 1612\u00E2\u0080\u00931620 (2008). 39. F\u00C3\u00BCnfhaus, A. & Genersch, E. Proteome analysis of Paenibacillus larvae reveals the existence of a putative S-layer protein. Environ. Microbiol. Rep. 4, 194\u00E2\u0080\u0093202 (2012). 40. Reybroeck, W., Daeseleire, E., De Brabander, H. F. & Herman, L. Antimicrobials in beekeeping. Vet. Microbiol. 158, 1\u00E2\u0080\u009311 (2012). 41. Evans, J. D. Diverse origins of tetracycline resistance in the honey bee bacterial pathogen Paenibacillus larvae. J. Invertebr. Pathol. 83, 46\u00E2\u0080\u009350 (2003). 93 42. Tian, B., Fadhil, N. H., Powell, J. E., Kwong, W. K. & Moran, N. A. Long-term exposure to antibiotics has caused accumulation of resistance determinants in the gut microbiota of honeybees. MBio 3, 1\u00E2\u0080\u00937 (2012). 43. Abdullah, I., Gary, S. R. & Marla, S. controlling European foulbrood with the shook swarm method and oxytetracycline in the UK. Apidologie 38, 67\u00E2\u0080\u009376 (2007). 44. Shankar, Pr. Book review: Tackling drug-resistant infections globally. Arch. Pharm. Pract. 7, 110 (2016). 45. Kuzy\u00C5\u00A1inov\u00C3\u00A1, K. et al. Testing of inhibition activity of essential oils against Paenibacillus larvae - The causative agent of American foulbrood. Acta Vet. Brno 83, 9\u00E2\u0080\u009312 (2014). 46. Du\u00C5\u00A1an, F., Mari\u00C3\u00A1n, S., Katar\u00C3\u00ADna, D. & Dobroslava, B. Essential oils-their antimicrobial activity against Escherichia coli and effect on intestinal cell viability. Toxicol. Vitr. 20, 1435\u00E2\u0080\u00931445 (2006). 47. Borba, R. S. & Spivak, M. Propolis envelope in Apis mellifera colonies supports honey bees against the pathogen, Paenibacillus larvae. Sci. Rep. 7, 1\u00E2\u0080\u00936 (2017). 48. Alonso-Salces, R. M. et al. Natural strategies for the control of Paenibacillus larvae, the causative agent of American foulbrood in honey bees: a review. Apidologie 48 387\u00E2\u0080\u0093400 (2017). 49. Ant\u00C3\u00BAnez, K. et al. Efficacy of natural propolis extract in the control of American Foulbrood. Vet. Microbiol. 131, 324\u00E2\u0080\u0093331 (2008). 50. Vrese, M. De. Probiotics, prebiotics, and synbiotics\u00E2\u0080\u0094approaching a definition. Am. J. Clin. Nutr. 73, (2001). 51. Plaza-Diaz, J., Ruiz-Ojeda, F. J., Gil-Campos, M. & Gil, A. Mechanisms of Action of Probiotics. Adv. Nutr. 10, S49\u00E2\u0080\u0093S66 (2019). 94 52. Daisley, B. A. et al. Novel probiotic approach to counter Paenibacillus larvae infection in honey bees. ISME Journal 14 476\u00E2\u0080\u0093491 (2020). 53. Corby-Harris, V., Maes, P. & Anderson, K. E. The bacterial communities associated with honey bee (apis mellifera) foragers. PLoS One 9, (2014). 54. Khilnani, J. C. & Wing, H. J. Protocols to test the activity of antimicrobial peptides against the honey bee pathogen Paenibacillus larvae. Journal of Microbiological Methods 117 54\u00E2\u0080\u009356 (2015). 55. Brodsgaard, C. J., Ritter, W. & Hansen, H. Response of in vitro reared honey bee larvae to various doses of Paenibacillus larvae larvae spores. Apidologie 29, 569\u00E2\u0080\u0093578 (1998). 56. Riessberger-Gall\u00C3\u00A9, U., Von Der Ohe, W. & Crailsheim, K. Adult honeybee\u00E2\u0080\u0099s resistance against Paenibacillus larvae larvae, the causative agent of the American foulbrood. J. Invertebr. Pathol. 77, 231\u00E2\u0080\u0093236 (2001). 57. Yoshiyama, M. & Kimura, K. Bacteria in the gut of Japanese honeybee, Apis cerana japonica, and their antagonistic effect against Paenibacillus larvae, the causal agent of American foulbrood. J. Invertebr. Pathol. 102, 91\u00E2\u0080\u009396 (2009). 58. Chan, Q. W. T., Melathopoulos, A. P., Pernal, S. F. & Foster, L. J. The innate immune and systemic response in honey bees to a bacterial pathogen, Paenibacillus larvae. BMC Genomics 10 (2009). 59. Evans, J. D. Transcriptional immune responses by honey bee larvae during invasion by the bacterial pathogen, Paenibacillus larvae. J. Invertebr. Pathol. 85, 105\u00E2\u0080\u0093111 (2004). 60. Cox-Foster, D. L. et al. A metagenomic survey of microbes in honey bee colony collapse disorder. Science (80-. ). 318, 283\u00E2\u0080\u0093287 (2007). 61. Kwong, W. K. & Moran, N. A. Gut microbial communities of social bees. Nat. Rev. 95 Microbiol. 14, 374\u00E2\u0080\u0093384 (2016). 62. Martinson, V. G., Moy, J. & Moran, N. A. Establishment of characteristic gut bacteria during development of the honeybee worker. Applied and Environmental Microbiology 78, 2830\u00E2\u0080\u00932840 (2012). 63. Kwong, W. K., Mancenido, A. L. & Moran, N. A. Immune system stimulation by the native gut microbiota of honey bees. R. Soc. Open Sci. 4, 1\u00E2\u0080\u00939 (2017). 64. Wong, A. C. N., Chaston, J. M. & Douglas, A. E. The inconstant gut microbiota of Drosophila species revealed by 16S rRNA gene analysis. ISME J. 7, 1922\u00E2\u0080\u00931932 (2013). 65. Donaldson, G. P., Lee, S. M. & Mazmanian, S. K. Gut biogeography of the bacterial microbiota. Nat. Rev. Microbiol. 14, 20\u00E2\u0080\u009332 (2015). 66. Shin, S. C. et al. Drosophila microbiome modulates host developmental and metabolic homeostasis via insulin signaling. Science (80-. ). 334, 670\u00E2\u0080\u0093674 (2011). 67. Lemanceau, P., Blouin, M., Muller, D. & Mo\u00C3\u00ABnne-Loccoz, Y. Let the Core Microbiota Be Functional. Trends Plant Sci. 22, 583\u00E2\u0080\u0093595 (2017). 68. Engel, P., Martinson, V. G. & Moran, N. A. Functional diversity within the simple gut microbiota of the honey bee. Proc. Natl. Acad. Sci. U. S. A. 109, 11002\u00E2\u0080\u009311007 (2012). 69. Koch, H., Abrol, D. P., Li, J. & Schmid-Hempel, P. Diversity and evolutionary patterns of bacterial gut associates of corbiculate bees. Mol. Ecol. 22, 2028\u00E2\u0080\u00932044 (2013). 70. Powell, J. E., Leonard, S. P., Kwong, W. K., Engel, P. & Moran, N. A. Genome-wide screen identifies host colonization determinants in a bacterial gut symbiont. Proc. Natl. Acad. Sci. U. S. A. 113, 13887\u00E2\u0080\u009313892 (2016). 71. Onchuru, T. O., Javier Martinez, A., Ingham, C. S. & Kaltenpoth, M. Transmission of mutualistic bacteria in social and gregarious insects. Curr. Opin. Insect Sci. 28, 50\u00E2\u0080\u009358 96 (2018). 72. Ke\u00C5\u00A1nerov\u00C3\u00A1, L. et al. Disentangling metabolic functions of bacteria in the honey bee gut. PLoS Biol. 15, 1\u00E2\u0080\u009328 (2017). 73. Li, J. H. et al. New evidence showing that the destruction of gut bacteria by antibiotic treatment could increase the honey bee\u00E2\u0080\u0099s vulnerability to nosema infection. PLoS One 12, 1\u00E2\u0080\u009318 (2017). 74. Cremer, S., Armitage, S. A. O. & Schmid-Hempel, P. Social Immunity. Curr. Biol. 17, 693\u00E2\u0080\u0093702 (2007). 75. Evans, J. D. et al. Immune pathways and defence mechanisms in honey bees Apis mellifera. Insect Mol. Biol. 15, 645\u00E2\u0080\u0093656 (2006). 76. Raheem, N. & Straus, S. K. Mechanisms of Action for Antimicrobial Peptides With Antibacterial and Antibiofilm Functions. Frontiers in Microbiology 10 (2019). 77. Danihl\u00C3\u00ADk, J., Aronstein, K. & Pet\u00C5\u0099ivalsk\u00C3\u00BD, M. Antimicrobial peptides: A key component of honey bee innate immunity. Journal of Apicultural Research 54 123\u00E2\u0080\u0093136 (2015). 78. CASTEELS, P. et al. Isolation and characterization of abaecin, a major antibacterial response peptide in the honeybee (Apis mellifera). Eur. J. Biochem. 187, 381\u00E2\u0080\u0093386 (1990). 79. Casteels, P., Ampe, C., Jacobs, F., Vaeck, M. & Tempst, P. Apidaecins: antibacterial peptides from honeybees. EMBO J. 8, 2387\u00E2\u0080\u00932391 (1989). 80. Ilyasov, R. A., Gaifullina, L. R., Saltykova, E. S., Poskryakov, A. V. & Nikolaenko, A. G. Defensins in the honeybee antiinfectious protection. J. Evol. Biochem. Physiol. 49, 1\u00E2\u0080\u00939 (2013). 81. Casteels, P., Ampe, C., Jacobs, F. & Tempst, P. Functional and chemical characterization of hymenoptaecin, an antibacterial polypeptide that is infection-inducible in the honeybee 97 (Apis mellifera). J. Biol. Chem. 268, 7044\u00E2\u0080\u00937054 (1993). 82. Wehbe, R. et al. Bee venom: Overview of main compounds and bioactivities for therapeutic interests. Molecules 24, 1\u00E2\u0080\u009313 (2019). 83. Fontana, R. et al. Jelleines: A family of antimicrobial peptides from the Royal Jelly of honeybees (Apis mellifera). Peptides 25, 919\u00E2\u0080\u0093928 (2004). 84. Rappsilber, J., Ishihama, Y. and M. M. Stop and Go Extraction Tips for Matrix-Assisted Laser Desorption/Ionization, Nanoelectrospray, and LC/MS Sample Pretreatment in Proteomics. Nature 422, 198\u00E2\u0080\u0093207 (2003). 85. Chan, Q. W. T. & Foster, L. J. Changes in protein expression during honey bee larval development. Genome Biol. 9, (2008). 86. Li, Z. H. et al. Comparative proteomic analysis of differentially expressed proteins in the Bombyx mori fat body during the microsporidia Nosema bombycis infection. J. Invertebr. Pathol. 149, 36\u00E2\u0080\u009343 (2017). 87. Vodovar, N. et al. Drosophila host defense after oral infection by an entomopathogenic Pseudomonas species. Proc. Natl. Acad. Sci. U. S. A. 102, 11414\u00E2\u0080\u009311419 (2005). 88. Scharlaken, B., de Graaf, D. C., Goossens, K., Peelman, L. J. & Jacobs, F. J. Differential gene expression in the honeybee head after a bacterial challenge. Dev. Comp. Immunol. 32, 883\u00E2\u0080\u0093889 (2008). 89. Wiegand, I., Hilpert, K. & Hancock, R. E. W. Agar and broth dilution methods to determine the minimal inhibitory concentration (MIC) of antimicrobial substances. Nat. Protoc. 3, 163\u00E2\u0080\u0093175 (2008). 90. Raymann, K., Shaffer, Z. & Moran, N. A. Antibiotic exposure perturbs the gut microbiota and elevates mortality in honeybees. PLoS Biol. 15, 1\u00E2\u0080\u009322 (2017). 98 91. Motta, E. V. S., Raymann, K. & Moran, N. A. Glyphosate perturbs the gut microbiota of honey bees. Proc. Natl. Acad. Sci. U. S. A. 115, 10305\u00E2\u0080\u009310310 (2018). 92. Kwong, W. K. & Moran, N. A. Cultivation and characterization of the gut symbionts of honey bees and bumble bees: Description of Snodgrassella alvi gen. nov., sp. nov., a member of the family Neisseriaceae of the betaproteobacteria, and Gilliamella apicola gen. nov., sp. nov., a memb. Int. J. Syst. Evol. Microbiol. 63, 2008\u00E2\u0080\u00932018 (2013). 93. Lee, C., Kim, J., Shin, S. G. & Hwang, S. Absolute and relative QPCR quantification of plasmid copy number in Escherichia coli. J. Biotechnol. 123, 273\u00E2\u0080\u0093280 (2006). 94. Ram\u00C3\u00ADrez-Camejo, L. A., Maldonado-Morales, G. & Bayman, P. Differential Microbial Diversity in Drosophila melanogaster: Are Fruit Flies Potential Vectors of Opportunistic Pathogens? Int. J. Microbiol. 2017, (2017). 95. Chib, S., Das, S., Venkatesan, S., Seshasayee, A. S. N. & Thattai, M. Using stochastic cell division and death to probe minimal units of cellular replication. New J. Phys. 20, 35004 (2018). 96. Forabosco, F., L\u00C3\u00B6hmus, M., Rydhmer, L. & Sundstr\u00C3\u00B6m, L. F. Genetically modified farm animals and fish in agriculture: A review. Livest. Sci. 153, 1\u00E2\u0080\u00939 (2013). 97. Forkus, B., Ritter, S., Vlysidis, M., Geldart, K. & Kaznessis, Y. N. Antimicrobial Probiotics Reduce Salmonella enterica in Turkey Gastrointestinal Tracts. Scientific Reports 7 (2017). 98. Hwang, I. Y. et al. Engineered probiotic Escherichia coli can eliminate and prevent Pseudomonas aeruginosa gut infection in animal models. Nat. Commun. 8, 1\u00E2\u0080\u009311 (2017). 99. Jenssen, H., Hamill, P. & Hancock, R. E. W. Peptide antimicrobial agents. Clin. Microbiol. Rev. 19, 491\u00E2\u0080\u0093511 (2006). 99 100. Rodr\u00C3\u00ADguez-Rojas, A., Makarova, O. & Rolff, J. Antimicrobials, Stress and Mutagenesis. PLoS Pathog. 10, (2014). 101. Blair, J. M. A., Webber, M. A., Baylay, A. J., Ogbolu, D. O. & Piddock, L. J. V. Molecular mechanisms of antibiotic resistance. Nat. Rev. Microbiol. 13, 42\u00E2\u0080\u009351 (2015). 102. Le, C. F., Fang, C. M. & Sekaran, S. D. Intracellular targeting mechanisms by antimicrobial peptides. Antimicrobial Agents and Chemotherapy 61 (2017). 103. Deng, T. et al. The heterologous expression strategies of antimicrobial peptides in microbial systems. Protein Expr. Purif. 140, 52\u00E2\u0080\u009359 (2017). 104. Li, Y. Recombinant production of antimicrobial peptides in Escherichia coli: A review. Protein Expr. Purif. 80, 260\u00E2\u0080\u0093267 (2011). 105. Kleiner-Grote, G. R. M., Risse, J. M. & Friehs, K. Secretion of recombinant proteins from E. coli. Eng. Life Sci. 18, 532\u00E2\u0080\u0093550 (2018). 106. Freudl, R. Signal peptides for recombinant protein secretion in bacterial expression systems. Microb. Cell Fact. 17, 1\u00E2\u0080\u009310 (2018). 107. Filloux, A. Secretion signal and protein targeting in Bacteria: A biological puzzle. J. Bacteriol. 192, 3847\u00E2\u0080\u00933849 (2010). 108. Tsirigotaki, A., De Geyter, J., \u00C5\u00A0o\u00C5\u00A1tari\u00C4\u0087, N., Economou, A. & Karamanou, S. Protein export through the bacterial Sec pathway. Nat. Rev. Microbiol. 15, 21\u00E2\u0080\u009336 (2017). 109. Izard, J. W. & Kendall, D. A. Signal peptides: exquisitely designed transport promoters. Mol. Microbiol. 13, 765\u00E2\u0080\u0093773 (1994). 110. Burdette, L. A., Leach, S. A., Wong, H. T. & Tullman-Ercek, D. Developing Gram-negative bacteria for the secretion of heterologous proteins. Microb. Cell Fact. 17, 1\u00E2\u0080\u009316 (2018). 100 111. Mathiesen, G., Sveen, A., Piard, J. C., Axelsson, L. & Eijsink, V. G. H. Heterologous protein secretion by Lactobacillus plantarum using homologous signal peptides. J. Appl. Microbiol. 105, 215\u00E2\u0080\u0093226 (2008). 112. Boekhorst, J., Wels, M., Kleeberezem, M. & Siezen, R. J. The predicted secretome of Lactobacillus plantarum WCFS1 sheds light on interactions with its environment. Microbiology 152, 3175\u00E2\u0080\u00933183 (2006). 113. Fagerquist, C. K. & Zaragoza, W. J. Proteolytic Surface-Shaving and Serotype-Dependent Expression of SPI-1 Invasion Proteins in Salmonella enterica Subspecies enterica. Front. Nutr. 5, (2018). 114. Lim, H. S., Yeu, J. E., Hong, S. P. & Kang, M. S. Characterization of antibacterial cell-free supernatant from oral care probiotic weissella cibaria, CMU. Molecules 23, (2018). 115. Leonard, S. P. et al. Genetic Engineering of Bee Gut Microbiome Bacteria with a Toolkit for Modular Assembly of Broad-Host-Range Plasmids. ACS Synth. Biol. 7, 1279\u00E2\u0080\u00931290 (2018). 116. Almagro Armenteros, J. J. et al. SignalP 5.0 improves signal peptide predictions using deep neural networks. Nat. Biotechnol. 37, 420\u00E2\u0080\u0093423 (2019). 117. Rodr\u00C3\u00ADguez-Ortega, M. J. \u00E2\u0080\u009CShaving\u00E2\u0080\u009D live bacterial cells with proteases for proteomic analysis of surface proteins. Methods in Molecular Biology 1722 (2018). 118. Herbert, B. Advances in protein solubilisation for two-dimensional electrophoresis. Electrophoresis 20, 660\u00E2\u0080\u0093663 (1999). 119. Krey, J. F. et al. Accurate label-free protein quantitation with high- and low-resolution mass spectrometers. J. Proteome Res. 13, 1034\u00E2\u0080\u00931044 (2014). 120. Boxberger et al., 2008. how many signal peptides are there in bacteria? Bone 23, 1\u00E2\u0080\u00937 101 (2008). 121. Kwong, W. K., Engel, P., Koch, H. & Moran, N. A. Genomics and host specialization of honey bee and bumble bee gut symbionts. Proc. Natl. Acad. Sci. U. S. A. 111, 11509\u00E2\u0080\u009311514 (2014). "@en . "Thesis/Dissertation"@en . "2020-11"@en . "10.14288/1.0394247"@en . "eng"@en . "Genome Science and Technology"@en . "Vancouver : University of British Columbia Library"@en . "University of British Columbia"@en . "Attribution-NonCommercial-NoDerivatives 4.0 International"@* . "http://creativecommons.org/licenses/by-nc-nd/4.0/"@* . "Graduate"@en . "A prophylactic probiotic to fight paenibacillus larvae infection in honey bees"@en . "Text"@en . "http://hdl.handle.net/2429/75906"@en .